Burning FET's and driver troubles

Things at the limits.

Re: Burning FET's and driver troubles

Postby Doug Coulter » Sat Nov 27, 2010 8:22 pm

Wow, if you're pulling the "on" fet a few volts off saturation there, you're going to have troubles later -- that's going to be a lot of watts loss into heat, with attending problems.
Hopefully I'm reading that trace wrong (is it 20v or so per div)? It could be an artifact of ground not being ground everywhere (check right across one fet, source to drain), so you may have to look at that one, in this case it would be good news actually, as it would mean something other than fets ->>>> smoke.

Good news on the floating scope test! That's unusual, it must be a good one. I know you can't pull that off with older Tektronix or Flukes. Haven't even tried it with my expensive new digitals.

Yes, you probably need to increase the snubber cap till the resistor get a little warm. I pulled those values from an offline switcher I took apart recently, and from the browning of the PCB, I'd say the R got warm to get the job done. but if you're not having other issues, it may not matter too terribly much. I've gone up to .01 in extreme cases, but that burns a lot of power at high bridge voltages and these speeds. Or burns up caps if they aren't really high quality ones.

It looks like you are hitting some issues with the transformer. You may need some series L. I get it here by just shifting some of the primary to another place on the core, it's free that way.
It looks like some kind of very high Q resonance at about 3x your drive F if it can really pull those large fets out of saturation. I'm hoping that you are seeing some other artifact, but I can't know.
To do that kind of Sherlocking, I generally have to be in the room, using all my senses to see the flaw that would be obvious in hindsight. And it really is Sherlocking -- you look at everything, not just the obvious stuff that's easy to put on paper, that hair on the back of the neck stuff, that fingerprint in an odd place, whatever (speaking metaphorically of course).

Do you know the resonance of your xfrmr? This sure looks like you have a fairly low voltage secondary and a pretty high resonance (because any high volt secondary will have a low resonance, that one's hard to overcome), you might have to go higher in drive frequency? That chip gets flakey over about 80khz, however. You might want to think about changing some details of the transformer. The fewer multiplier stages you can get away with, the better. If you need more volts/stage then your diodes and caps can easily do, you can simply series them up on a per stage basis -- same number of parts total, but fewer CW stages. The bottom diodes will thank you for doing that as current will be less on them in the first stage anyway doing that. This looks for all the world like you have a resonance about 3x your drive F, which can work, but it works better with leakage L in the primary so it doesn't all reflect back to the driver at very low effective impedance. Some of of this *should* get better with a load, by the way, as it will dampen the resonance a good bit.

The fusor environment is really hard on HV feedthroughs -- the hot H is a strong chemical reducing agent, and there's also the counterintuitive Paschen's law effects, where a discharge that cannot take place over a short path, can, over a longer one. If it's any consolation, my pal at CERN and my pal at ITER really couldn't help me much on that, I had to figure it out myself.
No one else really runs these conditions, so not much has been done here -- we are truly on our own doing new technology.

And it's still not completely figured out, but what I have is the best yet, easy to fix when it does fail, and I'm still working on improvements. It comes down to this. Alumina isn't a great insulator, but it's the only thing that can take getting hit with hot D, the chemical side product evaporates, and isn't conductive anyway. But since it's a crummy insulator, you need to make much of the feedthrough out of something else, and use the alumina to take the beating, while something else does the insulating. That wire in a hollow alumina cave design that is common just doesn't cut it here -- a couple thousand dollars later, that's pretty well established as fact. Also, even things with high dielectric constant get a lot of hits while the capacitor they create is getting charged up, and even that will make them fail after a number of runs. What I'm finding (and maybe I need to post over there some more about it) is that control over conductivity is key, having some non perfect insulator here, and some much better stuff there makes it all work as you'd like, or a lot closer. For example, in my design, the outermost tubing is pyrex, which is fairly conductive as these things go. The result is that it's grounded where it passes the tank walls, and nicely and evenly divides the voltage outside the tank (no dividers or corona rings needed, it's built in) and inside the tank largely stays near ground -- since it doesn't attract fast ions, it lives pretty well. Inside that I have quartz -- 3 orders magnitude better insulator, so that's pretty important, and it has a low dielectric constant to boot (but teflon would be better if it would take the conditions). The thin gap between the quartz and the inside and outside is important too -- prevents long-path arcs down the length of the thing, that pesky Paschen stuff, some surface effect I'm guessing. It is a tight fit on the main conductor, which in my case is a 3/8" copper or Al rod (depends on which one, I've made a few) that helps take out heat in the bargain. At the very end, I have the thick alumina over that rod instead of the quartz, which takes a beating, but it's only an inch and a half long piece (goes up inside the outer pyrex some instead of the inner quartz tube). So when it fails...no big deal, it's a few pennies to replace it -- I cut it with a diamond wheel in a toolpost grinder on my lathe, but I suppose there are other ways to get that done -- the stuff is hard to cut, and the old score and break doesn't work well with it (or quartz either).

