thyristor control circuit technical question

[motorhomer]

Hello All, I have a technical question about a circuit which has two thyristors which switch on/off to power two transformers which then drop down the voltage from 750v to 110vdc.

The part that fails over a short period of time are the thyristors which go to short circuit. Attached is the schematic but I cannot show the controlling circuitry to this part.

My main two questions are.

1. If they are shorting out then I assume switching the thryristors on permanently will not cause these to fail and I can assume there is no short to earth like a H bridge would do as this would cause an open circuit. With the Thyristors on permanently the transformers will be like an inductor in the circuit.

2. What would cause a thyristor to short across. Could this be the snubber circuit not working or the bank of capacitor not doing there job properly which are designed to smooth out any spikes, or could switching on/off to quickly cause the thyristor to overheat and fail.

I will assume over switching of the thyristors will increase the output frequency at 110v side which i assume will mean less magnetic flux cutting secondary coil and therefore either less voltage or no voltage on output.

The top right circuit is the 100Vdc rectified output and the bottom right is the 370Vdc output which goes on to be inverted to 240vac

Will update when I have more information on the circuit. What i can say is this is a flyback converter

[T3sl4co1l]

Yikes, where's the commutation?

Tim

[PartialDischarge]

At first glance I’d say the snubber is not working properly. Check if the thyristor is mounted properly with enough clamping force. Overvoltages are the achilles heel of thyristors and will short them easily, if its an old machine snubber caps do fail, so check or replace it, also check the parallel diodes

[motorhomer]

Thanks i did read over voltage could short thyristors, so i will get these parts tested. To add to the circuit D1 is there to stop any reverse spike on 750v side when switching on/off. So if this was shorted out then this might cause a reverse spike through the thyristor and short it, but this is a theory as the reverse diodes D2 and D3 should take this.

[NiHaoMike]

Have you considered upgrading to a modern design based around IGBTs? Thyristor DC/DC converters for less than 1kV input have practically disappeared due to the difficulty getting proper switching.

[motorhomer]

Would be nice to upgrade but this is a installation on a work place who will not spend the money and rather get it fixes each time. What i am trying to do is improve my knowledge on why the thyristor would have failed in the first place. I can see if part of the snubber circuit was defective this could cause over voltage on the thyristor causing a short circuit, but i am not sure what the affect would be of a higher or lower trigger pulse on the gate. Also if we have an over current on the secondary side would this cause a thyristor to short out or go open circuit.

I understand this might be all guess work but in end improves my knowledge

[duak]

If you look closely at the symbols for Th1 & Th2 you'll see an extra tick on the gate terminal.  I believe these are GTO (gate turn off) thyristors that are used in older inverters and motor drives.  These devices can be shut off and so don't need something else for commutation.

Sometimes thyristors are triggered on by a short, fast voltage spike, ie A-K dV/dt can induce enough current that, with leakage current, is enough to trigger the device.  Since the gate current isn't fast and high enough to snap the device on, it turns on slowly and may cause local hot spots that can damage it.  Thyristors can also be damaged by too much current, which can be caused by the transformer core saturating, the anode current quickly rising and then not being shut off in time to prevent damage.  It can also be caused by a secondary overload as the only thing limiting a current in a transformer is the series resistance of the windings.

At a race recently I bumped into a fellow that was in charge of the electrical system for one of our Rapd Transit lines.  One thing he said was that track voltage was about 750 VDC.  Would this inverter be for some sort of traction equipment?

[motorhomer]

Hi duak, yes this is very old equipment and track equipment so the input voltage can go up and down. The input part is protected from over voltage so this is not a problem.

The secondary current should be monitored by the control circuit but this might not be working correctly. I expect output overcurrent if not a factor.

As for core saturation causing GTO to fail i assume this could be caused by over triggering the gate to the GTO, this could be a fault with the GTO driver card which is something i am looking at changing. This is helpful as i was thinking over triggering will just cause the circuit to see transformers as an inductor.

[Circlotron]

Thirty years ago I used to work on UPSs ranging from 2.5 to 20kW that used force commutated SCR inverters running off a 135VDC buss and feeding a ferroresonant transformer. They were about as sophisticated as a steam locomotive, and equally as reliable. AC/AC efficiency was about 85% and year in, year out, except for one model that had occasional problems with a DC solenoid contactor overheating, the only thing that would ever need changing is the batteries. SCR inverters when done properly are as reliable as guts. As reliable as a good transformer. Old timey doesn't necessarily mean no good.

[sibeen]

Thirty years ago I used to work on UPSs ranging from 2.5 to 20kW that used force commutated SCR inverters running off a 135VDC buss and feeding a ferroresonant transformer. They were about as sophisticated as a steam locomotive, and equally as reliable. AC/AC efficiency was about 85% and year in, year out, except for one model that had occasional problems with a DC solenoid contactor overheating, the only thing that would ever need changing is the batteries. SCR inverters when done properly are as reliable as guts. As reliable as a good transformer. Old timey doesn't necessarily mean no good.

Not Thytec UPSs by any chance?

I'm about to remove from service a 300 kVA UPS that was manufactured in 1986. Utilisers a 12 pulse input Graetz bridge followed by a constant current thyristor inverter. The only reason that it's being removed is the efficiency of the unit and the fact that even though the manufacturer is still in business this line of UPSs was discontinued in the early 90s and parts are unavailable. The only thing that's ever been replaced on the system is a few fans and the batteries. It's never had an actual fault. They were reliable beasts, that's for sure.

[Circlotron]

Not Thytec UPSs by any chance?

Nah, Sola Basic. 13 Healey Rd Dandenong Sth. Was there from mid '88 until end '89.

[sibeen]

Not Thytec UPSs by any chance?

Nah, Sola Basic. 13 Healey Rd Dandenong Sth. Was there from mid '88 until end '89.

I remember the Sola units. Actually my memory has improved and I realised that the Thytec units had a mosfet inverter using really slow PWM to feed a ferro transformer.