Delivered-To: terrellf@mail-dnvr.uswest.net Resent-Date: Sat, 20 Jul 2002 15:38:58 -0600 Resent-From: tesla@pupman.com Resent-Sender: tesla-request@pupman.com Date: Sat, 20 Jul 2002 15:33:24 -0600 From: "Tesla list" To: tesla@pupman.com X-Sender: twftesla@pop.dnvr.qwest.net X-Mailer: QUALCOMM Windows Eudora Pro Version 4.1 Subject: Off-Line Tesla coils (OLTC) X-Mailing-List: archive/latest/25071 X-Loop: tesla@pupman.com Original poster: "Terry Fritz" Hi All, I have been doing some studying of off-line Tesla coils. These are (would be) Tesla coils that have "no" HV transformers but run directly off line voltages. What I found so far (purely by computer modeling) is somewhat encouraging. While a far cry from a 38 foot streamer unit, it may easily approach an NST powered coil in power. The only real new components would be an IGBT solid state spark gap and rather odd MMC caps. Everything else is quite conventional. So far, the design would run off 240VAC at around 700 watts. This is the schematic as it now stands: http://hot-streamer.com/temp/Image7.gif The the top is 120-0-120 VAC (single phase split 240 as is common in US homes). That voltage is rectified by two diodes and charges two 100uF caps (50uF primary capacitance) through some ballast resistors. This gives a firing voltage at 60 BPS SYNC of: 240 x SQRT(2) x 2 = 679 VDC The capacitor waveform looks like this: http://hot-streamer.com/temp/Image3.gif Since the caps need a full AC cycle to charge, the natural firing rate is limited to 60 BPS. The current through the charging diodes looks like this: http://hot-streamer.com/temp/Image4.gif Only about 4 amps RMS and very easily in the range of any 240VAC circuit. One concern is the RF spikes at about 20 amps which would have to be filtered. The resonant frequency is 16 kHz. A sweep of the circuit revels a dual hump noting that the circuit is rather simplified so they appear higher Q than in real life. http://hot-streamer.com/temp/Image1.gif The current in the primary and secondary are shown in: http://hot-streamer.com/temp/Image6.gif The IGBT "spark gap" would have to withstand a 2000+ amp spike 60 times per second. I would quench on the first notch electronically to trap the energy in the secondary. This first notch quenching also reduced the stress on the IGBT and caps considerably. The primary RMS current is near 120 amps(!!) so turning off the primary as soon as possible really helps losses and stress on the primary caps and IGBT. The secondary currents are conventional. The primary coil would be a single turn of 1 inch copper pipe. Surprisingly, the MandK program predicts I can still get very good coupling with just a big single turn primary at about 0.15. I thought having only one primary turn would devastate the coupling but that does not appear to be the case! The secondary is at 2.17H (2170mH) which is much large than usual to accommodate the low Fo frequency. Aside from having many turns, it is conventional. Since the IGBT can switch with very low loss, the vast majority of the energy is sent to and trapped in the secondary coil. In this case, the energy is: 1/2 x 50uF x 679^2 = 11.52 joules At 60 BPS, the power delivered to the secondary is 11.52 x 60 = 691 watts which is more than most LTR 15/60 NST systems can claim. The models does seem to fall short in one area. The 16kHz the streamer load estimate model of 220k+1pF per foot of streamer capacitance seems to give a very light streamer load... This is a big issue since the estimate may be "right" and streamers operating at only 16kHz may not work well!! However, with 700 watts trapped in the secondary system, one would think it would go somewhere? A major theoretical hole exists here that at least "i" can't predict... The model gives a very lightly loaded coil output that I really do not trust for accuracy: http://hot-streamer.com/temp/Image5.gif So it seems to work in the computer ;-) No guarantee that it will "really" work, but it has passed the significant hurtle of working in computer models. The computer also eliminated a number of other designs in an afternoon that it would have taken months (and thousands of $$$) to eliminate by testing real hardware... Note that the Ls/Lp ratio is 1,000,000:1 so Vs/Vp =1000 or 679000 volts. This addresses Antonio's concern that the primary circuit must have very low inductance at such low voltages. Fortunately, the coupling was not a great concern so just a "one big turn" primary could work. The special concerns that computer modeling predicts are as follows: 1. The AC line is poorly isolated from the Fo frequency in the primary caps as the circuit now stands. The 16kHz can easily push high currents back into the line. One solution may be to replace the didoes with FETs that would switch off during firing. Simple filters may take care of it too. The charging circuit works, but it could use much refinement. 2. The primary capacitors would have to be 700VDC 120Arms 2500Apeak at 100uF. Since it is a voltage doubler, we have about 4X the amount of capacitor. Capacitor arrays can easily get to these specs but they get costly at about $300. 3. The primary resistance would have to be very low at about <0.05 ohm at 16kHz. With a big single turn primary "pipe" and a very close IGBT and cap, this seems very possible. Many thanks to Gary Lau for his excellent(!!) work on this at: http://www.laushaus.com/tesla/primary_resistance.htm 5. The primary currents are "high" suggesting a 2000 amp peak 1000V IGBT and a cap RMS rating of 120 amps. However, this seems to be well within the reach of existing devices. 10 Digikey IGBTs would do that and in quantity their price is very good. 4. Depending on what the streamer loading really turns out to be, the coupling may have to be adjusted for maximum power transfer. This should be simply accomplished by raising the secondary in any case. Special notes about such a coil are: 1. Being an IGBT spark gap coil, the gap is very low loss. The only sound it makes is the high currents softly "clicking" in the circuit from magnetic constriction. But the streamers should be very loud especially at 16kHz ;-) 2. With such a low AC input current, conventional (cheap) small fuses can be used allowing very fast fusing action if something goes wrong. 3. The low 16kHz is in the audio range. The low Fo frequency and the IGBT drive may vastly reduce radio interference. However, the low Fo frequency may make the output streamers considerably more dangerous. I am not sure how nerves react at 16kHz? Also the human body's impedance at this frequency may look like a very nice "load". Of course, there is no skin effect in any case. 4. No transformer, spark gap motor, or variac. 100+pounds lighter ;-)) The electronics can easily be run from a few 9 volt batteries. A truely "different" coil... The questions I have for those that may know: 1. Does anyone have any data on Tesla coils running at near 16kHz and their streamers? Do streamer still form ok? 2. Does the human body "feel" 16kHz (nice sine wave) high voltage signals? Not that I am planning on finding out the hard way ;-) 3. Does anyone know of a reliable source for say 1000V 2500Apeak IGBTs with built in reverse diode. Quantity one or two... Affordable... IGBT arrays may be the way to go here too... MMIGBT (sorry ;-)) 4. Any ideas on a better charging circuit? Note that the diodes are conducting during the firing of the gap at the negative part of the cycle and only the 20 ohms of resistance provides isolation (but pretty good). If all seems well I will try to make it. The caps and IGBTs will cost real money but not too bad. Since it is low voltage, I have a much better chance of not blowing up a lot of IGBTs. Of course, those big ones don't blow up easy... Hope this is of interest and at least amusing ;-)) Cheers, Terry Delivered-To: terrellf@mail-dnvr.uswest.net Resent-Date: Sat, 20 Jul 2002 17:47:06 -0600 Resent-From: tesla@pupman.com Resent-Sender: tesla-request@pupman.com Date: Sat, 20 Jul 2002 17:43:27 -0600 From: "Tesla list" To: tesla@pupman.com X-Sender: twftesla@pop.dnvr.qwest.net X-Mailer: QUALCOMM Windows Eudora Pro Version 4.1 Subject: Re: Off-Line Tesla coils (OLTC) X-Mailing-List: archive/latest/25081 X-Loop: tesla@pupman.com Original poster: "Jim Lux by way of Terry Fritz " This starts to look like a high rep rate xenon strobe circuit. There are a number of flash tubes that can take this kind of load (several hundred watts). I'd definitely look into photoflash type capacitors for this.. For the secondary, you might want those tapewound air core inductors about which there were some postings a few months back in an oil tank. Fair amount of inductance in a small package at reasonable price. I'll look for the link. Having built a few strobe circuits, i'd say you definitely want resonant charging here (without the "constant current" effect of a NST output, you'll dissipate as much power in the ballast resistors as you put into the TC). An offline voltage doubler (a 110:110 transformer could get you 240V into the doubler without having to wire 240V supply.. or if you live in the rest of the world, just plug it in) feeding a BIG reservoir cap (several tens of thousand uF, so the voltage doesn't sag much during the charging cycle (450V electrolytic caps turn up surplus all the time)) and a conventional choke/diode charging circuit. Then, you're not constrained to 60 Hz rep rate. (actually, maybe a double doubler might work off 110V (a quadrupler..) normally terrible performance, but in this application you don't care about regulation, and you can have a ton of stored energy to smooth the ripple. Actually.. Say you use a offline multiplier to get the primary voltage up a bit (a x10?), then you might be able to make your primary C smaller, raising your frequency a bit. ----- Original Message ----- From: "Tesla list" To: Sent: Saturday, July 20, 2002 2:33 PM Subject: Off-Line Tesla coils (OLTC) > Original poster: "Terry Fritz" > > Hi All, > > I have been doing some studying of off-line Tesla coils. These are (would > be) Tesla coils that have "no" HV transformers but run directly off line > voltages. What I found so far (purely by computer modeling) is somewhat > encouraging. While a far cry from a 38 foot streamer unit, it may easily > approach an NST powered coil in power. The only real new components would > be an IGBT solid state spark gap and rather odd MMC caps. Everything else > is quite conventional. So far, the design would run off 240VAC at around > 700 watts. This is the schematic as it now stands: > > > The questions I have for those that may know: > > 1. Does anyone have any data on Tesla coils running at near 16kHz and their > streamers? Do streamer still form ok? > > 2. Does the human body "feel" 16kHz (nice sine wave) high voltage signals? > Not that I am planning on finding out the hard way ;-) > > 3. Does anyone know of a reliable source for say 1000V 2500Apeak IGBTs with > built in reverse diode. Quantity one or