One thing I've found (hopefully I save you some sorrow here) is that a lot of layers, other than needed because you need the different properties, isn't that great an idea. If you stand off 50kv with ten 5kv insulators, it's not going to work. One will punch through, and the rest will cascade fail immediately. Methinks that Kohl book would do you some good with this, worth it at the high price, even.

This is a real problem -- like you said, we're putting the sun (actually, something a lot nastier in many ways) in a box (hey the box for the sun is space, remarkable resilient stuff), so we have to learn to build the box. And since I'm about to go to the 100kv+ range, I suspect I'll have to do more box building myself soon. Right now, all this gets into the tank with an O ring compression tubing coupler, and it's looking like I'll need larger diameter than the 1" I'm using now, sigh -- those things, and big tubing, aren't inexpensive.
Posting as just me, not as the forum owner. Everything I say is "in my opinion" and YMMV -- which should go for everyone without saying.
User avatar
Doug Coulter
 
Posts: 3515
Joined: Wed Jul 14, 2010 7:05 pm
Location: Floyd county, VA, USA

Re: Burning FET's and driver troubles

Postby JonathanH13 » Thu Dec 02, 2010 5:08 am

>It could be an artifact of ground not being ground everywhere (check right across one fet, source to drain)

Yes, that is probable - my ground setup has become a bit mysterious, I just connect things up so as to prevent the power tripping in my apartment - I keep promising myself to figure it out properly. I will include an existing ground setup in the new schematic. For now, I have taken a new measurement across one fet (HO2), source to drain. See pic below. The 'mexican wave' artifact seems to be a lot better.

Still getting no joy from the snubber - I am up to 3nF on the cap side. I have just received some new caps, they will push it all the way up to 10nF - that will probably do it. I may try and change the gate resistor values as well. Or include a gate resistor for 'turn on' and then another gate resistor of a lower value for 'turn off' (with a diode in series, all in parallel with the 'turn on' resistor). That way I can dampen that turn on spike, but still ensure that the fets turn off fast enough.

>It looks like some kind of very high Q resonance at about 3x your drive F

This transformer is designed to resonate at 40KHz. What the value actually is I'm not sure (and am not sure how to determine this value, outside of having it all wired up in circuit). Is it possible to do some stand-alone measurements with a scope or multimeter to work this out without having an expensive LCR meter?

>The fewer multiplier stages you can get away with, the better.

I intend to rebuild the stack anyway - rearrange it into a 'triangle', put in some resistors and copper rings, and dump the entire thing in oil. That would be a good time to reduce the stages...
Attachments
drain to source ho2.jpg
drain to source ho2, 20V/div
User avatar
JonathanH13
 
Posts: 53
Joined: Tue Aug 10, 2010 3:20 pm

Re: Burning FET's and driver troubles

Postby Doug Coulter » Thu Dec 02, 2010 9:15 am

That's a much nicer looking waveform, I must say. Notice you can see the time-varying current draw during the fet on time? It's a tiny resistor after all, and in the right place to use for that.
That spike on turnoff is due probably to turning off too fast, or mistuning. The trick for asymetric on-off times uses a single resistor with a diode across it as the gate R most often, though a diode is a "hard" conductor and sometimes needs a resistor in series with it too (usually much smaller than the one it is paralleling). But again, most of that spike might not really be there if the current is high at the time of turnoff -- could be picked up on the scope probe via magnetic or cap coupling.

