DRSSTC, yes another! / Projects / Forums

Forums

4hv.org :: Forums :: Projects

« Previous topic | Next topic »

DRSSTC, yes another!

previous 1 2 3 4 next

Move Thread

LAN_403

Part Scavenger

Fri Jun 08 2007, 09:51PM

Registered Member #79 Joined: Thu Feb 09 2006, 11:35AM
Location: Arkansas
Posts: 673

Yikes!

Marko

Fri Jun 08 2007, 10:14PM

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

Steve Ward wrote ...

Does this same breaker supply power to your control logic? If so, then that is likely a problem. The fix (for my case) was to install a 120V coil relay across my interrupter signal. If my control logic loses power, the relay grounds out the signal (the interrupter has 100 ohms output impedance for safety from this short). This is how i lost about 4 CM300s awhile back

Steve, I don't see a failure mode with your classic circuit wich could lead to this, spare the brick driver.

At point where UCC's go UVLO IGBT's should still ave 10 or so V of gate drive and should be fine.

I was thinking about what would happen if LV supply lost power and I never seen such problem happening to anybody.

Scavenger, please give us some insight into your power supply.

If you are scoping the DC bus, you must be using isolation transformer, right?

And I guess you aren't powering the control logic from a variac, unless it has a spare outlet?

BTW, I think I also know what was blowing your GDT.

Part Scavenger

Sat Jun 09 2007, 01:15AM

Registered Member #79 Joined: Thu Feb 09 2006, 11:35AM
Location: Arkansas
Posts: 673

My power supply:
120V to isolation xfrmr to variac to doubler to bus. The controller has a separate outlet, but that outlet is shared with the isolation transformer. The controller never lost power, and I was not scoping the bus at the time (no probes were connected).

The 5A breaker on the relay popped, the coil went silent. I actually expected to turn it back on and run it again, but no. Maybe the IGBT's did die first. I upped the current to 160A and I was tuning it. At the time I was getting a very small notch and my current was jumping up so quickly that my burst length and spark length was being limited alot by the OCD. Something is still weird though, my 0'scope was reading exactly 320A peak, which was about where I would have expected failure. So, if I was running right on the edge, and my variac tripped, would the voltage increase for a few cycles enough to spike the current? I've got kinda small filter caps you know... Like I said, they could have died first, but it NEVER happens that way. It also could have been that my breaker was about to trip when the IGBT's let loose and so it fired immediately... Really, I need to provide more sound information to officially chaulk this failure to anything. I just wanted to know if this was a common thing, and if there was some way I should get around it.

I think I figured out what was causing that 1A draw when I suddenly shut off the interuppter, but I need to re-check the scope shot to see if I remember it correctly.

"BTW, I think I also know what was blowing your GDT"
Sure! Another brain is better than mine. ;)

Marko

Sat Jun 09 2007, 11:08AM

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

Hi.

"BTW, I think I also know what was blowing your GDT"
Sure! Another brain is better than mine. ;)

As I was running my coil with isolation transformer I noticed arcing between PCB traces on my bridge. The entire DC bus, being isolated, was elevating it's potential to probably several kilovolts by capacitive coupling to coil's output.

When I disconnected my scope cables, wich were grounding the circuit board, the arcing stopped.

These could have equally as easy blown my GDT or silicone insulators.

I think very similar thing was happening to you, you vere running the entire thing from an isolation transformer and your variac, your capacitors and wiring were elevating it's potential over the control logic until ultimately blowing the GDT.

The small arcs from capacitive coupling don't seem to have enough energy to do any harm to onboard components, but they will break IGBT gate and carbonise the insulation just enough to start up a power arc from your bus capacitors.

I think that was a naturalistic outcome.

The simple solution to this could be just using a nonisolated power supply, or much better, using proper DC blocking caps from DC bus to ground, and ensure that it is always grounded.

Your new GDT may be better but you are still risking chances of blowing it up.

Regarding your second failure, I'm not yet sure what could it be... I hoope Steve will put some more enlightment on it.

PS. I just put two 680nF caps between DC bus and ground and arcing stopped.
I'm now pretty sure that this was what was blowing your GDT.

Steve Conner

Sat Jun 09 2007, 05:30PM

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

That was a nice piece of detective work, Firkragg

I never thought of that failure mode myself, but I always used a small capacitor between the DC bus negative and the AC line ground that the driver board was grounded to. Also, when running off an isolation transformer, I went one step further and jumpered across the capacitor to short the negative rail solidly to ground.

I never thought it through, it just seemed to be a sensible kind of thing to do.

Marko

Sat Jun 09 2007, 07:46PM

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

Meh, not so fast..

I still get arcing in that place when drawing arcs from toroid and hitting higher peak current.
I did a bunch of changes but I still can't help it... see my DRSSTC thread :(

Obligatory cool pic of arcing between bus and CT output:

(Maybe I should write 'OWNED' with big black letters in the footer of this pic)

Danielle

Sat Jun 09 2007, 10:09PM

Registered Member #632 Joined: Mon Apr 09 2007, 01:09AM
Location:
Posts: 267

I wounder the voltage across there if its low enough you could separate the traces also you should look into further decoupling of the gate that may help.

I may sound stupid saying this so please tell me. I am just giving sudustions because that has never happened to me.

Part Scavenger

Sun Jun 10 2007, 12:20AM

Registered Member #79 Joined: Thu Feb 09 2006, 11:35AM
Location: Arkansas
Posts: 673

Firkragg, Dude! Another example of why I tell people you are brilliant. Your comment reminded me that disconnecting my o'scope probes made my heatsinks run cold instead of warm... Man, I had no idea what I was getting into with all this SSTC mess. I am TOTALLY not qualified, but it's REALLY addicting.

Looking at your pics, the traces are nice and parallel.

Sure you don't want to make a TEA laser instead? All joking aside, that totally stinks, and I hope you find an easy fix.

Shucks, I was enjoying being tired and beat up after spelunking under my house all day (air conditioner). Now I want to go down in the shop and get back to work!

...

Sun Jun 10 2007, 02:19AM

Registered Member #56 Joined: Thu Feb 09 2006, 05:02AM
Location: Southern Califorina, USA
Posts: 2445

hehe, you know that you need to rethink your layout when you get UV laser beams shooting out of the bridge

Could you just spray the board with a little polyurethane or something?

Marko

Sun Jun 10 2007, 09:39AM

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

Your comment reminded me that disconnecting my o'scope probes made my heatsinks run cold instead of warm...

Part scavenger scope probes are fine as long as you are using isolation transformer.

Your heatsinks shouldn't be getting hot at those power levels at all.

One thing I noticed, you aren't using silicone pads on IGBT's at all?
This would explain even more why your GDT blew.

And guys, please, go to my thread for suggestions, just not to jack' scavenger's.

Marko

previous 1 2 3 4 next

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