two... Affordable... IGBT arrays > may be the way to go here too... MMIGBT (sorry ;-)) > > 4. Any ideas on a better charging circuit? Note that the diodes are > conducting during the firing of the gap at the negative part of the cycle > and only the 20 ohms of resistance provides isolation (but pretty good). > > > If all seems well I will try to make it. The caps and IGBTs will cost real > money but not too bad. Since it is low voltage, I have a much better > chance of not blowing up a lot of IGBTs. Of course, those big ones don't > blow up easy... > > Hope this is of interest and at least amusing ;-)) > > Cheers, > > Terry > > > > > > > > > > > > > > > > > > Delivered-To: terrellf@mail-dnvr.uswest.net Resent-Date: Sat, 20 Jul 2002 21:10:03 -0600 Resent-From: tesla@pupman.com Resent-Sender: tesla-request@pupman.com Date: Sat, 20 Jul 2002 21:06:42 -0600 From: "Tesla list" To: tesla@pupman.com X-Sender: twftesla@pop.dnvr.qwest.net X-Mailer: QUALCOMM Windows Eudora Pro Version 4.1 Subject: Re: Off-Line Tesla coils (OLTC) X-Mailing-List: archive/latest/25087 X-Loop: tesla@pupman.com Original poster: "by way of Terry Fritz " Hope this is of interest and at least amusing ;-)) > > Cheers, > > Terry All those newbies asking why you need a high-voltage transformer... We'll see if ya really do. I think this is very interesting, and I want to know everything about it. i wonder if anyone has attempted something like this before? Even if they failed, it would be interesting to know what setup they used, and what happened. Keep up the good work --------------------------------------- Jonathon Reinhart hot-streamer.com/jonathon Delivered-To: terrellf@mail-dnvr.uswest.net Resent-Date: Sun, 21 Jul 2002 16:29:42 -0600 Resent-From: tesla@pupman.com Resent-Sender: tesla-request@pupman.com Date: Sun, 21 Jul 2002 13:11:16 -0600 From: "Tesla list" To: tesla@pupman.com X-Sender: twftesla@pop.dnvr.qwest.net X-Mailer: QUALCOMM Windows Eudora Pro Version 4.1 Subject: Re: Off-Line Tesla coils (OLTC) X-Mailing-List: archive/latest/25111 X-Loop: tesla@pupman.com Original poster: "by way of Terry Fritz " Well i for one am deffinately interested!!! Matt G Delivered-To: terrellf@mail-dnvr.uswest.net Resent-Date: Mon, 22 Jul 2002 09:01:10 -0600 Resent-From: tesla@pupman.com Resent-Sender: tesla-request@pupman.com Date: Mon, 22 Jul 2002 07:19:09 -0600 From: "Tesla list" To: tesla@pupman.com X-Sender: twftesla@pop.dnvr.qwest.net X-Mailer: QUALCOMM Windows Eudora Pro Version 4.1 Subject: Re: Off-Line Tesla coils (OLTC) X-Mailing-List: archive/latest/25164 X-Loop: tesla@pupman.com Original poster: "boris petkovic by way of Terry Fritz " Hi, > > 1. Does anyone have any data on Tesla coils > running at near 16kHz and > their > > streamers? Do streamer still form ok? --- They should form "ok".For that kind of fr you should have a big coil with a considerable power or relativelly smaller resonator with many turns and "large topload".If smaller coil choosen for design I think it will have somewhat shorter spark lenght compared with usual coil of same size for a given input power. One way or another higher bps are not recomended due to various reasons for that low resonant systems. ---- > > > > 2. Does the human body "feel" 16kHz (nice sine > wave) high voltage signals? --- Oho-ho it does.150 mA rms & 16 khz you'll not be able of willingly letting go a grasped electrode. regards, Boris __________________________________________________ Do You Yahoo!? Yahoo! Health - Feel better, live better http://health.yahoo.com Delivered-To: terrellf@mail-dnvr.uswest.net Resent-Date: Mon, 22 Jul 2002 18:06:32 -0600 Resent-From: tesla@pupman.com Resent-Sender: tesla-request@pupman.com Date: Mon, 22 Jul 2002 18:02:14 -0600 From: "Tesla list" To: tesla@pupman.com X-Sender: twftesla@pop.dnvr.qwest.net X-Mailer: QUALCOMM Windows Eudora Pro Version 4.1 Subject: Off-Line Tesla coils (OLTC) X-Mailing-List: archive/latest/25189 X-Loop: tesla@pupman.com Original poster: "Terry Fritz" Hi All, Today I substantially rethought the front end of my off-line Tesla coil. The schematic is at: http://hot-streamer.com/temp/Image1-1.gif Although a bit "rough" at the moment, the basic system seems sound. The cap values in the old design were simply "poor" choices considering commonly available caps. The advantages of this change are as follows: 1. The AC line is now very highly filtered from the coil's Fo frequency. No line filtering is really needed. 2. The primary cap cost is substantially reduced. Ten CD-940C6W4P7K series caps would work. Such a system would probably be tuned by adding or removing primary caps since the super heavy duty single turn primary "coil" can't be tapped for tuning. 3. The BPS rate can arbitrarily be anything the rest of the circuit could support. Since it is now a true DC charger, there is no need to worry with 60 or 120Hz BPS rates. It is far simpler theoretically now that the old way which was tuned to the line frequency. 4. There is now no real power limit. :-)) If you beef up the components, 10 to 20+ kW operation is possible. True pole pig power levels are relatively easy to obtain. You basically just get larger storage caps and increase the BPS rate (turn the knob up on the 555 timer). You could simply add components to an existing system to increase the power. Perhaps just bigger fuses and more electrolytic caps since the other parts could already be made high power with little added trouble. 5. The caps are now relatively easy to get new or surplus. The primary caps need only be cheap metalized ones since so many are in parallel. However, high BPS rates may need metal foil types. Not a big deal either way. Large electrolytics are common surplus items, but DigiKey sells these too ;-)) 6. The IGBT really does not care if you increase the BPS rate, so nothing else really changes. You may be able to go to a smaller primary cap to increase the Fo frequency. Its value no longer depends on the input circuit. 7. The primary tank circuit is no longer loaded by the charging circuit. That allows more and cleaner energy and more power throughput. The only real disadvantage is that the two large 5600uF caps (250 joules each) need a lot of care. Fusing, soft start, safety discharging, etc. is needed (don't wire them backwards ;-)). The large caps values are really not necessary and are arbitrary, but we need HIGH ripple currents. That naturally calls for thousands of uF of capacitance in common big caps. Ripple current is their key specification here. Right now, the system as shown should do 6+ foot streamers (very theoretical). The ESR in the caps and primary coil are very low. With the IGBT operating very efficiently, only the streamers should be burning off much power. I do have a bunch of power resistors in there too, but they are mostly for protection in case something in "Rev A" should go bad... Eventually, they should be removed. Much of the control circuits would be based on my previous SSgap work with the electronics hand held and connect via fail safe fiber-optics. Probably IR glass fiber since the added cost would probably we worth it. Thanks to Jim Lux for some great ideas! BTW - I just could not get photoflash caps to work. The specs just did not "fit" and they are polarized which was mess to deal with. Cheers, Terry >Hi All, > >I have been doing some studying of off-line Tesla coils. These are (would >be) Tesla coils that have "no" HV transformers but run directly off line >voltages. What I found so far (purely by computer modeling) is somewhat >encouraging. While a far cry from a 38 foot streamer unit, it may easily >approach an NST powered coil in power. The only real new components would >be an IGBT solid state spark gap and rather odd MMC caps. Everything else >is quite conventional. So far, the design would run off 240VAC at around >700 watts. This is the schematic as it now stands: > >http://hot-streamer.com/temp/Image7.gif > >The the top is 120-0-120 VAC (single phase split 240 as is common in US >homes). That voltage is rectified by two diodes and charges two 100uF caps >(50uF primary capacitance) through some ballast resistors. This gives a >firing voltage at 60 BPS SYNC of: > >240 x SQRT(2) x 2 = 679 VDC > >The capacitor waveform looks like this: > >http://hot-streamer.com/temp/Image3.gif > >Since the caps need a full AC cycle to charge, the natural firing rate is >limited to 60 BPS. The current through the charging diodes looks like this: > >http://hot-streamer.com/temp/Image4.gif > >Only about 4 amps RMS and very easily in the range of any 240VAC circuit. >One concern is the RF spikes at about 20 amps which would have to be >filtered. The resonant frequency is 16 kHz. A sweep of the circuit revels snip.... Delivered-To: terrellf@mail-dnvr.uswest.net Resent-Date: Mon, 22 Jul 2002 19:13:30 -0600 Resent-From: tesla@pupman.com Resent-Sender: tesla-request@pupman.com Date: Mon, 22 Jul 2002 19:08:45 -0600 From: "Tesla list" To: tesla@pupman.com X-Sender: twftesla@pop.dnvr.qwest.net X-Mailer: QUALCOMM Windows Eudora Pro Version 4.1 Subject: Re: Off-Line Tesla coils (OLTC) X-Mailing-List: archive/latest/25194 X-Loop: tesla@pupman.com Original poster: "Jim Lux by way of Terry Fritz " Tesla list wrote: > > Original poster: "Terry Fritz" > > Hi All, > > Today I substantially rethought the front end of my off-line Tesla coil. > The schematic is at: > > http://hot-streamer.com/temp/Image1-1.gif > > 3. The BPS rate can arbitrarily be anything the rest of the circuit could > support. Since it is now a true DC charger, there is no need to worry with > 60 or 120Hz BPS rates. It is far simpler theoretically now that the old > way which was tuned to the line frequency. Don't you need a diode in series with L35 and L34, as in a typical resonant charging circuit.. There's also no need for both L35 and L34 (since they are in series.. and the same goes for R63 and R64) just one 20 mH choke and 1 40 ohm resistor should do it. OTOH maybe you've got the components sitting around... > > 6. The IGBT really does not care if you increase the BPS rate, so nothing > else really changes. You may be able to go to a smaller primary cap to > increase the Fo frequency. Its value no longer depends on the input circuit. > > 7. The primary tank circuit is no longer loaded by the charging circuit. > That allows more and cleaner energy and more power throughput. > > The only real disadvantage is that the two large 5600uF caps (250 joules > each) need a lot of care. Fusing, soft start, safety discharging, etc. is > needed (don't wire them backwards ;-)). The large caps values are really > not necessary and are arbitrary, but we need HIGH ripple currents. That > naturally calls for thousands of uF of capacitance in common big caps. > Ripple current is their key specification here. The series resistors in the charging might give you soft start.... The 2 ohm ones you've got will limit the current to around 100A, and the caps will charge quickly. Granted the little 1Amp 1N400x