You can see current is peaking right at turnon, once in the middle, and again at turnoff -- and that's why you see the big spike at turnoff, at that point a lot of stored energy in the output circuit that has to go someplace.

The easy way to measure resonance (at low levels) is simply to hook a resistor in series with the primary and use a signal generator to sweep over a large frequency range. At resonance, the drop across the resistor (or the volts across the primary) will show that the transformer is drawing less current. I've got at least one that shows several resonances. What I see in the current waveform is one about 3x the drive frequency. I've made that work here, but it's a probable cause of the kickback at turnoff. You want to fine tune for minimum quiescent current draw (with no bleeder load on the stack) in any case. That resonance test is pretty easy (if you have a signal generator) and will get you close. It will change a bit at full power drives as the magnetic material isn't perfectly linear. But not a whole lot, just enough to make a slight retuning a good idea.

You may not see a ton of change on the waveform with a snubber until you're getting the resistor a bit warm at full supply voltage. Whether you see it or not, it's helping the fet have a nicer working environment by controlling some of the wacko motions during the deadtime (which in my opinion is a little too long in this chip).

If you change your stack to fewer stages with more parts (and volts) per stage, you'll need a higher transformer output voltage to get the same results as before. I'd not go to oil if I had any other choice. Very messy when repairs are needed (soldering oily stuff is no fun either)...good for cooling things, but in general, these things shouldn't get too hot anyway. The only time I've had heat issues is with 12 stage multipliers, where the diodes at the bottom are seeing 12x the output current, that got them pretty warm. The downside of oil is also that it has a few times the dielectric constant of air, so will act like further capacity across the transformer when in oil. That can be good or bad -- it depends on what resonant frequency you had, and want.
Posting as just me, not as the forum owner. Everything I say is "in my opinion" and YMMV -- which should go for everyone without saying.
User avatar
Doug Coulter
 
Posts: 3515
Joined: Wed Jul 14, 2010 7:05 pm
Location: Floyd county, VA, USA

Re: Burning FET's and driver troubles

Postby JonathanH13 » Sun Dec 05, 2010 8:34 am

I have attached the latest schematic, complete with earth connection, zeners, snubber, etc. I may still add return diodes on the fet gates. As you can see from the schematic, both supplies are floating. This is the only arrangement that I can get to work. I also performed that resonance measurement on the transformer - there is a definite volt drop across the resistor (110ohm) at 63KHz. This is a big surprise for me! I modified the driver to be tuneable at around 60KHz, then I reconnected everything and tuned for lowest current on no load around that frequency. I will post some wave new voltage waveforms and some current measurements ASAP. For the moment, it seems to be working fine at its new frequency...
Attachments
Hbridge Schematic2.jpg
Xfmer Freq setup.jpg
Xfmer Freq setup.jpg (29.96 KiB) Viewed 4933 times
KVxfmer Freq Response.jpg
Last edited by JonathanH13 on Sat Dec 11, 2010 5:17 am, edited 3 times in total.
User avatar
JonathanH13
 
Posts: 53
Joined: Tue Aug 10, 2010 3:20 pm

Re: Burning FET's and driver troubles

Postby Doug Coulter » Sun Dec 05, 2010 6:55 pm

Oh, you'll always have some loss -- luckily (!) the transformer doesn't have infinite Q or you could only drive it with a sine wave and have the drivers live. In fact, that's what the series LC in the trick circuit I posted above kind of accomplishes.

I'm not surprised you need to float the supply at all. When you get currents as high as the bridge is seeing, it's all too easy to have ground circuit drops and noises get into things you'd rather they didn't. There are tricks (star grounding) that often work, but sometimes take more effort that just having what amounts to an extra wall wart in the thing. I've been known to do just that and simply solder to the AC input pins to avoid the wasted space on some power strip I plug the whole thing into. Cheap, quick, floating, all the good stuff.

