Kapanadze Cousin - DALLY FREE ENERGY

[NickZ]

Nick,

those 300V spikes are there even with a 24V (or 2 x 12V in series) bulb in the 24V supply line?
Try to go down to 12V for now to lower these spikes or you might go for a higher voltage MOSFET like some IRFP460.

We are not trying to dampen the spikes yet by adding a drain/source cap, we first try to get half of the ringing signal by adding a cap.
This is a first step to continue to build / calculate an RC snubber.

Guess with a 271 tagged cap you mean 471  (470pF), if so, 2 of them in parallel (940pF) would probably be a correct value to half your ringing frequency, but
use your new scope to take a screenshot of that ringing both expanded as un-expanded.

Itsu

  Itsu:
  Yes, those reading were with the two 10k resistors on the yoke input. Not the full 24v.
  The cap that I mentioned is a 271, not the 471. It's a blue HV 14D271K capacitor, possibly 2000v, but it doesn't give the voltage, just the K at the end. The 152 cap is a 1000v, ceramic capacitor. I also have lots of the blue 472M caps, and some others as well.
   Ok then I'll try to half the ringing spikes, and not worry to much about the voltages for now.
   Once I have the snubbers calculated and working, I'll install the new fets on just 12v, to see what happens. Or even try it with a small 12v 5 watt solar panel with has an open reading of 22v.
    So also for now, I'll assume that there is a false reading of about 25% higher than the actual reading. Which may make the difference of being able to drop the running voltages to under 200v.

  Hoppy my UC4420 drivers are only about 1.5 inches away from my fets. So, I'm not going to change that as yet, as it's means a whole new board, with new drivers, and much work, also. I think that I can get the voltages down, without having to do that. We'll see in any case.
   I'll be going to another town today, and will report back concerning these changes later tonight.

   Itsu: maybe try to see if you can get your spikes down under 200v,

[Hoppy]

  Itsu:
  Yes, those reading were with the two 10k resistors on the yoke input. Not the full 24v.
 
So also for now, I'll assume that there is a false reading of about 25% higher than the actual reading. Which may make the difference of being able to drop the running voltages to under 200v.

  Hoppy my UC4420 drivers are only about 1.5 inches away from my fets. So, I'm not going to change that as yet, as it's means a whole new board, with new drivers, and much work, also. I think that I can get the voltages down, without having to do that. We'll see in any case.
   

Nick,

Please explain how the 2 x 10K resistors are connected to the yoke input?

Don't assume, measure correctly to be sure!

[itsu]

Itsu:
  Yes, those reading were with the two 10k resistors on the yoke input. Not the full 24v.

Huh??  not sure what you mean by that.  Do you have 300V spikes when running your MOSFETs into a 10K resistive load??
I have seen your screenshots of those, they had these rounded off turnon times,
like here:  http://overunity.com/12736/kapanadze-cousin-dally-free-energy/msg508556/#msg508556
But no spikes seen there.

So what do you mean with the above statement?

The cap that I mentioned is a 271, not the 471. It's a blue HV 14D271K capacitor

A 271 tagged capacitor is 270pF, but if you mention:  HV 14D271K   then i think its NO capacitor but a varistor:
http://www.bourns.com/docs/Product-Datasheets/MOV14D.pdf

Itsu: maybe try to see if you can get your spikes down under 200v,

Working on that, i have seen a lot of smoke in the mean time, these spikes are not easy to tame.


I don't think is sensible to use (partly) defective MOSFETs in calculating a snubber.
Chance is that you have to redo the calculations again when the new MOSFETs are in.

Just put in the new MOSFETs and put the drains on 12V with some 12V bulb protection.
Use your new scope to show the ringing, and try to half that ringing using a capacitor across the drain / source.
Then we take it from there.

Itsu

[Grumage]

Nick
the old girl will always be needed ..to keep your new one safe during unknown or sensitive testing [unwanted scope sizzle]

also never hesitate to ask for connection advise for "sniffing" [not direct connections] during possible scope sizzling events.

TK and others have good protocols for this [on the cheap], the newer scopes can be less forgiving in the "unintentional abuse" department.

--------------------------

Matt is a good man indeed !!



with gratitude and respect

Chet K

Dear Chet.

Wise words, indeed....

This one's for Matt.

Cheers Graham.

[NickZ]

Huh??  not sure what you mean by that.  Do you have 300V spikes when running your MOSFETs into a 10K resistive load??
I have seen your screenshots of those, they had these rounded off turnon times,
like here:  http://overunity.com/12736/kapanadze-cousin-dally-free-energy/msg508556/#msg508556
But no spikes seen there.

So what do you mean with the above statement?


A 271 tagged capacitor is 270pF, but if you mention:  HV 14D271K   then i think its NO capacitor but a varistor:
http://www.bourns.com/docs/Product-Datasheets/MOV14D.pdf

Working on that, i have seen a lot of smoke in the mean time, these spikes are not easy to tame.


I don't think is sensible to use (partly) defective MOSFETs in calculating a snubber.
Chance is that you have to redo the calculations again when the new MOSFETs are in.

Just put in the new MOSFETs and put the drains on 12V with some 12V bulb protection.
Use your new scope to show the ringing, and try to half that ringing using a capacitor across the drain / source.
Then we take it from there.

Itsu

  No, you're right. I was running the full load on now, when using the older fets, I was on the full 24v, and not running them through the 10k resistors, anymore. And now, I do see 250 to 300v spikes, without any snubbers caps on.
   So, ok I'll put the new fets on, and run them off of a 12v, 7ah battery. And leave the 24v, 18w car bulb on for now, (which has not been taken off, yet).
    So, I was running the full 24v, but through the car bulb first, up to now. So, maybe it was on full 24 volts, but it was not running on full amperage. I hope that clears thing up. Sorry for the confusion, my fault.

   I'll check the blue HV 271 (which does not look like a varistor), but it may be, as it has different lettering or numbers on it, and if so I won't use it. I'll see if I can find one more 1.5nf 1000v, to match the single one of those that I do have.

   Although it's been 10 days since the digital scope was sent to me, I haven't got it yet. It should arrive any day now.

   I did pick up 3 bigger 200w, 130v bulbs at a hardware store, along with some plastic pvc 1 1/4 inch 90 degree curbs to make some new side supports for my grenade. 
   I'll also see what I can do to get the ringing down to 1/2.  As you can see, it's not easy to do, without killing the ringing, and still be able to get the voltage under 160v or so.