SSTC CURRENT DRAW TESTING
February 20, 2003
I decided to try and work out actually how much current the SSTC board was sucking from the mains. It is difficult to measure and even though my tests readings are correct, they are probably nothing like the actual true RMS power.
I tried to monitor the current draw on a DMM though this proved impossible since the meter was displaying almost a random current. I had a old AMM so I connected that up to see what would happen.
At first the current showed what looked like 0amps. The AMM showed this and didn't even flicker. I thought at first I wasn't making contact somewhere though my DMM showed 20VDC input. I reset my variable frequency pot and swept though from around 80khz to 300khz. I found as I moved up the needle started to move a little. I used my 12" flo-tube light as a indicator as normal. This lite up and I was drawing at this point around 500ma. I moved on and current dropped down again to 0.
I then came across another point where the tube lite up and current draw moved upto about 1amp. Tube was a lot brighter this time. I moved frequency on again and current again went to 0amps. I then came across the strongest frequency which I assume is the resonant frequency and the tube was brighter still. Current draw went up to 2amps. Out of curiosity I moved on with the frequency and current again went to 0. At this point another coil in the room started to buzz, couldn't work out which one though must have been on a harmonic of the secondary coil on test.
I moved on more and I got current of about 500ma again and it started to fade out in current just as it did when I first started the tests. The coil used had a frequency of about 110khz and I went from around 80khz to about 300khz. It was clear I could excite the secondary with other sub frequencies though they proved to be of little use.
I then moved up the voltage to around 50VDC and where the maximum current draw was 2amps before, it went up to 7amps. I can only assume this is peek current since a few days ago I was running on as little as a 1amp fuse which didn't blow for a while. I also tried the full frequency range again and they came out much the same though current readings were of course doubled as from the first tests.
I think with a AMM you can see peek currents which are probably not far off being accurate. I would assume I can run my SSTC right upto 20amps peek since this is their rating, Depending on how much voltage I would need to draw 20amps peek of course. Going by these tests I would say running the driver stage on 100VDC would pretty much equal 20amps peek current which is of course the Mosfet rating. I haven't tried higher than 40VDC yet so hopefully I can check on this in some latter tests.
Any voltages over 100VDC would start to overload the Mosfets and they may take 40amps peek for 200VDC though I wouldn't like to try it. They maybe safe since RMS power could well be only a fraction of that. I probably won't go over 150VDC until I can source some higher power Mosfets. The readings could be wrong though I think I'm prettymuch on track with the peek currents here.
One way to test it with poormans equipment is to take voltage readings from the smoothing capacitor bank. I have had some good results with this method and I think makes a good test. The smoothing caps work out to a value of 0.5F. These are of course pretty beefy caps. Off hand I think I worked out this value to give a 2amp supply for 1 second with a voltage drop of 10volts. I don't have the formula to hand to check on it accurately but go with it for now...
If I were then to draw 2amps constant DC for 1 second then I would of course expect to see the voltage drop on the capacitors from 100VDC to 90VDC. Which of course should be true. So all which needs to be done if to time the 1 second, with a small 555timer or a stopwatch, Though timing 1 second by eye isn't really a good test though it can still work if you repeat the tests and take a approx. reading.
Now if the SSTC was drawing a constant DC load then the voltage would drop to 90volts just like a resistor would drop the voltage in 1 second. Since the SSTC is running at high frequency and drawing a peek current it won't draw a constant 2amps. The trick is now to run the coil for 1 second and check the voltage in the smoothing caps. I would repeat this test about 10times to make sure nothing was in error.
If I found 90volts in the capacitors then I know the RMS power would work out to be 2amps/second. If I found 95volts then I would estimate that the current draw was 1.5amps/second. Of course if I found the voltage was down to 80volts then I would then estimate that there would be more than a 2amp current draw, maybe 4amps. Repeating the tests you should gain about the the voltage in each test. If I found something like 98volts left in the cap bank then I would probably half the value to 0.25F or less since that large capacitance isn't really needed.
This idea can also be used in reverse. One time I wanted to find out actual current capability of a PSU I built. It wasn't anything special and only gave a few mA output. I calculated that it would take around 22 seconds to charge a 68,000uF cap to 60VDC. I can't remember the current I used for the calculation, maybe 10ma CCS would charge it in 22 seconds.... Anyway, After a few tests I found that the capacitor charge in about the same time. The actual time was 20 seconds so I could prettymuch work out that there was say 11ma CCS there. Of course this was a few years ago when I did those tests but using capacitors seems to provide a good indication of current draw.