If I were to find I needed more gate drive, I'd use more or less a circuit I saw in the Pressman book. Basically just two transistors, one NPN, one PNP, emitters together, bases together, collectors at the (driver, 15v) rails. I would have about a 47 ohm resistor between the bases and the emitters to avoid a bit of crossover glitch, and some small (10-20 ohm) series R to the fet gate. For the lower sides of the bridge, the "rails" for these emitter follower buffers are obvious - the negative rail and the same power the chip is running on. For the upper halves, the lower "rail" is the fet source connection for the upper fet, the the upper rail for the buffer will be the Vb pin on the chip. If it turns out that this is too much current drain for the bootstrap switch to drive, you could add a diode like they do with the older chips that didn't have that built in. For example, this datasheet shows one. This needs to be a fast, HV diode to work, as it must stand off peak input HV minus the chip supply when reverse biased.

irs2304spbf.pdf
Old style half bridge driver that needs a diode
(1.29 MiB) Downloaded 329 times


I'm planning to use this chip BTW for the turbo driver pump controller I'm breadboarding right now. It will drive a half bridge to control the input main DC and control maximum motor current.
Using two fets in a buck converter is a little better than one and a diode as the when the bottom fet is on, the drop is much less than a diode drop, and hey, I have these chips around, and wanted the high side driver anyway so as to be able to use all N (in other words, cheap) fets.

Now, if you're really on the resonance, if you look at a bridge output with one scope channel, and put the probe from the other near enough to the secondary to pick up some of that, you should see a nice 0 or 180 degree phase relationship -- with the secondary waveform looking a lot more like a sine wave than the input to the primary does. You can expect the resonant frequency to change a bit with drive voltage -- if it starts going up fast with drive, it means you're getting into the bad part of the ferrite BH curve and are going to have other "issues" due to that. Near saturation is good, actual saturation is very bad. There's a fudgy area there in ferrites (they don't saturate "hard") where the main penalty is a lot of core losses and heating.

The reason you'd just as soon be near saturation is that in an arc situation, less of the bad stuff can blow back through the transformers into the drivers, and you're getting your money's worth out of the winding wire as well as you can -- max volts per turn. If the transformer is already nearly saturated, it's not as possible for 10x the normal rail voltage to appear blowing back from a messed up voltage multiplier -- I have experienced this and it's time for all new semiconductors when that happens.
Which is another reason to not skimp on the capacitors across that main rail -- to make them hard to overcharge to high voltages in that event.
That one's a tradeoff, like always, since if the spike is hot enough, it simply fries the diodes in the circuit instead with forward current.
Posting as just me, not as the forum owner. Everything I say is "in my opinion" and YMMV -- which should go for everyone without saying.
User avatar
Doug Coulter
 
Posts: 3515
Joined: Wed Jul 14, 2010 7:05 pm
Location: Floyd county, VA, USA

Re: Burning FET's and driver troubles

Postby JonathanH13 » Fri Dec 10, 2010 3:19 am

Some more results: After having tuned the driver circuit to 63KHz, I had a look at the voltage waveforms on the gate of one of the bottom fets in the bridge, compared with the output (drain) of that same fet (see first pic). Whilst watching the waveforms, I tuned for lowest quiescent current in this frequency region. At 66.8KHz I heard a distinct buzzing, sizzling sound (which did not sound very healthy). I'm presuming that this is the entire system hitting resonance - the variac current dropped right down to 175mA and the output waveform became really unstable. I reduced the frequency just off this point, down to 65KHz, which pushed up the current slightly (226mA), but stopped the noise, and stabilized the waveform. So I think I understand what you mean now - you want to be close to resonance, but not on resonance. The more I increased the voltage, the more I had to reduce the frequency to keep the waveform stable (and prevent it hitting resonance), and at 40KV the frequency is down to 60KHz (out -of-circuit measurements made on the transformer in my previous post put the Fres at 63KHz). There is a slight glitch on the bridge output under no load, but it does not seem to disturb the secondary much. In fact when the glitch is not there, the secondary voltage starts to look almost triangular, and the whole thing is too close to resonance. Under load the bridge output is distorted and there is some ringing (see last wave pic)

I found a similar low quiescent current point at around 36.5KHz, although the output waveform is not as smooth - this seems to deliver the same KV voltage for the same current draw on the variac (but I did not test it under load).

