[sachu]

second board, the negative rail failed spectacularly on turn on. The regulator MOSFET burned out..took out R4 and R2 with it. Its possible I didn't have the load connected properly on the negative rail and was probably reading open.

Already removed all the offending parts, R2, R4,R6, Protection zener, mosfet. Probably will replace Q2 as well while am at it.
Unfortunately i don't have extra N channel mosfets with me..anyone got a spare? Need to scrounge up a 10R resistor as well..the others i have already replaced.

 

[Beefy]

Quote:

Originally Posted by sachu /img/forum/go_quote.gif second board, the negative rail failed spectacularly on turn on. The regulator MOSFET burned out..took out R4 and R2 with it. Its possible I didn't have the load connected properly on the negative rail and was probably reading open.

OUCH! But it don't think it will be the MOSFETs that burn up if the load isn't connected. It would probably be the shunt transistors; Q7 and 8 on the negative rail.

Something else might be wrong on that side......

 

[sachu]

Quote:

Originally Posted by Beefy /img/forum/go_quote.gif OUCH! But it don't think it will be the MOSFETs that burn up if the load isn't connected. It would probably be the shunt transistors; Q7 and 8 on the negative rail. Something else might be wrong on that side......

hmm..well could it have been the trim pot being turned a bit too high up on turn on? in any case i need to rebuild that section.

Wondering if it is safe to test the positive rail at least without the negative rail regulator mosfet in the loop.

 

[runeight]

Quote:

Originally Posted by sachu /img/forum/go_quote.gif hmm..well could it have been the trim pot being turned a bit too high up on turn on? in any case i need to rebuild that section. Wondering if it is safe to test the positive rail at least without the negative rail regulator mosfet in the loop.

Only if you want to risk more components.

Beefy is correct. If the NFET was in the right orientation and if its associated zener/bjt were also correct then without a load it will be the shunt devices that have been toasted. Certainly the MJEs, possibly the MPSAs.

 

[Beefy]

Well my power supply is up and checking out A-O-K. No smoke, and no need for CPR

Mains 122 VAC
Transformer unloaded 282 VAC
Transformer loaded 269 VAC

T1-T2 ~360 mV
T3-T4 ~ 410 mV

T1-GND 340 VDC
T3-GND 338 VDC

+ rail 308 VDC
- rail 307.5 VDC

Q3,4 heatsinks hot but can easily hold my fingers on them
Q5,6 and 7,8 heatsinks just warm
Load resistors HOT

My measured secondary voltage doesn't seem to make much sense relative to my post-cap voltage..... but I have confirmed it with two different multimeters and get the exact same result, so meh. The custom transformer seems to work perfectly though!

 

[Beefy]

Quote:

Originally Posted by runeight /img/forum/go_quote.gif You know, the value for R23/R24 was calculated for the nominal filter voltage of around 325-330V. If you have much higher than that then these resistors will get quite a bit hotter than I had expected.

IMHO, the default value of R23/R24 has to go up - let alone for those with higher post-cap voltage.

With 340V post-caps, my resistors are reading in excess of 120C - that is even hotter than the load resistors.. The resistors are not touching the PCB, but the PCB near the resistor legs is still getting much hotter than I am comfortable with.

 

[runeight]

I believe you are right. Would you have time/inclination to try 100k to see if that is a good value for LED brightness vs power dissipation?

 

[Beefy]

Quote:

Originally Posted by runeight /img/forum/go_quote.gif I believe you are right. Would you have time/inclination to try 100k to see if that is a good value for LED brightness vs power dissipation?

It will probably be a long time before I place another Mouser order, but I will definitely grab something when I do.

*

Just a bit more of an update with respect to my SumR transformer. The secondary AC voltage and post-cap voltage posted above do drop slightly as the transformer heats up; everything settles at ~335V post-cap after an hour or so. Thus I can definitely recommend a 2x240VAC @ 62.5mA (30VA) toroid as an option for the SS builds.

 

[fierce_freak]

I'm actually very suspicious of my DMM, so I've ordered a new one and will retest IDSS on the J271's. Scott, I'll get ahold of you after I've had a chance with the new meter and send the <10mA ones your way.

 

[BoilermakerFan]

Quote:

Originally Posted by fierce_freak /img/forum/go_quote.gif I'm actually very suspicious of my DMM, so I've ordered a new one and will retest IDSS on the J271's. Scott, I'll get ahold of you after I've had a chance with the new meter and send the <10mA ones your way.

Can't speak highly enough of my Amprobe:

The Amprobe 37XR-A. Made right next to Flukes in the same factory.
Amazon.com: Amprobe 37XR-A True RMS Digital Multimeter: Home Improvement

Resistance, Impedance, Continuity, Capacitance, and Inductance as well as true RMS voltage...

 

[runeight]

There are still five open slots for SS boards on this next run. Are there any additional new builders who want to have one?

 

[Keithpgdrb]

Quote:

Originally Posted by runeight /img/forum/go_quote.gif There are still five open slots for SS boards on this next run. Are there any additional new builders who want to have one?

what is the cost of the two boards together? I might just snatch up a sand as well for another build along the line.

 

[runeight]

The cost per set will probably be similar to before, about $25. However, I'd still like to limit the buys to one board set per builder. But, we'll see how this rolls out over the next week or so while we wait for the boards.

 

[M3NTAL]

runeight - please add me to the solid state board list!

Cheers
-M3NTAL-

 

[runeight]

Done.