Wireless powering SSTC / Tesla Coils / Forums

Forums

4hv.org :: Forums :: Tesla Coils

« Previous topic | Next topic »

Wireless powering SSTC

first previous 2 3 4 5 next

Move Thread

LAN_403

Marko

Sat Nov 24 2007, 11:11AM

Registered Member #89 Joined: Thu Feb 09 2006, 02:40PM
Location: Zadar, Croatia
Posts: 3145

Great replies guys... but apparently I'm not happy if they contradict each other!

Well now you have. It does indeed work. I made an sstc with primary current feedback.

Any more information on the coil? Pics? Schematics? Because it shouldn't work according to Steve.

Why doesn't the secondary force the primary to resonate in this case? The way i see it is because the coupling is always too low to make it work. The system has to have a full current reversal in order to make it oscillate (otherwise you never reverse the output). If you apply a step input to your primary coupled with secondary (and assuming no DC blocking cap, or its infinitely large) then you will get something that oscillates, but not a full swing like you do for a second order LC system (where you get a natural harmonic oscillation). Instead you will see a typical

Steve, I understand what are you talking about, but still it (for my understanding

) ''doesn't fit'' with what I thought to be normal SSTC operation...

Those were things I was never able to simulate or get a picture off. How did you come to need 0f k > 0.98 in order to make the current ''undershoot''?

In that case primary feedback would probably only be feasible with ferrite transformer -> base feed to resonator.

Now I think, doesn't current in most normal SSTC's (low k) go down to zero on itself, anyway?
If it didn't it would be hard switching, right?

(Isn't that what you would call ''full current reversal'')?

So if primary current can ring down to zero even if it has 325V over it implied that one could use feedback. (I must admit it isn't so crystal clear then

)

So who of you guys is right?

Surely this can only be caused by the mutual inductance between primary / secondary, in which case could the primary feedback work in theory for an SSTC if the coupling were high enough? I would have thought it would be possible, but not without some slight phase lag on the feedback.

(or maybe I should just turn my computer off and go to the pub..!)

Avalanche: you think exactly as I do. I don't know about the phase lag... but SSTC's with feedback taken directly by an air windings near secondary appear to have oscillated well.

Marko

Steve Conner

Sat Nov 24 2007, 12:32PM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

What the other Steve said!

The problem in a nutshell is the magnetizing current, which never reverses. If the reflected sinusoidal current from the secondary is too small to overcome it, the primary current will never pass through zero, so the switches will never flip, and the current will keep rising indefinitely.

In previous posts I've explained how streamer loading damps the secondary and limits the amount of current it can reflect. So the result would be a coil that seems to work fine at low power but explodes at high power, which is the case for most hobbyist coils anyway

Steve Ward

Sat Nov 24 2007, 08:19PM

Registered Member #146 Joined: Sun Feb 12 2006, 04:21AM
Location: Austin Tx
Posts: 1055

I must say im a bit confused as to how Tom's primary feedback SSTC works. Some more details on the setup would be useful.

Marko, i got the K > .98 from a simple pspice simulation. Thats where it looked feasible to work, but its probably very dependent on the values i used, but its likely to never look "really good" with a K of .5 like SSTCs are capable of.

One thing i didnt take into account is that its possible to make a current feedback transformer that only "sees" the secondary resonance "imposed" back upon the primary magnetizing current (since the Imag will be of much lower natural frequency). The issue here is to not saturate your CT before it can work properly. This seems about as reliable (or not as reliable) as secondary base feedback, though. In all cases you are subject to loading the secondary too far, and losing the conditions necessary for oscillation (in which case its nice to have a PLL take over for awhile).

As Conner mentioned about most SSTCs blowing up at high power, my big one only works because of using massive 80A mosfets. The primary current starts looking "bad" at near full power (no longer ZCS by a long shot!). In fact, it *looks* like it switches early, but really its because the Imag > Ireflected (meaning the primary current may not actually cross zero naturally), so you dont get real ZCS, instead it switches when the current drops to maybe 50% of its peak value. During arc loading conditions, the primary current looks like an RC low passed square wave (in this case its really L/R low passed, but you get the point) where the switching occurs at the maximum primary current of ~50-60A.

Marko

Sun Nov 25 2007, 01:29AM

Registered Member #89 Joined: Thu Feb 09 2006, 02:40PM
Location: Zadar, Croatia
Posts: 3145

OK - if I figured you guys out, I might make the primary feedback work while I'm fully soft switching (high Q, sine current) but I would lose it and blow up as soon as any hard switching appears?

I'm also interested what is with Tom's coil. Hope to see a reply from him soon.

Steve Ward

Sun Nov 25 2007, 02:52AM

Registered Member #146 Joined: Sun Feb 12 2006, 04:21AM
Location: Austin Tx
Posts: 1055

I might make the primary feedback work while I'm fully soft switching (high Q, sine current) but I would lose it and blow up as soon as any hard switching appears?

Im not 100% sure, but it may be possible to make the thing work with primary feedback if you can make it start up. I think what happens is that as the secondary stores up more energy, its impact on the primary current is greater (which is what we want). But, as soon as you draw an arc or whatever, the Q drops off and you get the hard switching conditions i mentioned above, and at that point it will fail to be an oscillator. So i think you need both high Q and sufficient energy stored in the secondary circuit for it to work (that is, Ireflected >= Imag... i think). Oh, and always use a DC blocking cap, since it will save your bridge if you do fail to oscillate. I think thats what fixed my big SSTC when it blew up before, though i dont want to risk removing the caps and blowing it up again to find out :P.