Itried to look at the frequency response of the transformer on the network analyzer at work, but the analyzer only starts at 300KHz. This driver circuit appears far more stable and resilient than the previous one (controlled by a TL494), despite driving fets that are somewhat large for it. The driver IC seems to take the hit before any fets burn, which is fine by me (and which has not happened since I put the zeners in). I have attached the current artwork.
Attachments
2.jpg
red = gate, blue = output (drain) @60KHz
3.jpg
red = gate, blue = output (drain) @36.5KHz
5.jpg
black = deadtime, red = gate, blue = output (drain) @60KHz
6.jpg
red = xfmer primary, blue = xfmer secondary @60KHz
7.jpg
red = gate, blue = output (drain) @60KHz, with plasma load
kv driver final scaled.jpg
current version pcb layout
User avatar
JonathanH13
 
Posts: 53
Joined: Tue Aug 10, 2010 3:20 pm

Re: Burning FET's and driver troubles

Postby Doug Coulter » Fri Dec 10, 2010 9:44 am

Pulling the fets out of the on state looks un-good to me if I'm interpreting that scope shot correctly.
If you look at the on-resistance specs for the fet, this implies some amazing-big currents you shouldn't have no-load or low load. It doesn't add up with the quiescent current you're reporting.

Something is still a bit odd. I'd love to see a pic of that transformer.

You might need to widen your sweep of the transformer characteristics. One of the ones I'm working with shows 3 resonances and only one of them is the right place to run.
Having the resonance rise with drive is normal (a little bit) but too much means you're trying to get more than the core can do in V/turn at this frequency.
Posting as just me, not as the forum owner. Everything I say is "in my opinion" and YMMV -- which should go for everyone without saying.
User avatar
Doug Coulter
 
Posts: 3515
Joined: Wed Jul 14, 2010 7:05 pm
Location: Floyd county, VA, USA

Re: Burning FET's and driver troubles

Postby JonathanH13 » Fri Dec 10, 2010 11:37 am

Which scope shot are you referring to? The last image (with the blue trace slopes down and to the right after turn on) is with load, and is pulling at least 2 amps.

I will post an image of the transformer and some specs - it is a bit unconventional. And I will check those on-resistance specs...
User avatar
JonathanH13
 
Posts: 53
Joined: Tue Aug 10, 2010 3:20 pm

Re: Burning FET's and driver troubles

Postby Doug Coulter » Fri Dec 10, 2010 12:56 pm

Oh, sorry, the bottom one -- a few amps should not pull a fet out of saturation by volts -- that would imply an ohm or more "on" resistance on the fet -- and the ones you're using are in the milliohm range when on I think -- the smaller ones I use are. That or some measurement issue (ground not really ground or something else dropping volts). I see you're having the same issues with close ups on scope screens as I -- hard to read the legend on volts/div in these. I'm usually to lazy to use the scope itself to save a bitmap on its SD card, as it takes a long time to get into the computer for posting here, myself. But that bottom shot shows (I think) volts happening across a fet that's supposed to be "on" at the time, and at most you should see millivolts there.
Posting as just me, not as the forum owner. Everything I say is "in my opinion" and YMMV -- which should go for everyone without saying.
User avatar
Doug Coulter
 
Posts: 3515
Joined: Wed Jul 14, 2010 7:05 pm
Location: Floyd county, VA, USA

Re: Burning FET's and driver troubles

Postby chrismb » Fri Dec 10, 2010 1:13 pm

Jonathan, what is feeding your supply rail, and what current draw do you think the load is pulling? The chart says 'rectified AC' with some capacitors. I tend to employ an excess of capacitance 'downstream', so to speak, and a low-ish ESR cap of a uF or two up close to the FET so it has a good 'stiff' supply. I'm just wondering if I'm looking at some plots here showing a supply on the rail that is a bit weak and folding back under load.
chrismb
 
Posts: 620
Joined: Thu Aug 05, 2010 6:32 pm

PreviousNext

Return to High Voltage, High Power

Who is online

Users browsing this forum: No registered users and 6 guests

cron