rectifiers might pop, but an off the shelf 25-40Amp 600PIV bridge (use only half) probably has a healthy enough Isurge that it would work. Or, a couple of stud mount diodes. Another trick is the "100W light bulb in series" one (probably about 10 ohms, cold), or a relay/resistor soft start. > > Cheers, > > Terry > > >Hi All, > > > > Delivered-To: terrellf@mail-dnvr.uswest.net Resent-Date: Mon, 22 Jul 2002 20:21:14 -0600 Resent-From: tesla@pupman.com Resent-Sender: tesla-request@pupman.com Date: Mon, 22 Jul 2002 20:17:07 -0600 From: "Tesla list" To: tesla@pupman.com X-Sender: twftesla@pop.dnvr.qwest.net X-Mailer: QUALCOMM Windows Eudora Pro Version 4.1 Subject: Re: Off-Line Tesla coils (OLTC) X-Mailing-List: archive/latest/25200 X-Loop: tesla@pupman.com Original poster: "Terry Fritz" Hi Jim, At 05:20 PM 7/22/2002 -0700, you wrote: > > >Tesla list wrote: >> >> Original poster: "Terry Fritz" >> >> Hi All, >> >> Today I substantially rethought the front end of my off-line Tesla coil. >> The schematic is at: >> >> http://hot-streamer.com/temp/Image1-1.gif >> >> 3. The BPS rate can arbitrarily be anything the rest of the circuit >could >> support. Since it is now a true DC charger, there is no need to worry with >> 60 or 120Hz BPS rates. It is far simpler theoretically now that the old >> way which was tuned to the line frequency. > >Don't you need a diode in series with L35 and L34, as in a typical resonant >charging circuit.. Do you mean have a inductor in series (sort of) with the primary cap (or buffer cap) and drive the voltages up by resonating the thing at 60Hz off the Ac line? I never considered that but I guess you could run up some darn high voltages that way... Hehehe But don't you need several cycle before the voltage starts to get really high? I guess I don't quite understand how you mean for it to work? > >There's also no need for both L35 and L34 (since they are in series.. and >the same goes for R63 and R64) just one 20 mH choke and 1 40 ohm resistor >should do it. I do that so the voltages and such in the model stay balanced. It just makes the computer models a little easier to work with. > >OTOH maybe you've got the components sitting around... > > >> >> 6. The IGBT really does not care if you increase the BPS rate, so >nothing >> else really changes. You may be able to go to a smaller primary cap to >> increase the Fo frequency. Its value no longer depends on the input circuit. >> >> 7. The primary tank circuit is no longer loaded by the charging circuit. >> That allows more and cleaner energy and more power throughput. >> >> The only real disadvantage is that the two large 5600uF caps (250 joules >> each) need a lot of care. Fusing, soft start, safety discharging, etc. is >> needed (don't wire them backwards ;-)). The large caps values are really >> not necessary and are arbitrary, but we need HIGH ripple currents. That >> naturally calls for thousands of uF of capacitance in common big caps. >> Ripple current is their key specification here. > >The series resistors in the charging might give you soft start.... The 2 ohm >ones you've got will limit the current to around 100A, and the caps will >charge quickly. Granted the little 1Amp 1N400x rectifiers might pop, but an >off the shelf 25-40Amp 600PIV bridge (use only half) probably has a healthy >enough Isurge that it would work. Or, a couple of stud mount diodes. > >Another trick is the "100W light bulb in series" one (probably about 10 >ohms, cold), or a relay/resistor soft start. The 1N4002 is the only diode supplied in the "free version" of MicroSim. I modified the spice internal model so it acts like a big high power diode. A little modeling kludge... I think DigiKey has some 1200V 100 amp high speed puppies ;-) I think for a lower powered coil you could get away with resistors. However, I think I will use a lot of relays and switched resistors. I am sure it will work and I don't want to have too many things to "fiddle" with in the "first" one. Having worked with a bunch of prototype high power stuff in my day, I like the thought of a bunch of big contactors slamming home when I hit the "STOP!!" switch ;-)) I'll probably over design and spend a lot more money just to make things more R&D safe and friendly since I really don't know what I am doing ;-)) If it all works, then we can worry about subtle stuff ;-) However, I do want to get the circuits basically realistic and practical so anyone can do it. Cheers, Terry Delivered-To: terrellf@mail-dnvr.uswest.net Resent-Date: Mon, 22 Jul 2002 22:16:24 -0600 Resent-From: tesla@pupman.com Resent-Sender: tesla-request@pupman.com Date: Mon, 22 Jul 2002 22:13:34 -0600 From: "Tesla list" To: tesla@pupman.com X-Sender: twftesla@pop.dnvr.qwest.net X-Mailer: QUALCOMM Windows Eudora Pro Version 4.1 Subject: Re: Off-Line Tesla coils (OLTC) X-Mailing-List: archive/latest/25209 X-Loop: tesla@pupman.com Original poster: "Jim Lux by way of Terry Fritz " > >> 3. The BPS rate can arbitrarily be anything the rest of the circuit > >could > >> support. Since it is now a true DC charger, there is no need to worry with > >> 60 or 120Hz BPS rates. It is far simpler theoretically now that the old > >> way which was tuned to the line frequency. > > > >Don't you need a diode in series with L35 and L34, as in a typical resonant > >charging circuit.. > > Do you mean have a inductor in series (sort of) with the primary cap (or > buffer cap) and drive the voltages up by resonating the thing at 60Hz off > the Ac line? I never considered that but I guess you could run up some > darn high voltages that way... Hehehe No, just in series with the charging inductor (10 mH) that you have. Check out Richie Burnett's site for a good explanation. > > But don't you need several cycle before the voltage starts to get really > high? I guess I don't quite understand how you mean for it to work? > > > > >There's also no need for both L35 and L34 (since they are in series.. and > >the same goes for R63 and R64) just one 20 mH choke and 1 40 ohm resistor > >should do it. > > I do that so the voltages and such in the model stay balanced. It just > makes the computer models a little easier to work with. > > > > >OTOH maybe you've got the components sitting around... > > > > > >> > > > The 1N4002 is the only diode supplied in the "free version" of MicroSim. I > modified the spice internal model so it acts like a big high power diode. > A little modeling kludge... I think DigiKey has some 1200V 100 amp high > speed puppies ;-) Don't need high speed.. it's 60 Hz.. > Delivered-To: terrellf@mail-dnvr.uswest.net Resent-Date: Tue, 23 Jul 2002 20:29:57 -0600 Resent-From: tesla@pupman.com Resent-Sender: tesla-request@pupman.com Date: Tue, 23 Jul 2002 20:26:43 -0600 From: "Tesla list" To: tesla@pupman.com X-Sender: twftesla@pop.dnvr.qwest.net X-Mailer: QUALCOMM Windows Eudora Pro Version 4.1 Subject: Off-Line Tesla coils (OLTC) X-Mailing-List: archive/latest/25264 X-Loop: tesla@pupman.com Original poster: "Terry Fritz" Hi All, Today I substantially rethought the front end of my off-line Tesla coil, again... With a great clue from Jim Lux ;-)), I checked out Richie's site... The "present" idea is at: http://hot-streamer.com/temp/IGBTcoil01.gif I did away with the big storage caps and the about $400 worth of control equipment (many big contactors) it needed. It just was not working well. The big line caps were starting to need a lot of safety critical items and boring stuff like that which would make it hard for the average fellow to make and run safely. The power resistors in the old design were cooking off a tremendous amount of heat and I could not get around it easily. The source power vs. load power thing... You can get to very high power the old way with resonant charging, but things start to get expensive and hard. I am going for the easiest natural way to make an off-line Tesla coil, and leaving it to the future to find great improvements. Just have to get a nice first working one in the beginning. This design looks very promising! It filters the line and even has some power factor correction through C4. The firing voltage is only about 540 volts but worth the vast reduction in complexity (nice range for cheap IGBTs!) When you get down into this "conventional" voltage range, a giant arsenal of devices becomes available!! All the voltages and currents look good but timing and such is really important. Not a big deal since it is fully electronically controlled anyway. No big power storage elements and no heat dissipaters! Should easily be the most efficient Tesla coil out there. Nothing should get hot (or even warm) except the streamers ;-))) Basically, a bridge rectifier made from some big power diodes (probably in range of a typical bridge rectifier now). C4 acts to help filter the line and power factor. The value is fairly arbitrary. L1, L2 and C3 are resonant tuned to 60Hz to get the voltage up to about 530 volts on a single cycle (it is running at 120BPS). The current through L1 and L2 is about 6 amps peak and sort of sinusoidal. Probably a simple to make pair of inductors. The value is important. Lots of modeling work is still needed especially regarding "when things go bad..." I looked into IGBTs today. The big single brick ones run about $700!! However, if we get "cheap" IGBTs from DigiKey, we can parallel up the equivalent for about $100 (15 X 200 amp peak TO-247s) . The cheap IGBTs have more loss and such, but they are 1/3 the cost of the best ones and they should work perfectly fine. A "MMIGBT", (sorry :-))) I think I can make the primary loop less than 1uH now bit still with about a 0.2 coupling factor. The caps and IGBTs are actually part of the inductor's loop! This would allow higher frequencies and/or greater power. I think I would like higher frequency at this point. Maybe close to Fo = 30kHz... Have to watch out for induced voltage in the gates of the IGBTs, but those nice driver chips are good at that. The coil's power can be estimated. A 50uF primary cap firing at 120 BPS is: 1/2 x 50E-6 x 540^2 = 7.29 Joules per bang. At 120 BPS, we get a primary power of 7.29 x 120 = 875 watts. Assuming 90% of that makes it to the streamers, the streamer power is 790 watts. If a normal coil is 60% efficient, that would be equivalent to a 1300 watt coil. Using John's formula, that implies a streamer length of 60 inches. Not bad at all and hopefully it will be fairly simple to make. It's ALL a bit "imaginary" right now, but time will tell... BTW - Without MicroSim, the Internet, and friends on the list, I don't think a design like this would be possible! There is zero "prior art". MicroSim can crunch through idea after idea for only pennies in electricity, but it can find subtle fatal flaws in a design instantly! It would take weeks and a lot of cash to prove any one of these designs was worthless without it! With the Internet, I can suck up data sheets, component values, who has it in stock, and for how much money in a few minutes. This stuff relies an a lot of little details I could not find "fast" without the Internet. Of course, many thanks to all 800+ of you for little hints and tips! Jim's tips have been especially helpful! Nothing like having a few hundred experts helping with a project like this ;-)) Hopefully, it will all work out well and we will soon have a brand-new, high- tech, class of Tesla coil to play with ;-)) Cheers, Terry Delivered-To: terrellf@mail-dnvr.uswest.net Resent-Date: Fri, 5 Jul 2002 18:18:02 -0600 Resent-From: tesla@pupman.com Resent-Sender: tesla-request@pupman.com Date: Fri, 05 Jul 2002 18:12:03 -0600 From: "Tesla list" To: tesla@pupman.com X-Sender: twftesla@pop.dnvr.qwest.net X-Mailer: QUALCOMM Windows Eudora Pro Version 4.1 Subject: Okay.. Tesla Cabinets lower price... X-Mailing-List: archive/latest/24200 X-Loop: tesla@pupman.com Original poster: "BunnyKiller by way of Terry Fritz " Hi All... Maybe I asked too much for the cabinets ... I had no responces ... Sooo.. here ya go... I have 2 very nice cabinets available .... all it will cost is shipping I REALLY hate to see them sitting on the curb for trash pick-up....... Go see them at http://members.telocity.com/~bigfoo39/ goto bottom of page to Tesla cabinets for sale forget about the 40$ price tag.... these things are too nice to pass up or even worse.... trash ....... Scot D Delivered-To: terrellf@mail-dnvr.uswest.net Resent-Date: Sat, 10 Aug 2002 07:04:00 -0600 Resent-From: tesla@pupman.com Resent-Sender: tesla-request@pupman.com Date: Sat, 10 Aug 2002 06:51:05 -0600 From: "Tesla list" To: tesla@pupman.com X-Sender: twftesla@pop.dnvr.qwest.net X-Mailer: QUALCOMM Windows Eudora Pro Version 4.1 Subject: OLTC X-Mailing-List: archive/latest/667 X-Loop: tesla@pupman.com Original poster: "Mr Gregory Peters by way of Terry Fritz " Terry, I must say that I just cannot wait to see if your OLTC works! I'm probably more excited about it than you are :) It's good to see someone actually trying an idea out, rather than dismissing it as impossible at the theorising phase. I wish you all the best in your conquest. I am/was currently designing and collecting parts for a solid state CW coil, but I'm going to hold off for the moment, and see how your OLTC works - if it does, your system will be cloned in no time :) Good luck and best wishes, Greg Peters Department of Earth Sciences, University of Queensland, Australia Phone: 0402 841 677 http://www.geocities.com/gregjpeters Delivered-To: terrellf@mail-dnvr.uswest.net Resent-Date: Sat, 10 Aug 2002 08:56:24 -0600 Resent-From: tesla@pupman.com Resent-Sender: tesla-request@pupman.com Date: Sat, 10 Aug 2002 07:41:14 -0600 From: "Tesla list" To: tesla@pupman.com X-Sender: twftesla@pop.dnvr.qwest.net X-Mailer: QUALCOMM Windows Eudora Pro Version 4.1 Subject: Re: OLTC X-Mailing-List: archive/latest/686 X-Loop: tesla@pupman.com Original poster: "by way of Terry Fritz " In a message dated 8/10/02 6:01:58 AM Pacific Daylight Time, tesla@pupman.com writes: << Terry, I must say that I just cannot wait to see if your OLTC works! >> Hey Terry, I'm going to have to show my ignorance and ask what the "OL" in OLTC stands for? I have started following this thread with in- terest but can't recall what the "OL" stands for for :-/ David Rieben Delivered-To: terrellf@mail-dnvr.uswest.net Resent-Date: Sat, 10 Aug 2002 08:54:47 -0600 Resent-From: tesla@pupman.com Resent-Sender: tesla-request@pupman.com Date: Sat, 10 Aug 2002 08:48:04 -0600 From: "Tesla list" To: tesla@pupman.com X-Sender: twftesla@pop.dnvr.qwest.net X-Mailer: QUALCOMM Windows Eudora Pro Version 4.1 Subject: Re: OLTC X-Mailing-List: archive/latest/685 X-Loop: tesla@pupman.com Original poster: "Terry Fritz" Hi David, At 09:16 AM 8/10/2002 -0400, you wrote: > >Hey Terry, > >I'm going to have to show my ignorance and ask what the "OL" >in OLTC stands for? I have started following this thread with in- >terest but can't recall what the "OL" stands for for :-/ > >David Rieben > No worries, if this works, that question may get asked again someday :o)))) It stands for "Off-Line". This coil runs right 'off line' voltage AC without step up transformers. The primary circuit is not 10,000 to 20,000 volts like in a conventional coil but rather about 500 volts resonated up from the 240 VAC line voltage. http://hot-streamer.com/temp/OLTC8-8-01.gif (BTW - I think hot-streamer.com will be working much better now ;-)) In this diagram (a littles sketchy I guess, prolly should make a nice one...) the top two circles are the AC line voltage like from a dryer outlet. 120 - 0 - 120 split single phase. That voltage is simply rectified with a bridge rectifier and feed through the inductors to C7 which is the main primary cap. It is like 28uF in this case since the voltage is only about 500 volts. The switch in the lower left is an IGBT array in real life that can switch the 500 volts with very low loss (2500 amps!). The rest of the coil is conventional but the primary is only a single big turn and the secondary has many more turns since the operating frequency is lower. This shows the main waveforms: http://hot-streamer.com/temp/OLTC-08-07-1.gif The green is the AC from the rectifiers, the red is the voltage on the main cap, and the blue is 100X the charging current. The advantage is that there are no transformers, variacs, PFC caps... I am trying to get the entire system (control cabinet, and coil, to weigh very little, like 30 pounds) The main gap is solid state and eliminates the giant loss of a conventional spark gap. Therefor, if this coil has 700 watts input, that is easily equivalent to a 1kW+ conventional coil. If the power is scaled up, there is very little additional weight. I think such a system could do 7 foot arcs if the cap and IGBT array are beefed up to take about 5000 amps peak. But that is all one would have to do. There really is just not much "stuff" to such a coil and the theory is actually simpler but "new". Right now, it is just all theory and computer models but I am trying to build it up to see if it really works. It really "has" to work. It just a matter of getting components to do what the models says they have to do. Modern caps and IGBTs seem to be very capable of the task. I should make a nice web page and explain all the details in a nice format. But I could spend my time doing that or building the coil. So... this is all you get :o)) I have kept and reported all the details to the list if you sort for posts beginning with "OLTC". BTW - Some of us remember the days when people said computers could never predict a Tesla coil's functioning (and you know who you are! :o)))). Now it seems we can't build Tesla coils without them ;-)) LTR coils were first predicted by computer, but we would have figured it out fairly quickly without them. Computers certainly figured out the best component values for them far faster than say "trial and error..." However, this OLTC thing would have been virtually impossible to "whip up" without a LOT of computer power. There are some pretty tight regions it needs to work in and specs it has to meet to operate. One could never figure out how to get every thing to "sing" together without lots of study by computer. Also, many of the ideas that make it work, where from others here on the list. I sort of had the idea, but great ideas from many people could be quickly gathered here that really provided the set of keys to make it run. On-line data sheets and parts ordering saves a few months too :-)) If this works, it will truly be a modern state of the art Tesla coil that will be very new and different. It may also be far better in many ways. If we can get it at 30 pounds, keep it simple, $400, and seven foot sparks... It would "change things"... Cheers, Terry Delivered-To: terrellf@mail-dnvr.uswest.net Resent-Date: Sun, 11 Aug 2002 10:49:45 -0600 Resent-From: tesla@pupman.com Resent-Sender: tesla-request@pupman.com Date: Sun, 11 Aug 2002 10:35:29 -0600 From: "Tesla list" To: tesla@pupman.com X-Sender: twftesla@pop.dnvr.qwest.net X-Mailer: QUALCOMM Windows Eudora Pro Version 4.1 Subject: OLTC Thoughts X-Mailing-List: archive/latest/749 X-Loop: tesla@pupman.com Original poster: "Wall Richard Wayne by way of Terry Fritz " Terry, Here are some other OLTC design thoughts. Your TC primary, 47 mF cap(s) and IGBT switch(s) form a large LC tank. Once fired this thank will have huge oscillatory currents. What are the reverse stand off voltages of the IGBTs? Hopefully they have a built in protection diode. But, I would not depend on this small diode, if it exists, for protection. Which brings up another point. Richard Hull's H2 thyratron magnifier, by necessity, quenched at the first zero crossing of the wave form. So only the positive half of the energy cycle was passed to the primary. While a very interesting design, it is not a spectacular performer, if one measures performance by spark length. It's a phenomenal research device though. You may consider adding free running back diodes across your IGBTs for protection and for a full one cycle energy transfer. Have you calculated the resonate oscillatory frequency of the primary tank without the switches? Also consider moving the switch(s) outside the LC tank altogether to allow the tank to oscillate freely and transfer even more energy before you electronically quench it. I'll leave these design thoughts up to your fertile electronics imagination. RWW > Hi All, > > To update the IGBT current sharing in the OLTC... > > We have say 2000 amps and ten 200 amp IGBTs to switch it. There is a > problem in how to be sure the IGBTs share that current equally. If one > IGBT takes like 400 amps, it is a "bad" thing! The circuit looks like this: > > http://hot-streamer.com/temp/OLTC8-10-01.gif > > Basically, you slam all those IGBTs on with very low inductance, impedance > balanced "buss work". Done well, it is simple and easy, once some nasty > details are worked out. I have put a lot of thought about how I would do > this for the OLTC. Lots of worry over the 10nH wire bonds inside the IGBTs > and stuff... However, I guess that was all wasted now since the solution > is...: > > http://hot-streamer.com/temp/OLTC8-10-02.gif > > Yep! We have ten independent primary caps too. If we hook them up like > this, the cap's impedance naturally hard balances the current to the IGBTs. > It is just a done deal, "painfully" simple... nothing can go wrong. It's > gotta work. Even if just one IGBT turns on I "think" it self protects... > If one stays out, it self protects... (the frequency is wrong, but who > cares...) > > This solution to current sharing of the OLTC's IGBTs is so beautiful, it > makes me wanna cry... > > Cheers, > > Terry > > --- Richard Wayne Wall --- rwall@ix.netcom.com Delivered-To: terrellf@mail-dnvr.uswest.net Resent-Date: Sun, 11 Aug 2002 11:29:23 -0600 Resent-From: tesla@pupman.com Resent-Sender: tesla-request@pupman.com Date: Sun, 11 Aug 2002 11:26:48 -0600 From: "Tesla list" To: tesla@pupman.com X-Sender: twftesla@pop.dnvr.qwest.net X-Mailer: QUALCOMM Windows Eudora Pro Version 4.1 Subject: Re: OLTC Thoughts X-Mailing-List: archive/latest/755 X-Loop: tesla@pupman.com Original poster: "Terry Fritz" Hi Richard, At 11:16 AM 8/11/2002 -0400, you wrote: >Terry, > >Here are some other OLTC design thoughts. > >Your TC primary, 47 mF cap(s) and IGBT switch(s) form a large LC tank. >Once fired this thank will have huge oscillatory currents. What are the >reverse stand off voltages of the IGBTs? 