DRSSTC killing DC power supplies
chris_inkubate, Thu Mar 05 2015, 06:40PMHi All,
I have a strange problem with my DRSSTC and would appreciate some advice please. When using the DRSSTC at low voltage into the bridge so only small arcs the whole system operates fine for ages. However once I power the bridge up to mains voltage and are getting the larger 1m plus arcs its only a few seconds before the 24V power supply running the UD2.5 board blows. Its now blown a 1A mains iron core transformer and a 2.5A 24V switched mode PSU. The IGBT's and bridge circuit remain un-damaged.
The coil is using two half bridge bricks, UD2.5 with both gate drive output stages in place one running each IGBT brick.
Any ideas would be most welcome as its running nice apart from this happening.
Many Thanks
Re: DRSSTC killing DC power supplies
chris_inkubate, Fri Mar 06 2015, 07:00PM
I have a little more information to add now. I've tried adding more shielding around the electronics and also a ferrite on the DC lead from the PSU. However another PSU has blown.
This PSU upon inspection though shows that the damage is to the input sections with bad burning around its PTC and varistor. This leads me to believe that high voltage is somehow getting into the mains input of the PSU. To make it more confusing the PSU gets its mains supply from a socket across the opposite side of the room from where the bridge is powered from and also has no earth connection.
Does anyone know of any circumstance where HV gets back into the mains via the Live / Neutral leads? I have TVS diodes from the bridge o/p to the -Ve supply of the bridge and a 100nF cap from the -Ve rail to ground. I'm wondering if the issues lies here?
Many Thanks
chris_inkubate, Fri Mar 06 2015, 07:00PM
I have a little more information to add now. I've tried adding more shielding around the electronics and also a ferrite on the DC lead from the PSU. However another PSU has blown.
This PSU upon inspection though shows that the damage is to the input sections with bad burning around its PTC and varistor. This leads me to believe that high voltage is somehow getting into the mains input of the PSU. To make it more confusing the PSU gets its mains supply from a socket across the opposite side of the room from where the bridge is powered from and also has no earth connection.
Does anyone know of any circumstance where HV gets back into the mains via the Live / Neutral leads? I have TVS diodes from the bridge o/p to the -Ve supply of the bridge and a 100nF cap from the -Ve rail to ground. I'm wondering if the issues lies here?
Many Thanks
Re: DRSSTC killing DC power supplies
GrantX, Sat Mar 07 2015, 07:10AM
Hmm, I've seen minor noise on the live and neutral due to cheap SMPS units with no/inadequate filtering. A DRSSTC isn't cheap, but it certainly isn't "low noise" either ;)
Perhaps place a bigger common-mode EMI filter (one with decoupling caps tying all three conductors together) between the 24V PSU and its mains connection? It might be worth trying before resorting to a more expensive option. Even just three capacitors across the mains conductors without the common-mode choke could help.
GrantX, Sat Mar 07 2015, 07:10AM
Hmm, I've seen minor noise on the live and neutral due to cheap SMPS units with no/inadequate filtering. A DRSSTC isn't cheap, but it certainly isn't "low noise" either ;)
Perhaps place a bigger common-mode EMI filter (one with decoupling caps tying all three conductors together) between the 24V PSU and its mains connection? It might be worth trying before resorting to a more expensive option. Even just three capacitors across the mains conductors without the common-mode choke could help.
Re: DRSSTC killing DC power supplies
Mads Barnkob, Sun Mar 08 2015, 11:11AM
I am thinking more along the lines that the galvanic isolation in your GDT is broken at higher bridge voltages and you inject full DC bus voltage backwards into the driver section.
Mads Barnkob, Sun Mar 08 2015, 11:11AM
I am thinking more along the lines that the galvanic isolation in your GDT is broken at higher bridge voltages and you inject full DC bus voltage backwards into the driver section.
Print this page