Setup/Tuning Multicoil Machines

[James Milner]

TRIMMER RESULTS
I built a 5 filar board and added 10 ohm 20 turn trimmers in series with the base resistors.
I had set all the trimmers the same at 5 Ohm so I could adjust 5 Ohm in each direction. My transistors were already matched as were the resistors. As it turned out I had already done a good job and all 5 gates on the transistors were opening in unison. So I gave one of them a few turns and the signal drifted every now and then and as I continued to increase the turns the two signals became less in phase.
I turned the pot back till the drift started to go the other way and and then found the spot where it was stable and perfectly in phase.
I'll upload some video to this post tomorrow to more adequately describe what I mean.

The interesting thing is the questions that are provoked by getting these transistors perfectly aligned. When aligned the wave addition is fantastic! Nice big pulses
I found when I tested the signal on the other side of the diode the signal was no where as clear or as strong or even appeared to be as in phase as it was on the collector side. Obviously because of the voltage drop across the diodes! So why are we using a diode on each output? And reducing the amplitude of all the pulses outbound. Why not just use one larger diode or better still a full wave bridge rectifier and use both the high and the low to increase the punch and waste less, particularly if going down the cap pulser path.
Tom, John K. Any idea why JB uses separate diodes on each collector output?
If not tuned I could understand but tuned to a high degree it seems a waste! Particularly if radiant is more present with larger voltage gradients. By simple wave addition even when attached to the battery I am getting approx 30 v per transistor, that's 150v top to bottom on the h-wave using simple wave addition, obviously if I was to test without a battery it would be much higher but I don't fancy blowing my precious trannies.

I think. My next test will be to remove the diodes in series and replace them with a single 3A or 5 A diode. Shotky even better! Then a FWBR and I'll keep you posted!
Cheers
James

[aln]

FWBR is the road I am going to take. Why not get everything and send it to battery? Aln

[James Milner]

SCHOTTKY DIODES
I had a play with Schottky diodes in preference with regular 1n4007's. Looking at the waveforms. The regular diodes let through some residual ripple voltage which surprised me. When I switched in the Schottky diodes the resulting waveform was far more crisp in its cut off an preserved more of the fine radiant spikes and at a greater magnitude. I would love to see these fitted to all the outputs on my sg. The downside is a slip up that caused the neons to go on would likely burn up the diodes as well. They tend to not have a high voltage tolerance. The item of interest is that faster diodes have something to contribute to improving outcomes. I think Bro Mikey mentioned this in another thread!

[John_Koorn]

Hi James, I'm using ETX0806 in some of my experiments and they seem to be working very well. I don't know if they're Shottky but they are ultra-fast and rated at 8A 600V.

John K.

[James Milner]

SHOTTKY DIODES.

Well I fried my 40V 1A shottky's. They just don't seem to handle the output. I am now trialing some Ultra Fast Diodes 1A 400V and they seem to be handling it so far.......
I might consider trying some higher voltage/amp shottky's in the future and will post results as I get to the testing.

[bigmotherwhale]

Hi James, I'm using ETX0806 in some of my experiments and they seem to be working very well. I don't know if they're Shottky but they are ultra-fast and rated at 8A 600V.

John K.

Im using ebay 200915797977 UF5408 DIODE ULTRA FAST 1KV 3A, they seem prety cheap and good,
i had a diode out of a switch mode psu that i was using for a while in stan meyer hho experiments it began with FEP and it was a t220 package and that seemed to be about the best i have come across. it was 2 diodes in 1 package also.

[James Milner]

Sorry I've been so slow to get things done but that's what you get with a large family and shift work!
I have finally got round to dismembering and re-gluing my Monopole. This time with the two coils far more precisely positioned.
Works like a dream! Significantly more RPM, runs so cold as to not need a heatsink (although I'll put one on to mitigate against stuff-ups).

Like always its about precision!

PHASE LAG
When I looked closely at each coil I found that I had some excessive tolerance in the coil holder, particularly around the hole where the coils core material poked through. When I glued it together I unintentionally biased both to one side ( about 5 mm each). When done on the two coils this resulted in a variance of around 1cm. Because I was working on it from the rear of the machine I hadn't noticed (blocked by the coil holder!).
This would explain the phase of the secondary coil being out of step with the first. And why it generated heat and slowed down when correctly connected to the primary coils trigger. In essence it was working in repulsion mode for the slave and attraction mode for the primary.
James

[Volty]

Sorry I've been so slow to get things done but that's what you get with a large family and shift work!
I have finally got round to dismembering and re-gluing my Monopole. This time with the two coils far more precisely positioned.
Works like a dream! Significantly more RPM, runs so cold as to not need a heatsink (although I'll put one on to mitigate against stuff-ups).

Like always its about precision!

Hey James,

What did you choose? 1N4007 or 17nS 600V / 8A VS-ETX0806-M3? Did you ever learn why not one SUPER DIODE for all Trani's ganged to it?

http://www.mouser.com/Search/Product...-VS-ETX0806-M3 ? I just soldered up 4 8x SG Traini Busses with matched 1N4007's, and wish I'd have seen John's recommended 17nS Ultra Fast recovery diodes @ 600V / 8A.. Mouser.com 0.94 each US.

I found with SS SG w. One Wire w. CPD Mod, that 1N4006 diodes sound much crisper at lower audible frequencies, and I do not know why. I know 1N4007 conducts backwards some if above 17KHz. Anyway, JB shows moving Charge Battery Gnd from Run Battery + to Run Battery -, and IF done, Run Battery must be lower Volts, else drains DC to Charge Battery.

https://www.youtube.com/watch?v=TRLVhKzamUE
JB shows that here between 0:55 and 1:07 he has moved the Charge Battery (FWBR for Cap Dump) from Run Battery + to Run Battery -. So I did this for my Solid State SG w. CPD Cap Mod, and went nuts trying to learn why it ran terrible and sounded not crisp in firing any more. Then I accidentally discovered that for the Output Diode on Collector, that a 1N4006 was supremely crisp and superior in sound and results. Then I realized that Run Battery must be less Volts than Charge Battery else sucks Amp like piggy piggy. I kept the 1N4006 for my SS SG's, and moved the Charge Battery - back to the traditional Run Battery +. Solid State likes it that way i found.

See my babies.