900 volts. They make higher but these are really cheap. I ran a bunch of models studying what might happen to the voltage and where they run at and they should be very safe. I am also putting in some big MOVs to catch any over voltage swings at 750V. here is a pic of a transient and how the coils handles it: http://hot-streamer.com/temp/OLTC-08-07-2.gif >Hopefully they have a built in >protection diode. But, I would not depend on this small diode, if it >exists, for protection. Which brings up another point. Richard Hull's H2 >thyratron magnifier, by necessity, quenched at the first zero crossing of >the wave form. So only the positive half of the energy cycle was passed to >the primary. While a very interesting design, it is not a spectacular >performer, if one measures performance by spark length. It's a phenomenal >research device though. You may consider adding free running back diodes >across your IGBTs for protection and for a full one cycle energy transfer. The IGBTs have big powerful anti-parallel diodes in them already ;-)) >Have you calculated the resonate oscillatory frequency of the primary tank >without the switches? Also consider moving the switch(s) outside the LC >tank altogether to allow the tank to oscillate freely and transfer even >more energy before you electronically quench it. I'll leave these design >thoughts up to your fertile electronics imagination. They are fairly well outside it now. I have to watch out for induced currents from the primary on control wiring. Cheers, Terry > >RWW > >> Hi All, >> >> To update the IGBT current sharing in the OLTC... >> >> We have say 2000 amps and ten 200 amp IGBTs to switch it. There is a >> problem in how to be sure the IGBTs share that current equally. If one >> IGBT takes like 400 amps, it is a "bad" thing! The circuit looks like >this: >> >> http://hot-streamer.com/temp/OLTC8-10-01.gif >> >> Basically, you slam all those IGBTs on with very low inductance, impedance >> balanced "buss work". Done well, it is simple and easy, once some nasty >> details are worked out. I have put a lot of thought about how I would do >> this for the OLTC. Lots of worry over the 10nH wire bonds inside the >IGBTs >> and stuff... However, I guess that was all wasted now since the solution >> is...: >> >> http://hot-streamer.com/temp/OLTC8-10-02.gif >> >> Yep! We have ten independent primary caps too. If we hook them up like >> this, the cap's impedance naturally hard balances the current to the >IGBTs. >> It is just a done deal, "painfully" simple... nothing can go wrong. It's >> gotta work. Even if just one IGBT turns on I "think" it self protects... >> If one stays out, it self protects... (the frequency is wrong, but who >> cares...) >> >> This solution to current sharing of the OLTC's IGBTs is so beautiful, it >> makes me wanna cry... >> >> Cheers, >> >> Terry >> >> > > > >--- Richard Wayne Wall >--- rwall@ix.netcom.com > > > Delivered-To: terrellf@mail-dnvr.uswest.net Resent-Date: Sun, 11 Aug 2002 14:41:00 -0600 Resent-From: tesla@pupman.com Resent-Sender: tesla-request@pupman.com Date: Sun, 11 Aug 2002 14:36:31 -0600 From: "Tesla list" To: tesla@pupman.com X-Sender: twftesla@pop.dnvr.qwest.net X-Mailer: QUALCOMM Windows Eudora Pro Version 4.1 Subject: Re: OLTC Thoughts X-Mailing-List: archive/latest/767 X-Loop: tesla@pupman.com Original poster: "Terry Fritz" Hi Richard, At 04:14 PM 8/11/2002 -0400, you wrote: >Terry, > >> The IGBTs have big powerful anti-parallel diodes in them already ;-)) > >Excellent, do you know the speed (Reverse Recovery time) of these diodes? >If they're too slow you may need to add supplemental fast diodes. Your >fast CE MOVs or snubbers are also good ideas. The data sheet is at: http://hot-streamer.com/temp/irg4pf50wd.pdf > >> >> >Have you calculated the resonate oscillatory frequency of the primary >tank >> >without the switches? Also consider moving the switch(s) outside the LC >> >tank altogether to allow the tank to oscillate freely and transfer even >> >more energy before you electronically quench it. I'll leave these design >> >thoughts up to your fertile electronics imagination. >> >> They are fairly well outside it now. > >Your schematic places the IGBTs right in the LC tank circuit. See >http://hot-streamer.com/temp/OLTC8-10-01.gif et subsequent. Yes, the IGBTs act just like a normal spark gap. The coils basic function is just like that of any disruptive coil. > >Another consideration is that with ten IGBTs and ten 4.7 mF caps, not only >does turn on have to be right on the money, BUT turn off has to also be >exactly the same for all IGBTs. It can really be very sloppy. If one turns on 1.3uS early, no problem. The key is that the current in the tank is pretty slow to come up due to the tank inductor. The IGBT's will easily outrun it. >Post firing, as the tank oscillates and it >is also charged by the PS, differences in IGBT turn off timing will affect >residual charge imparted on each 4.7 mF cap. I can actually leave them charged negative for even higher power. http://www.pupman.com/listarchives/2002/August/msg00424.html > >What is your proposed IGBT firing rate and dwell time? 120BPS and around 130uS. The firing rate is automatically controlled and the dwell is from a 555 timer. http://www.pupman.com/listarchives/2002/August/msg00391.html http://www.pupman.com/listarchives/2002/August/msg00414.html > >RWW > >--- Richard Wayne Wall >--- rwall@ix.netcom.com > > > Delivered-To: terrellf@mail-dnvr.uswest.net Resent-Date: Sun, 11 Aug 2002 14:40:56 -0600 Resent-From: tesla@pupman.com Resent-Sender: tesla-request@pupman.com Date: Sun, 11 Aug 2002 14:36:58 -0600 From: "Tesla list" To: tesla@pupman.com X-Sender: twftesla@pop.dnvr.qwest.net X-Mailer: QUALCOMM Windows Eudora Pro Version 4.1 Subject: Re: OLTC Thoughts X-Mailing-List: archive/latest/766 X-Loop: tesla@pupman.com Original poster: "Wall Richard Wayne by way of Terry Fritz " Terry, > The IGBTs have big powerful anti-parallel diodes in them already ;-)) Excellent, do you know the speed (Reverse Recovery time) of these diodes? If they're too slow you may need to add supplemental fast diodes. Your fast CE MOVs or snubbers are also good ideas. > > >Have you calculated the resonate oscillatory frequency of the primary tank > >without the switches? Also consider moving the switch(s) outside the LC > >tank altogether to allow the tank to oscillate freely and transfer even > >more energy before you electronically quench it. I'll leave these design > >thoughts up to your fertile electronics imagination. > > They are fairly well outside it now. Your schematic places the IGBTs right in the LC tank circuit. See http://hot- streamer.com/temp/OLTC8-10-01.gif et subsequent. Another consideration is that with ten IGBTs and ten 4.7 mF caps, not only does turn on have to be right on the money, BUT turn off has to also be exactly the same for all IGBTs. Post firing, as the tank oscillates and it is also charged by the PS, differences in IGBT turn off timing will affect residual charge imparted on each 4.7 mF cap. What is your proposed IGBT firing rate and dwell time? RWW --- Richard Wayne Wall --- rwall@ix.netcom.com Delivered-To: terrellf@mail-dnvr.uswest.net Resent-Date: Sun, 4 Aug 2002 22:17:45 -0600 Resent-From: tesla@pupman.com Resent-Sender: tesla-request@pupman.com Date: Sun, 04 Aug 2002 22:04:18 -0600 From: "Tesla list" To: tesla@pupman.com X-Sender: twftesla@pop.dnvr.qwest.net X-Mailer: QUALCOMM Windows Eudora Pro Version 4.1 Subject: OLTC update X-Mailing-List: archive/latest/392 X-Loop: tesla@pupman.com Original poster: "Terry Fritz" Hi All, I have been working pretty hard on my Off-Line Tesla Coil project. Mostly ordering things and running computer models to check things out. But I am starting to get some "real" stuff going too ;-) All still looks well... I built the 240VAC controller today. http://hot-streamer.com/temp/OLTCcontFRONT.jpg http://hot-streamer.com/temp/OLTCcontBACK.jpg http://hot-streamer.com/temp/OLTCcontSCHEM.gif http://hot-streamer.com/temp/OLTCcontTOP1.jpg http://hot-streamer.com/temp/OLTCcontTOP2.jpg It came in at 4 pounds 15 ounces instead of two pounds ;-)) Hey, that is still lighter than just a big variac knob like you "big iron" guys use :o))) It is designed for about 1+kVA but it could do about 4kVA if it had to. All heavy duty stuff that can take a big "hit" if things go... "bad"... Really not much to it. Just a nice safe way to control 240 VAC into a Tesla coil. I have not had a chance to test it yet since the dryer outlet is being used for the stupid dryer ;-)) I got to use the Brother P-Touch label maker a lot :-)) I am also collecting parts for the IGBT switch: http://hot-streamer.com/temp/P8040005.jpg I guess it does not look like much yet :-) Actually, only the 10 IGBTs in the tube, the bag of MOVs, and the section of heat sink I sliced out right quick in the band saw are for the IGBT spark gap. The box and protoboard are for the IGBT firing control. Hopefully it will look cool pretty soon ;-) I was going to get the milling machine working on the heat sink tonight, but getting too late... I found that I need to fire all the IGBTs within 1.3uS of each other. Dave Sharpe clued me into the Toshiba TLP250 driver chip (just another DigiKey part...) which will be great! I still need to work a lot on the SYNC timing stuff... The IGBTs will be paralleled together with less than 15nH of inductance. More than that and their phase angles start to get goofy. It is really very simple in practice... I guy at work gave me a big beat up 300 amp brick but I was surprised it was rated at "only" 600 amps peak :-p Seems the big bricks are not designed well for really high peak currents like my project will need. No matter, since many small IGBTs are like 1/10th the cost... I wonder if an extreme Tesla coil use of IGBTs will teach the "big boys" a few trick like the MMC caps did ;-)) The bricks are really not designed all that well... they are mostly worried with slow speed high thermal dissipation stuff... I think I will go with a primary coil of three 14 inch diameter copper tubing loops (1/2 inch diameter tubing) spaced 2.5 inches apart. Everyone can find this "stuff" and the IGBTs can evenly push current into such a structure. Tuning will probably be through a variable number if primary caps. Might need a multi-taped resonant inductor too, but I'll see... A lot depends on the inductance of the cap and switch structures... Maybe I can have an adjustable inductor in there