DC spark gapper.. rectifier issues! / Tesla Coils / Forums

Forums

4hv.org :: Forums :: Tesla Coils

« Previous topic | Next topic »

DC spark gapper.. rectifier issues!

Move Thread

LAN_403

Kizmo

Sat May 26 2012, 11:26PM

Registered Member #599 Joined: Thu Mar 22 2007, 07:40PM
Location: Northern Finland, Rovaniemi
Posts: 624

Hello all.

I have medium size DC resonant charging TC which is eating rectifier diodes. Im currently using 3 phase power and 6 pulse rectifier. Each of the 6 diodes are made from two paralled strings of 50x 3A/1kV diodes and power resistors for current sharing. So far i have had 2 rectifier failures.

First one happened when i was using microwave oven diodes and why it failed was pretty obvious:

My only safety gap at TC end was not up to the task, it wasnt protecting against common mode overvoltage.

Then i went and bought 600 3A 1kV diodes and spent couple nights soldering them to strings of 50 diodes. For every diode two strings was put in parallel with power resistors to form "50kV 6A diode" I know that 50x1kV diodes wont make 50kV diode but my output voltage was only 10kVAC between phases.

Safety gap at the coil end was upgraded with center electrode that was connected directly to same ground rod with secondary base and strike rail.

Everything was good, i gave it huge amount of power after tuning (right before failure i was reading about 10kVDC and 1Amp charging current). I didnt see any ground strikes to the metal mesh that was covering my feed / rotary gap wires (which was also tied to RF ground rod). Safety gap fired every now and then which is pretty usual since it was set quite close. Right before failure i had massive air streamers, safety gap went couple times into machine gung mode (continuous firing for Â½ second maybe) and then variac breaker tripped.

After some fiddling around i found out that one of 6 diodes was dead short and its game over.. again.

So what the heck is eating my diodes this time? With 24kV 750mA microwave oven diode strings this thing worked for good Â½ hours of run time and one set of rotary electrodes but this larger rectifier failed in 30 seconds!

This coil is wired like this:

3x400V mains input -> 3 phase variac -> 3 phase 1:2 step down transformer -> 3 wire 3 phase emi filter (tied to house ground) -> 3x400V/3x20000V pole transformer (case tied to house ground) -> 6 pulse rectifier -> Dq diode -> 30H Charging reactor -> feed wires covered with RF grounded metal mesh -> Safety gap with RF ground center electrode -> tank circuit

Im little confused now about what the heck happened and more importantly how to stop it from happening again...

Steve Conner

Sun May 27 2012, 08:25AM

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

When the safety gap fires, there's nothing to limit the DC current. The rectifier could get overstressed. A DC resonant charging system can't exceed twice the DC bus voltage, so does your system really need safety gaps?

If you're worried about common-mode overvoltage, there may be another way to control that without the risk of shorting the DC bus. For instance, ground the negative side, and then put a RF bypass cap between negative and positive.

Another possibility is excessive RF voltages. Large power resistors in series with the rectifier's DC output terminals will help to damp these. The idea is similar to the Terry Filter.

Kizmo

Sun May 27 2012, 09:11AM

Registered Member #599 Joined: Thu Mar 22 2007, 07:40PM
Location: Northern Finland, Rovaniemi
Posts: 624

Thanks :)

Im not very big fan of ground referenced high voltage supplies. In drsstc world people dont ground their DC lines but they use capacitors between house ground and negative/positive. Could similar thing work here? Im also planning to wind high voltage common mode choke. My charging reactor already has about 800ohms of DC resistance. Before TC duty we did some jacobs ladders and other stuff with this transformer-rectifier contraption and it worked well, HV side current meter peaked over scale at 5Amps and nothing broke during that. With TC as load we never saw more than 1 amp average.

Safety gap is from the time when this coil was AC driven, maybe i could get rid of it.

Steve Conner

Sun May 27 2012, 12:22PM

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

Hi Kizmo

Yes, grounding the DC bus through capacitors should work much the same as grounding it directly. The case of the pole pig might be a suitable "house ground".

I think Terry Fritz showed that resistors work better than chokes. A resistor can damp oscillations by absorbing energy, but a choke can only change the frequency of the oscillations.

Kizmo

Sun May 27 2012, 03:08PM

Registered Member #599 Joined: Thu Mar 22 2007, 07:40PM
Location: Northern Finland, Rovaniemi
Posts: 624

Ok.

I replaced one faulty diode stack with ugly microwave oven diode contraption and the high voltage supply is up&running.

Now, these RF bypass capacitors. What kind of capacitance i have to aim for. I have couple 16kV 20nF strings from old mmc which i could pop in there with minor modifications. One between DC rails and one between DC- and transformer case. This coil in question runs at 70kHz

In terry filter these power resistors are 1k 100W resistors, at 1Amp 1k resistor would waste 1kW.

Can i go lower than 1k?

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