It maybe a crude way to check on current draw though I have have good results so far and hopefully I can shed some better light into actual current draw from my SSTC design at various voltage inputs. So far I would guess from what I have seen so far that the actual current draw for 30VDC input is only around 1amp. A 1amp fuse seems to last rather well so it shouldn't be too far off.
March 1, 2003
After a lot of testing I found that my 0.4mm wound coil didn't draw as much current from the supply. I tried a lot of coil in various dimentions and wire diameters to try and find out why. I tried a 0.2mm wound coil which always worked better than my 0.4mm wound coil which had the same length of wire. After talking to Jan and Richie I found out that the inductance was to draw more current from the supply, which is probably why it worked better. Since my 0.2mm coil had a larger inductance.
Latter testings actually showed that my 0.4mm coil actually worked a fraction better than my 0.2mm coil. The 0.2mm coil had 180uH inductance and the 0.4mm coil had 130uH inductance. I was finding that a higher inductance was going to be better though now I wasn't so sure.
I measured the supply current at 20VDC input for both coils. The 0.2mm coil pulled 4amps from the supply. The 0.4mm coil pulled 3amps from the supply. I repeated the tests not alterting the voltage at anytime and testing spark length. I found now that the 0.4mm coil gave a fraction more power output. Strange results since on inductances basis it shoulsn't work aswell since there is less input to the coil.
After another chat to Jan he suggested I try alterting my primary turns. It would seem fewer turns would make the driver see more inductance and pull more current. I wasn;t so sure about this so I put it to the test. I wound a quick primary just 2 turns of coax cable and tried that. To my amazement I could pull 10amps from the supply now even at 20VDC input. I now had 200watts power based on the readings. Mosfets got very hot quick so I didn't try this for very long. It didn't seem however to give anymore output despite the higher current input.
I also noticed that current just about always the same no matter if I put in 20VDC input or 100VDC input. It would draw say 3amps ( normal yellow primary and 0.4mm coil ) at 20VDC input and also 3amps at 100VDC input. Of course the wattage in would go up ( 20VDC x 3amps = 60watts, 100VDC x 3amps = 300watts ). Of course 300watts overrates the wattage of the mosfets so care must always be taken!
March 2, 2003
More current monitoring tests today. After much pondering and testing frequencies I came to a odd conclusion.. I used my Yellow primary coil, 30VDC input. Now without knowing the running frequency of the coil I went from about 60Khz up to around 500khz to see what I could find. At 136khz I could draw some nice arcs about 2" long, Current input was 4Amps. I then found that 98khz lite up tubes well but no sparks. At 98khz I was pulling 2amps. Unknown until today my coils Fres was actually 107khz. My nice sparks at 136khz seemed impossible.
I latter ran the coil after setting up the frequency to 107khz. I was supprised that I got no current draw and no sparks or anything at all infact! This seemed very odd since I could draw a few inches when I had it running at 136khz! So then I put back a quick 2 turn primary to pull more current from the supply to see what would happen. I ran at 107khz and I was pulling 3.7amps at 30VDC input. I got some arcs ( not as good as my 136khz it would seem ) which seemed to be right. I then upped the voltage to 50VDC input and it pulled 6amps. This was about the rating of the mosftet's 250watts rating.
I found that the Mosfets were getting very hot and also every diode on the board was also getting very hot even after just a few seconds run time. I then realised that this is why I had some problems with Jan's auto tune design since most times I couldn't get it to start up. Until I used my 2 turn test primary there just wasn't enough EMF strength to start it oscilating. I had to move my small pickup antenna back about 10" from the coil since arcs were comming off the end off it!
Auto tune seemed to be working so I was back in action... or so I thought until I couldn't draw any arcs from the secondary at all. I then realised what Jan had said about the primary if connected the wrong way round it could tune into a harmonic. This I found to be true. I tested this idea with a small 12" flo-tube which I move up along the side of the coil. About every 2inches the tube would go dim, then go bright again, then go dim again. This happend about 4 times along the coil from bottom to top.
I swapped the wires around and I was back in action. I got some sparks an everything seemed to be running fine again. I also noticed that the mosfets and diodes were all getting very hot very fast again with my 2 turn primary. It seems clear now that I must try a primary such as my yellow one and try spark output, then take off 1 turn and try spark output again. Also keeping track of the input current.
I will add more as I do more tests. For more interesting tests relating to this page click HERE