somehow... Talking like 300nH... Still thinking... It is all brand new territory here... OLTCs are sort of complex from a theoretical point of view, but the hardware is really very very simple once the gnarly details are figured out. Hopefully, it might actually work!! I am trying to keep it really simple so others could easily make one. There are fancy improvements and stuff I am neglecting in the name of "keep it simple"... We will see.... Cheers, Terry Delivered-To: terrellf@mail-dnvr.uswest.net Resent-Date: Sun, 4 Aug 2002 22:57:43 -0600 Resent-From: tesla@pupman.com Resent-Sender: tesla-request@pupman.com Date: Sun, 04 Aug 2002 22:53:42 -0600 From: "Tesla list" To: tesla@pupman.com X-Sender: twftesla@pop.dnvr.qwest.net X-Mailer: QUALCOMM Windows Eudora Pro Version 4.1 Subject: OLTC update X-Mailing-List: archive/latest/396 X-Loop: tesla@pupman.com Original poster: "Terry Fritz" Hi All, > > >I built the 240VAC controller today. > > http://hot-streamer.com/temp/OLTCcontFRONT.jpg > > I just wanted to mention that the clothes finally got dry and I was able to test it. It works perfectly :-))) http://hot-streamer.com/temp/P8040010.jpg Cheers, Terry Delivered-To: terrellf@mail-dnvr.uswest.net Resent-Date: Mon, 5 Aug 2002 19:07:28 -0600 Resent-From: tesla@pupman.com Resent-Sender: tesla-request@pupman.com Date: Mon, 05 Aug 2002 18:58:34 -0600 From: "Tesla list" To: tesla@pupman.com X-Sender: twftesla@pop.dnvr.qwest.net X-Mailer: QUALCOMM Windows Eudora Pro Version 4.1 Subject: Re: OLTC update X-Mailing-List: archive/latest/449 X-Loop: tesla@pupman.com Original poster: "David Sharpe by way of Terry Fritz " Terry Very nice looking controller. Will e-mail PDF of TLP250 datasheet to you either tonight or tomorrow... Can post on Hotstreamer. All the application info is there. Regards Dave Sharpe, TCBOR Chesterfield, VA. USA Tesla list wrote: > Original poster: "Terry Fritz" > > Hi All, > > > > > > >I built the 240VAC controller today. > > > > http://hot-streamer.com/temp/OLTCcontFRONT.jpg > > > > > > I just wanted to mention that the clothes finally got dry and I was able to > test it. It works perfectly :-))) > > http://hot-streamer.com/temp/P8040010.jpg > > Cheers, > > Terry Delivered-To: terrellf@mail-dnvr.uswest.net Resent-Date: Tue, 6 Aug 2002 22:15:52 -0600 Resent-From: tesla@pupman.com Resent-Sender: tesla-request@pupman.com Date: Tue, 06 Aug 2002 21:56:11 -0600 From: "Tesla list" To: tesla@pupman.com X-Sender: twftesla@pop.dnvr.qwest.net X-Mailer: QUALCOMM Windows Eudora Pro Version 4.1 Subject: OLTC update X-Mailing-List: archive/latest/507 X-Loop: tesla@pupman.com Original poster: "Terry Fritz" Hi All, I got the IGBT's mounted to a heat sink and the primary coil made up. http://hot-streamer.com/temp/P8050014.jpg http://hot-streamer.com/temp/P8050015.jpg http://hot-streamer.com/temp/P8060016.jpg http://hot-streamer.com/temp/P8060019.jpg The primary is three parallel 1/2 inch copper pipes 14 inches in diamater, spaced 2.5 inches apart (all center to center). Still needs copper iny and outy plates, caps, and the drive electronics before it really starts to look like something cool. But it's getting there ;-) I hope the milling machine was not cheating :-)) They don't make things better, they just make things faster ;-)) Cheers, Terry Delivered-To: terrellf@mail-dnvr.uswest.net Resent-Date: Wed, 7 Aug 2002 19:32:17 -0600 Resent-From: tesla@pupman.com Resent-Sender: tesla-request@pupman.com Date: Wed, 07 Aug 2002 19:25:29 -0600 From: "Tesla list" To: tesla@pupman.com X-Sender: twftesla@pop.dnvr.qwest.net X-Mailer: QUALCOMM Windows Eudora Pro Version 4.1 Subject: OLTC update X-Mailing-List: archive/latest/548 X-Loop: tesla@pupman.com Original poster: "Terry Fritz" Hi All, I think I figured out a super simple, easy, and cheap way to control the firing timing of my OLTC project. In this diagram: http://hot-streamer.com/temp/OLTC-08-07.gif I simply put a 0.1 ohm resistor (Rsense) inline with the charging circuit. When the current across the resistor goes negative it happens to be the perfect time to fire the IGBT array. The cap, resistors and diode (a 555 timer in real life) will control the ~130uS pulse width to give first notch quenching and all. Here are the major waveforms: http://hot-streamer.com/temp/OLTC-08-07-1.gif The timing circuit automatically seeks the most powerful mode and is very stable and reliable (at least on the computer ;-)) More importantly, such a circuit reacts very well to odd faults like this: http://hot-streamer.com/temp/OLTC-08-07-2.gif I was thinking of all kinds of fancy things and ways to run the coil's rather critical IGBT timing but they all had their problems, especially if things we not going just right. This new circuit is only a 0.1 ohm resistor, an op-amp, a 555 timer, and a driver chip to run the TLP250 IGBT drivers. What I thought would be the most messy part of the project to figure out, suddenly appears to be all done :-)) If this circuit and method work, it is a major breakthrough!! No more major technical stuff to figure out now. Just have to build it and see if it really works... Cheers, Terry Delivered-To: terrellf@mail-dnvr.uswest.net Resent-Date: Thu, 8 Aug 2002 12:31:26 -0600 Resent-From: tesla@pupman.com Resent-Sender: tesla-request@pupman.com Date: Thu, 08 Aug 2002 12:29:06 -0600 From: "Tesla list" To: tesla@pupman.com X-Sender: twftesla@pop.dnvr.qwest.net X-Mailer: QUALCOMM Windows Eudora Pro Version 4.1 Subject: OLTC update X-Mailing-List: archive/latest/571 X-Loop: tesla@pupman.com Original poster: "Terry Fritz" Hi All, I drew up schematics for the timing control and IGBT gate drives. http://hot-streamer.com/temp/OLTC-TimingControl.gif http://hot-streamer.com/temp/OLTC-IGBTdrive.gif Not much to it ;-)) I'll just get a $3 current transformer from DigiKey to detect the current, run the output to a comparator to trigger a 555 timer and it's done. ;-) A current transformer provides isolation which is pretty handy in this case. Close to the IGBTs, I will have a the gate drive to push current into and out of all their gates. The TLP250s are opto isolators too so hopefully there will not be too much odd current running around. Might get tricky having electronics near a 2500 amp coil and 200,000 volt RF streamers running about ;-)) I'll have to try and keep things perpendicular to the big current loop. Cheers, Terry ================== Hi All, I think I figured out a super simple, easy, and cheap way to control the firing timing of my OLTC project. In this diagram: http://hot-streamer.com/temp/OLTC-08-07.gif I simply put a 0.1 ohm resistor (Rsense) inline with the charging circuit. When the current across the resistor goes negative it happens to be the perfect time to fire the IGBT array. The cap, resistors and diode (a 555 timer in real life) will control the ~130uS pulse width to give first notch quenching and all. Here are the major waveforms: http://hot-streamer.com/temp/OLTC-08-07-1.gif The timing circuit automatically seeks the most powerful mode and is very stable and reliable (at least on the computer ;-)) More importantly, such a circuit reacts very well to odd faults like this: http://hot-streamer.com/temp/OLTC-08-07-2.gif I was thinking of all kinds of fancy things and ways to run the coil's rather critical IGBT timing but they all had their problems, especially if things we not going just right. This new circuit is only a 0.1 ohm resistor, an op-amp, a 555 timer, and a driver chip to run the TLP250 IGBT drivers. What I thought would be the most messy part of the project to figure out, suddenly appears to be all done :-)) If this circuit and method work, it is a major breakthrough!! No more major technical stuff to figure out now. Just have to build it and see if it really works... Cheers, Terry Delivered-To: terrellf@mail-dnvr.uswest.net Resent-Date: Thu, 8 Aug 2002 20:26:10 -0600 Resent-From: tesla@pupman.com Resent-Sender: tesla-request@pupman.com Date: Thu, 08 Aug 2002 20:06:07 -0600 From: "Tesla list" To: tesla@pupman.com X-Sender: twftesla@pop.dnvr.qwest.net X-Mailer: QUALCOMM Windows Eudora Pro Version 4.1 Subject: OLTC update X-Mailing-List: archive/latest/579 X-Loop: tesla@pupman.com Original poster: "Terry Fritz" Hi All, Today I explored ways to raise the OLTCs resonant frequency and increase the power. I found that I could go to a smaller 28.2uF primary cap (6 x 4.7uF) which would raise the resonant frequency to around 35kHz. So, maybe I will only need "one" mile of secondary wire :-) Here is the schematic: http://hot-streamer.com/temp/OLTC8-8-01.gif Here is something interesting about power! Since the quenching is electronically controlled and we are using resonant charging, we can turn the IGBTs off any time we like. Such as, when the caps are at a far negative voltage: http://hot-streamer.com/temp/OLTC8-8-05.gif What is stunning here, is that this gives us an effective firing voltage of 1018 volts while still staying reasonably in the safe region for 900 volt IGBTs! 1080 volts, 28,2uF, 120BPS... That is a coil primary system power of 1973 watts (16.4 joules per bang)!! Assuming a 90% coil efficiency, that "suggests" a streamer length of over seven feet!! However, there is a problem with the primary current: http://hot-streamer.com/temp/OLTC8-8-04.gif Looks like about 4500 amps peak =:O One would have to parallel up about 25 IGBTs. Not impossible or that expensive... I will not go this high in power for this coil (yet :o)) since I want some pretty good safety margins during 1st explosion... I mean "light" :-)) But really, the only change would be another pound of IGBTs and heat sinks for another $75. It is bazaar what you can do when things are all electronically controlled. I have a little more confidence in the models at 25kHz. They predict a 400kV peak output voltage with about a 500uS ringdown: http://hot-streamer.com/temp/OLTC8-8-03.gif No problems there ;-)) Although, I imagine getting hit with a coil at that frequency and power would not be fun... The system charging current is very tame with no surprises: http://hot-streamer.com/temp/OLTC8-8-02.gif This is interesting. Here is the AC line voltage and current into the system: http://hot-streamer.com/temp/OLTC8-8-06.gif No need for line filters or PFC caps :-)) I am think this coil will be very electrically quiet compared to the spark gap types. There really are not any terrible RFI generators in it except the streamers. Remember that the IGBT "main gap" gives off no real sound, heat, or light (at least, not more than once :o)) There is a slight "clicking" sound as 2000+ amps pulse through it and maybe 20 to 30 watts of heat, but that's it... System complexity is still very low. I am a bit worried that one may need a scope to check things in such a coil during construction, but I'll try to make it "just work" without any fancy stuff. Protection against faults is fairly good but only "real" testing can verify that... The theory may be "new" but it is much simpler than a conventional coil. There