I hope someone else can chime in here, since im a bit uncertain about what i said ^^.

Tom540

Sun Nov 25 2007, 09:06AM

Banned on 3/17/2009.
Registered Member #487 Joined: Sun Jul 09 2006, 01:22AM
Location:
Posts: 617

The circuit I used was the exact same as I used for my other coils. Just a 60 turn CT clamped into a 74hc14 then fed into the gate drivers and my interrupter was my signal generator. It wasn't very reliable. the fets got hotter than hell and the big sparks you see in the video were only when I put my fingers on the CT did they go from about 2 inches to 8 or so long.

I'd have to give this another try and although I didn't actually tune the primary it might have been close by chance or maybe on a harmonic. I might give this another try tomorrow. Either way the coil was definitely NOT a DRSSTC and my interrupter on times were very long.

CT2

Sun Nov 25 2007, 09:14AM

Registered Member #180 Joined: Thu Feb 16 2006, 02:12AM
Location: Ontario, Canada
Posts: 187

If the circuit has a 555 as a start up signal on the input aswell as the primary CT then when the coil starts to break out and it stops oscillating the 555 should start it oscillating again, shouldn't it? But I guess thats no longer feedback, if it's always arcing then huh?

Marko

Sun Nov 25 2007, 10:16PM

Registered Member #89 Joined: Thu Feb 09 2006, 02:40PM
Location: Zadar, Croatia
Posts: 3145

OK, I'll understand this as _don't ever use primary feedback on a SSTC_ :p ... still, I'd like to see anything about tom's coil!

Steve: Didn't your coil fail because of too high magnetizing current caused by too few primary turns?

Reading your posts a while ago and pondering actually made me exclude the DC blocking caps from the design. What do you now say about that? I don't know how would I put them back now... but still, I'm not using PLL, and topology is actually going to be weird and not-well-tested... maybe I should really use DC blocking caps

I wonder, what can go wrong on that part? Especially if you are using PLL...

Still GDT would stop any significant DC going through...

Don't DRSSTC's start up simply by putting a quite large pulse (full on time) onto the GDT and this way pinging the system up?

I would use a startup pulse not longer than few us, and if oscillation collapses both drivers should go to 0.

(Look at my hand-drawn schematic.)

Hmmm....

Steve Ward

Sun Nov 25 2007, 10:42PM

Registered Member #146 Joined: Sun Feb 12 2006, 04:21AM
Location: Austin Tx
Posts: 1055

Steve: Didn't your coil fail because of too high magnetizing current caused by too few primary turns?

My memory is starting to go on this one... It did fail when i tapped my primary down to 3 or 4 turns i think. But, i also added a DC blocking cap in after the failure, and its been fine since then, but i dont think i tried tapping it at "3 or 4" turns again. At the time i figured there could be a DC bias conveyed even through GDT driving. The reason i suspected this was that my PLL chip always seemed to give non 50% duty cycle (always a little off), and this just results in gate drive waveforms that are shifted up/down, but the output duty cycle from the bridge is not 50/50. I figured that this might have caused excessive DC current in the output, so put a 1.32uF film cap in there to be safe. But, the only thing that seems to invalidate my DC offset theory is that the primary circuit pretty much looks like zero impedance to any DC current (surely the 3/8" diameter tubing i used isnt very resistive), so it should have been a problem no matter what tapping point. Perhaps it was a combined effort (larger AC current and some DC current) that really made it fail.

I think that now my opinion is that DC blocking caps are a good thing when they can be used such that they dont interfere with normal operation. If i made you think otherwise, then i take it back. I seem to have a bad habit of saying things, and then learning some other finer details, and then i have to change my stance on the topic.

I dont see why it would be difficult to include a DC blocking cap in the design? Just string it in series anywhere, right?

Don't DRSSTC's start up simply by putting a quite large pulse (full on time) onto the GDT and this way pinging the system up?

Yeah, its like applying a step input to the primary system. No harm done even if the output transistors never switch around (you just end up with a charged tank cap to Vcc). The same would apply to a SSTC with DC blocking cap, without the cap the DC current (same as magnetizing current in this case) will just shoot up until the resistance of the circuit limits it (really too high). I wouldnt rely on saturating GDT or decaying output from CT to turn the fets back off, a DC blocking cap would be the most sure-fire way.

BTW, your control setup looks like it should work to me (at least fairly well). I suggest you try it out and see how it works, and i would recommend using a DC blocking cap of ~1uF (something large enough to avoid resonance).

Marko

Sun Nov 25 2007, 11:25PM

Registered Member #89 Joined: Thu Feb 09 2006, 02:40PM
Location: Zadar, Croatia
Posts: 3145

Steve: don't you have a DC block cap on your GDT anyway (it would saturate first on uneven duty cycle).

I think that now my opinion is that DC blocking caps are a good thing when they can be used such that they dont interfere with normal operation. If i made you think otherwise, then i take it back. I seem to have a bad habit of saying things, and then learning some other finer details, and then i have to change my stance on the topic.

Not really directly... but after you advised me to use bifilar primary for 2 paralleled bridges I thought they are just a hassle for the topology... But yeah, definitely better to use them.

BTW, your control setup looks like it should work to me (at least fairly well). I suggest you try it out and see how it works, and i would recommend using a DC blocking cap of ~1uF (something large enough to avoid resonance).

I will... as soon as I get on with school... and polyprojectitis. DRSSTC is still sitting broken and I even started a third thing...

Marko

first previous 2 3 4 5 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob