DRSSTC is popping rectifier bridges. / Tesla Coils / Forums

Forums

4hv.org :: Forums :: Tesla Coils

« Previous topic | Next topic »

DRSSTC is popping rectifier bridges.

Move Thread

LAN_403

Daniel Uhrenholt

Thu Jul 30 2009, 08:00PM

Registered Member #125 Joined: Fri Feb 10 2006, 01:52PM
Location: Aalborg, Denmark
Posts: 155

Hi All,

I have a small problem with the DRSSTC I have at the childrenâ€™s fare at the University, the bridge rectifier keeps blowing upâ€¦

The bus capacitor is 4600uF, and itâ€™s charged to 320VDc thru a resistor with a delay timer, so I donâ€™t use any variac here. I doubt that the inrush current kills the rectifier bridge, as I have used the same capacitors and rectifiers on other coils without any problems.

Can the field inside the cage kill a rectifier bridge??

Tomorrow morning I will change the blown 1000V 35A with a new one, and put some kind of box around it. If this doesnâ€™t work, I will order a big a** bridge rectifier and hope it will survive.

BTW, here is a movie of the coil

Cheers, Daniel Uhrenholt

Pylon

Thu Jul 30 2009, 10:13PM

Registered Member #2065 Joined: Sun Apr 05 2009, 06:32PM
Location:
Posts: 26

If the rectifier bridge failure is caused by high voltage induced in you circuit, the faraday cage around electronics is a good idea, maybe you should use TVS diode on 320VDC bus.

Daniel Uhrenholt

Fri Jul 31 2009, 01:39PM

Registered Member #125 Joined: Fri Feb 10 2006, 01:52PM
Location: Aalborg, Denmark
Posts: 155

Hi Pylon,

There is definitely something going on in the isolation transformer and filter network outside the Faraday cage. I took a bunch of 420V TVS and had them mounted across the rectifier bridge, and on the bus capacitors, and they died suddenly as the relay switched to full mains.
So there must be induced a voltage spike somewhere in the filters, but this I have to measure Sunday.

I can build a bigger rectifier that can take 5kV 3A 200A pulsed, but as I have no idea how big the voltage spike is, it may easily destroy the filter caps over time!

So I will probably try to tame the spike with a big wire wound resistor, which can burn off some power in the 5 seconds the coil runs every 15 minutes.

The filters is some rather big ones that are made for HF cages, they were used at the University in a laboratory where they develop antennas for mobile phones. Maybe this was a bad choice, I could easily have used some smaller ones, but these were recommended to me by the guys at the HF labâ€¦

Cheers, Daniel

GeordieBoy

Fri Jul 31 2009, 05:12PM

Registered Member #1232 Joined: Wed Jan 16 2008, 10:53PM
Location: Doon tha Toon!
Posts: 881

Make sure that the bridge rectifier is rated to allow for the lousy power factor of a passive line-connected rectifier. What I mean by this is that the mains line current will not be sinusoidal when the TC runs. It will in fact consist of a series of brief but incredibly intense current spikes just before the peaks of the mains supply voltage. All of the power must be transferred in these brief instants when the line voltage goes higher than the DC bus voltage. This means that the current spikes can be very large in magnitude. On top of this, there can be a further increase in the current if your DRSSTC controller fires the bridge at that moment!

Also bare in mind that filter networks typically consist of L's and C's and are well known to ring if not chosen correctly for the application. A badly chosen line filter can be worse than having no line filter at all. A popular misconception is that choosing the biggest RFI filter will give the best results. In fact it is best to choose a line filter for the anticipated current draw, not an over-rated one!

Finally if it is actually voltage spikes that are killing the rectifier bridge, you could try connecting 10nF 1000V ceramic capacitors across all of the rectifier diodes. If the cause of the problems is short term RF hash this will likely solve the problem. If the problem is more prolonged over-voltage, the capacitors will probably worsen the problem by providing local stored charge to dump into the diodes when they avalanche!

-Richie,

Steve Conner

Sat Aug 01 2009, 09:52AM

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

GeordieBoy wrote ...

Also bare in mind that filter networks typically consist of L's and C's and are well known to ring if not chosen correctly for the application. A badly chosen line filter can be worse than having no line filter at all. A popular misconception is that choosing the biggest RFI filter will give the best results. In fact it is best to choose a line filter for the anticipated current draw, not an over-rated one!

I thought this too, but wasn't sure if it made sense. Those big filters have an awful lot of inductance and so on. Maybe try without the filter and see if the rectifier lasts longer...

My Mjolnir DRSSTC uses just an ordinary 600V, 35A bridge. It also has an inductor in series with the line, though, to keep the SCRs in the "solid-state variac" happy, and this would improve the power factor a little. It's just a stack of iron powder toroids with as much wire as I could fit on there, of a gauge suitable to carry the expected line current.

It also has a small 0.01uF capacitor from DC bus negative to mains ground. This bypasses RF hash from primary strikes, etc, that would otherwise try to fight its way back into the mains through the rectifier bridge and GDTs.

If using one of these super EMI filters, you should probably bond its casing to the Faraday cage. Then the whole inside of the cage is mains ground, and also RF ground: the two grounds are the same at the point where the filter casing meets the cage.

Daniel Uhrenholt

Mon Aug 03 2009, 10:17AM

Registered Member #125 Joined: Fri Feb 10 2006, 01:52PM
Location: Aalborg, Denmark
Posts: 155

Richie,

Thanks for you input, itâ€™s much appreciated as Iâ€™m rather stressed out about this problem.

You got a point about rectifier, so I have just ordered a more robust bridge from RS. And I will try and put some caps over the diodes this time, to see if it helps.

Steve McConner wrote ...

My Mjolnir DRSSTC uses just an ordinary 600V, 35A bridge. It also has an inductor in series with the line, though, to keep the SCRs in the "solid-state variac" happy, and this would improve the power factor a little. It's just a stack of iron powder toroids with as much wire as I could fit on there, of a gauge suitable to carry the expected line current.

This I also have to take a look at, I have some rather big iron powder toroids, which can be used for this application. I also have a big inductor that Finn gave me years ago, that was in service when I was messing around with MOT`s

Steve McConner wrote ...

If using one of these super EMI filters, you should probably bond its casing to the Faraday cage. Then the whole inside of the cage is mains ground, and also RF ground: the two grounds are the same at the point where the filter casing meets the cage.

I use two of these â€œsuperâ€ EMI filters on the cage, and they are mounted as you say. But they are mounted underneath the floor, and I canâ€™t bypass them without I have to take the whole cage apartâ€¦

But as the coil and cage has passed some EMI tests without any comments, I donâ€™t want to bypass the filters, because I will then have an antenna out of the cageâ€¦

Thanks for the help guys,

Daniel Uhrenholt

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