just isn't much hardware to it. System weight is also very low. There are two transformers. The current sense CT, but it is less than an inch square ;-) There will have to be an AC resonant inductor that will probably be about 4 pounds. I could, and may, use a small variac for testing, but the inductors value will never change normally, so it could be a simple fixed or tapped inductor. I have a variac core (4 pounds) and some smaller ones that would probably work. I was trying to sneak by without using any iron stuff ;-)) However, the "iron" weighs only 3% of what a 15/120 NST, SRSG motor and big variac does (160 pounds!) Of course, the super seven foot arc version would weigh "nothing" compared to a pig, 25 amp variac, controller, and an arc welder :o))) I will guess the whole thing will be about 25 pounds with most of the weight in the secondary and top terminal. The secondary will probably be like a 12 x 48 inch Sonotube. The top terminal will be a big as possible to help push that frequency down. Cost also seems low. The secondary and top terminal are big and conventional. Probably blow $50 there. The controller is almost nothing to $100 if you make a fancy one like I did. The electronics is maybe $40 with nice die cast boxes. $30 just in power cords and the dryer plug. "stuff" is probably another $50. Caps and IGBTs are anybody's guess. I'll just say they are about $5 each if you look around. Probably $100 total. So that adds up to... $370 Probably could do it for $250. Of course, "extra" IGBTs can add up, but at least things are pretty well isolated so "failures" are contained in just that part. Insuring reliability of the IGBTs needs to be the big priority once it is working. Not really much else that could fail... Three areas of concern are: 1. Primary circuit resistance and loss. 28.2uF at 25kHz as a reactance of 0.226 ohm. Not much room for a lot of extra resistance there!! I am trying for a primary circuit resistance of about 0.01 ohm. That is really low, but the primary circuit does resemble a big copper short ;-)) This actually goes right along with a primary inductance of only ~750nH! 2. Control, faults, and the unexpected... The "computers" say it will work ;- )) But there are a lot of things that can go wrong between theory and reality. I think the big basics are solid, but there certainly may be some "technical" problems that will need work. 3. The IGBTs. Currents of thousands of amps and many hundreds of volts... Current sharing... If there is anyway for it to fail, it certainly will. I "think" I have things covered, but I can use pretty light fuses to save the parts for analysis if they don't. I think if one IGBT "goes away", the rest will be safe. No too bad really but at these powers and this level of "experimentation", blown IGBTs are expected till the bugs are worked out. Of course, this all only exists in hard drives and dreams at the moment, but parts are coming in ;-)) Many thanks for everyone's ideas and thoughts on this, I have used them all!! Cheers, Terry Delivered-To: terrellf@mail-dnvr.uswest.net Resent-Date: Thu, 8 Aug 2002 23:32:16 -0600 Resent-From: tesla@pupman.com Resent-Sender: tesla-request@pupman.com Date: Thu, 08 Aug 2002 23:25:39 -0600 From: "Tesla list" To: tesla@pupman.com X-Sender: twftesla@pop.dnvr.qwest.net X-Mailer: QUALCOMM Windows Eudora Pro Version 4.1 Subject: Re: OLTC update X-Mailing-List: archive/latest/593 X-Loop: tesla@pupman.com Original poster: "Eddie Burwell by way of Terry Fritz " Terry, Sound's like your pretty pump on this. I for one won't miss the weight of a NST! You could add some electrolytics and make a doubler (or rectify 220V) and then use an inductor to resonantly charge the primary cap at a frequency other than 60 hz. This would also double the voltage on the cap again. Then add some switch timing mischief.... Eddie Burwell Delivered-To: terrellf@mail-dnvr.uswest.net Resent-Date: Fri, 9 Aug 2002 23:41:31 -0600 Resent-From: tesla@pupman.com Resent-Sender: tesla-request@pupman.com Date: Fri, 09 Aug 2002 23:20:05 -0600 From: "Tesla list" To: tesla@pupman.com X-Sender: twftesla@pop.dnvr.qwest.net X-Mailer: QUALCOMM Windows Eudora Pro Version 4.1 Subject: Re: OLTC update X-Mailing-List: archive/latest/649 X-Loop: tesla@pupman.com Original poster: "B2 by way of Terry Fritz " Hi Terry, Would it be easier to build a secondary that couples to a higher harmonic of the primary? Cheers, Barry Delivered-To: terrellf@mail-dnvr.uswest.net Resent-Date: Sat, 10 Aug 2002 19:45:54 -0600 Resent-From: tesla@pupman.com Resent-Sender: tesla-request@pupman.com Date: Sat, 10 Aug 2002 19:38:22 -0600 From: "Tesla list" To: tesla@pupman.com X-Sender: twftesla@pop.dnvr.qwest.net X-Mailer: QUALCOMM Windows Eudora Pro Version 4.1 Subject: OLTC update X-Mailing-List: archive/latest/730 X-Loop: tesla@pupman.com Original poster: "Terry Fritz" Hi All, I have verified two import things the computer models predicted (apparently very well ;-)) I tested my primary coil today and it measures 439.5nH. So you "can" make a big single turn primary with less then 500nH inductance ;-)) That is critical for getting a high top voltage since Vo = ~1/2 SQRT (Ls/Lp). A lower Lp also allows higher operating frequency and a lower inductance secondary (which is already pretty high at about 500+mH). More importantly, I measured the coupling to my present big coil secondary at k=0.125!! I am not sure anyone realized that a single turn primary could have a high coupling like that! That is, of course, key for this thing to work. http://hot-streamer.com/temp/P8100020.jpg I used the simple (with nice test equipment) method on page 14-11 of John's "Tesla coil Construction Guide". A few weeks ago, I just happened to punch the single turn primary case into Mark R.'s program and was stunned the coupling was that high. That was a key factor that started this whole thing... Apparently, the coupling "really" is that good! These may seem like odd technical points, but they pretty much insure the OLTC will work. Just a matter of details now. In another note. I think I figured out how to insure that the IGBTs share current perfectly evenly!! No fancy stuff needed at all. I just thought of it like 15 seconds ago so let me think on it *:-)) It would be "tooooo" simple... Eliminates some rather critical IGBT mounting stuff... I also need some way to verify the very low loss of the primary circuit but I think I need the IGBTs and all running first since the loss is like 0.03 ohm and not at all easy to test. Probably need to test that for real but at low power... Back to work now ;-) Cheers, Terry Delivered-To: terrellf@mail-dnvr.uswest.net Resent-Date: Sat, 10 Aug 2002 21:48:28 -0600 Resent-From: tesla@pupman.com Resent-Sender: tesla-request@pupman.com Date: Sat, 10 Aug 2002 21:43:50 -0600 From: "Tesla list" To: tesla@pupman.com Cc: "David Sharpe" X-Sender: twftesla@pop.dnvr.qwest.net X-Mailer: QUALCOMM Windows Eudora Pro Version 4.1 Subject: OLTC update X-Mailing-List: archive/latest/736 X-Loop: tesla@pupman.com Original poster: "Terry Fritz" Hi All, To update the IGBT current sharing in the OLTC... We have say 2000 amps and ten 200 amp IGBTs to switch it. There is a problem in how to be sure the IGBTs share that current equally. If one IGBT takes like 400 amps, it is a "bad" thing! The circuit looks like this: http://hot-streamer.com/temp/OLTC8-10-01.gif Basically, you slam all those IGBTs on with very low inductance, impedance balanced "buss work". Done well, it is simple and easy, once some nasty details are worked out. I have put a lot of thought about how I would do this for the OLTC. Lots of worry over the 10nH wire bonds inside the IGBTs and stuff... However, I guess that was all wasted now since the solution is...: http://hot-streamer.com/temp/OLTC8-10-02.gif Yep! We have ten independent primary caps too. If we hook them up like this, the cap's impedance naturally hard balances the current to the IGBTs. It is just a done deal, "painfully" simple... nothing can go wrong. It's gotta work. Even if just one IGBT turns on I "think" it self protects... If one stays out, it self protects... (the frequency is wrong, but who cares...) This solution to current sharing of the OLTC's IGBTs is so beautiful, it makes me wanna cry... Cheers, Terry Delivered-To: terrellf@mail-dnvr.uswest.net Resent-Date: Sat, 10 Aug 2002 21:48:28 -0600 Resent-From: tesla@pupman.com Resent-Sender: tesla-request@pupman.com Date: Sat, 10 Aug 2002 21:44:09 -0600 From: "Tesla list" To: tesla@pupman.com X-Sender: twftesla@pop.dnvr.qwest.net X-Mailer: QUALCOMM Windows Eudora Pro Version 4.1 Subject: RE: OLTC update X-Mailing-List: archive/latest/737 X-Loop: tesla@pupman.com Original poster: "Jeff W. Parisse by way of Terry Fritz " Terry, I agree that single turn primary coupling is counterintuitive. Bill Wysock's 13M magnifier uses a single turn primary and has the characteristic tight pri/sec coupling found on most magnifier designs. Interesting. I Love the threads that result from your "pet" projects... (remember the strobe?). ;-) Since we (TCML) coin the phrases that are used ad nauseam, I wanted to offer my 2 cents (worth less here) on the "off line" term. Off line implies that it's "off line", not working, not ready... Perhaps, transformer-less, ac resonant or something like that. Yes, this is my pitiful offering... a way to name it... Sheesh... When you're farther along and I have more time, I'll jump in, build it and offer my help... Jeff Parisse > I am not sure anyone realized that a > single turn primary could have a high > coupling like that! That is, of course, > key for this thing to work. Delivered-To: terrellf@mail-dnvr.uswest.net Resent-Date: Sun, 11 Aug 2002 10:47:14 -0600 Resent-From: tesla@pupman.com Resent-Sender: tesla-request@pupman.com Date: Sun, 11 Aug 2002 10:33:52 -0600 From: "Tesla list" To: tesla@pupman.com X-Sender: twftesla@pop.dnvr.qwest.net X-Mailer: QUALCOMM Windows Eudora Pro Version 4.1 Subject: Re: OLTC update X-Mailing-List: archive/latest/744 X-Loop: tesla@pupman.com Original poster: "tesla by way of Terry Fritz " Great scheme to ensure current sharing Terry Thought FWIW Is it better to use more lower value caps to keep the ESR lower or are the 10 by 4.7uF caps a good compromise for ESR. Best Ted L in NZ Delivered-To: terrellf@mail-dnvr.uswest.net Resent-Date: Sun, 11 Aug 2002 10:47:07 -0600 Resent-From: tesla@pupman.com Resent-Sender: tesla-request@pupman.com Date: Sun, 11 Aug 2002 10:35:54 -0600 From: "Tesla list" To: tesla@pupman.com X-Sender: twftesla@pop.dnvr.qwest.net X-Mailer: QUALCOMM Windows Eudora Pro Version 4.1 Subject: Re: OLTC update X-Mailing-List: archive/latest/748 X-Loop: tesla@pupman.com Original poster: "Phil@TCBFW by way of Terry Fritz " Terry, Hat's off to you for the idea on current sharing . . . I've used this idea before in a non-TC application (my load to each IGBT was inductive not capacitive) and it worked great until trying to get the IGBT's to switch in the nano second range. You should be able to compensate for this by sending the trigger pulse to each IGBT through a PFN, do a bit of low power adjusting and I think you'll have "TC The Next Generation" Waiting with anticipation to see your results !!! Phil TCBFW ----- Original Message ----- From: "Tesla list" To: Cc: "David Sharpe" Sent: Saturday, August 10, 2002 10:43 PM Subject: OLTC update > Original poster: "Terry Fritz" > > Hi All, > > To update the IGBT current sharing in the OLTC... > > http://hot-streamer.com/temp/OLTC8-10-02.gif > > Yep! We have ten independent primary caps too. If we hook them up like > this, the cap's impedance naturally hard balances the current to the IGBTs. > It is just a done deal, "painfully" simple... nothing can go wrong. It's > gotta work. Even if just one IGBT turns on I "think" it self protects... > If one stays out, it self protects... (the frequency is wrong, but who > cares...) > > This solution to current sharing of the OLTC's IGBTs is so beautiful, it > makes me wanna cry... > > Cheers, > > Terry > > > Delivered-To: terrellf@mail-dnvr.uswest.net Resent-Date: Sun, 11 Aug 2002 10:54:56 -0600 Resent-From: tesla@pupman.com Resent-Sender: tesla-request@pupman.com Date: Sun, 11 Aug 2002 10:51:40 -0600 From: "Tesla list" To: tesla@pupman.com X-Sender: twftesla@pop.dnvr.qwest.net X-Mailer: QUALCOMM Windows Eudora Pro Version 4.1 Subject: Re: OLTC update X-Mailing-List: archive/latest/753 X-Loop: tesla@pupman.com Original poster: "Terry Fritz" Hi Ted, The 4.7uF 600V caps are the largest in the right range that they make. They just hapened to turn out matching up to the IGBTs well. It just sort of worked out just right ;-) Cheers, Terry At 10:11 PM 8/11/2002 +1200, you wrote: >Great scheme to ensure current sharing Terry >Thought FWIW >Is it better to use more lower value caps to keep the ESR lower or are the >10 by 4.7uF caps a good compromise for ESR. >Best >Ted L in NZ > Delivered-To: terrellf@mail-dnvr.uswest.net Resent-Date: Sun, 11 Aug 2002 13:12:50 -0600 Resent-From: tesla@pupman.com Resent-Sender: tesla-request@pupman.com Date: Sun, 11 Aug 2002 13:05:44 -0600 From: "Tesla list" To: tesla@pupman.com X-Sender: twftesla@pop.dnvr.qwest.net X-Mailer: QUALCOMM Windows Eudora Pro Version 4.1 Subject: Re: OLTC update X-Mailing-List: archive/latest/760 X-Loop: tesla@pupman.com Original poster: "Paul Nicholson by way of Terry Fritz " I guess the main difficulty with the OLTC is that a large primary cap is required in order to get a decent firing energy. Thus an impractically small primary inductance is called for. Barry wrote: > Would it be easier to build a secondary that couples to a > higher harmonic of the primary? Hmm. Putting the energy into a primary (or any coil) by connecting a charged cap across its ends tends to excite the coil at its lowest resonant mode, which is what we'd call its Fres. Only a little of the firing energy goes into higher modes of the primary. To excite the primary at a higher mode (not a harmonic), one way would be to simultaneously discharge several caps, strategically placed along the winding, alternated with ground connections. The initial firing energy would be shared across all the caps. This would seem to be feasible given the use of semiconductor switches. Apart from the difficulty with tuning, so far so good. The problem then is to couple effectively to the secondary. Now with the primary at a higher mode, current in parts of the primary will tend to be pulling against current in other parts of the primary as far as the inductive coupling to the secondary quarter wave is concerned. And as the mode number gets higher, the overall cancellation of the field from the primary becomes more complete. Therefore, it would be necessary, in order to obtain a cooperative field from each half-wave current segment of the primary, to reverse the winding sense between each firing cap. But doing so would destroy the transmission line properties which were giving you the particular mode, and the system degenerates almost into a set of parallel primary windings, each with its own personal cap. So to persue Barry's rather interesting idea, I think it would be necessary to make sure that the secondary was closely coupled to only one half-wave's worth of the primary, and that the rest of the primary contributes relatively less to the coupling. Acmi can be used to model the coupling to a secondary from a primary which has caps placed along it's length, so long as these caps are large compared to the self-C of the primary itself, which it's safe to assume. Note that this might not be possible with traditional spark gaps because of the need for simultaneous switching. Apart from tuning, I think the other main problem with this approach would be how to ensure that large amounts of energy don't converge onto a single cap of the primary, which it's perfectly free to do so once the thing is released into resonance - given unpredictable reflections from the top of the secondary. We wouldn't want any of the caps to suddenly find itself carrying a whole lot more energy than it started out with. But then we do have scope for controlling the timing of the switching, so maybe things could be controlled automatically somehow. If you get into that sort of territory, you can conjecture building up a traveling pulse waveform along a primary, by firing caps consecutively, so that the energy is built up into a single broadband pulse. You'd set things for a steady impedance transformation along the way, and the primary would couple to or merge into a secondary, which would continue the impedance transformation (by some cunning choice of coil profile) to give a single huge voltage pulse at the top. Picture a deep atlantic wave approaching a beach, rolling up in the shallows, trading width for height. The deep sea wave sees a gradual impedance change as the beach applies some extra boundary conditions. Surfers know the beaches with the best profiles - wonder what shape of coil would turn a long duration, low-voltage, high- current, broadband pulse applied at one end, into a short, high voltage, low current pulse at the other. Note that this is quite unlike a normal TC. We're now firing up multiple resonant modes, and timing them so that they momentarily converge to a single giant voltage pulse at one end (preferably the far end) of the coil. But before anyone gets too excited, there are some problems with this approach and it's not likely to give you any more topvolts per joule than a regular TC, if my sums are correct. But, I think I like Barry's idea, because it opens up some interesting possibilities. Whether it would bo any better than, say, a normal primary with lots of separate caps and switches paralleled across it would have to be considered. All in all, I'm following this OLTC stuff with great interest. -- Paul Nicholson -- Delivered-To: terrellf@mail-dnvr.uswest.net Resent-Date: Sun, 11 Aug 2002 10:51:46 -0600 Resent-From: tesla@pupman.com Resent-Sender: tesla-request@pupman.com Date: Sun, 11 Aug 2002 10:36:42 -0600 From: "Tesla list" To: tesla@pupman.com X-Sender: twftesla@pop.dnvr.qwest.net X-Mailer: QUALCOMM Windows Eudora Pro Version 4.1 Subject: [Fwd: OLTC update] X-Mailing-List: archive/latest/752 X-Loop: tesla@pupman.com Original poster: "David Sharpe by way of Terry Fritz " Terry wanted me to forward this to the list while he is sleeping... The OLTC thing is getting bigger...MUCH BIGGER, then anyone would have predicted. Like dynamic tuning by controlling how many IGBT's are switching (controling the tank C). Would now require the CEEC structure I mentioned in a previous post to allow bidirectional isolation. The IGBT / capacitor structures can now be modularized for plug in into busbars from primaries, which is exactly the same design feature used on high power Klystron drivers for linear accelerators. Another point is potentially (no pun) scale up to truly enormous power capabilities. Now if we could just get the voltages up, so we could reduce tank losses (i^2r and proximity)... Best Regards Dave Sharpe, TCBOR Chesterfield, VA. USA Return-Path: Received: from mx05.mrf.mail.rcn.net ([207.172.4.54] [207.172.4.54]) by mta04.mrf.mail.rcn.net with ESMTP id <20020811044326.ULHU385.mta04.mrf.mail.rcn.net@mx05.mrf.mail.rcn.net>; Sun, 11 Aug 2002 00:43:26 -0400 Received: from dnvrpop8.dnvr.uswest.net ([206.196.128.10]) by mx05.mrf.mail.rcn.net with smtp (Exim 3.35 #6) id 17dkZe-0002qO-00 for sccr4us@erols.com; Sun, 11 Aug 2002 00:43:30 -0400 Received: (qmail 70696 invoked by uid 0); 11 Aug 2002 04:43:26 -0000 Received: from unknown (HELO 2x400mhz) (63.229.238.61) by dnvrpop8.dnvr.uswest.net with SMTP; 11 Aug 2002 04:43:26 -0000 Date: Sat, 10 Aug 2002 22:43:49 -0600 Message-Id: <4.1.20020810222127.00a23fc8@pop.dnvr.qwest.net> From: "Terry Fritz" To: "David Sharpe" Approved: twftesla@qwest.net X-Sender: twftesla@pop.dnvr.qwest.net X-Mailer: QUALCOMM Windows Eudora Pro Version 4.1 Subject: Re: OLTC update In-Reply-To: <3D55E0C2.32751EEC@erols.com> References: <4.1.20020810212426.00a1da80@pop.dnvr.qwest.net> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" X-Mozilla-Status2: 00000000 Hi Dave, At 11:57 PM 8/10/2002 -0400, you wrote: >Terry > >AWESOME!!! You just made a self ballasted current source >where the currents are dependent on the statistical manufacturing >tolerances on capacitor values. What modeling I've done suggests >that 5% tolerance are acutally within 1% when taken as a grand >average. Need to scale up or drop frequency lower (or dynamically >tune the system!!!! :^) ) Adjust the number of active IGBT's on >the fly. You're right, I'm crying with you. :^) Oh God!!! Your right, you can tune the darn thing too by just "selecting" cap strings!!!!!!!!!! Also, there is no limit to the number of IGBTs in this method. So a 300uF, or whatever cap, with a big number of IGBTs is practical!!! What!!! Like 50,000 amp primaries :-))))) Suddenly, there is no limit to the power... :o))))))))))))) We could scale it up till the breaker box blows off the wall :- ))))))) > >BTW this concept is similarly used with inductors to tie together >very large bridge converters for large (100kW to 1MW+) SS Induction >heating appartus. Lumped inductors in load then become part of >total tank circuit, and FORCE current sharing similarly to caps in >this application. I can find URL with this application in it to illustrate >my point... Yep! It's not a new concept. I have used it to do current measurements on MMCs when I didn't have the ability to measure the full current. I just did one string... > >Excellent, now sleep on it, but I wanna see it FIRE!!! Yep!! I sleep a lot, but still not enough to handle all this :-))) I "thought" this may be "cool", but I never knew it was a "big thing" ;-)) Think we can obsolete variacs and transformers (NSTs, MOTs, and even PIGs) like we did them oil filled homemade poly caps :-))) Cheers, Terry > >Regards >Dave Sharpe, TCBOR >Chesterfield, VA. USA