Yes 1U with slim line 01/350 from Slim Line 01/350 10mm BLACK front panel - 3mm aluminium covers
Thanks j wilhelm, that is very good , i have the datasheets for the XLR Connectors, the only thing i am not sure is the drilling for the RCA connector's screw, i will use a countersunk allen screw. Looks like this is the same that the XLR connectors(0.126)(3.2 mm)..
I recently assembled this PSU using the standard parts as called out in the BOM. My board is v2.2. I placed jumpers as indicated on the PCB for 120v use. I used the standard AMGIS L01-6354 transformer. I adjusted the 5K trim pots on both rails for +-15VDC. All things checked out on my bench, with no magic smoke.
Here's my issue: After a few minutes in my application, I noticed an extra amount of bass that I had not heard before. Almost a boomy sound. I decided to touch one of the rectifier diodes (D4). I almost burned myself it was so hot. I took my infrared thermometer and measured the temp. D4 was reading ~155 degrees fahrenheit! I measured the other 3 diodes.
The readings were;
D3 = ~142 degrees
D5 = ~100 degrees
D6 = ~102 degrees
I also felt the regulator heatsinks, and both were cool to the touch. I thought these temps were extremely high and that something must be amiss - either with a bad part, or my assembly.
My question to you is; Are the temps uncharacteristically high? Or, are these temp readings in the normal range? Are you aware of any issue with this PSU, and/or other reports of extreme temps on the rectifier diodes?
I've double-checked my parts layout and everything looks correct. Given the higher temps on D3 and D4, is there a particular part that may be suspect, or causing this if indeed it is abnormal?
There have been a few builders that have had the problem that you describe. One of them wrote about it earlier in this thread. In both cases they replaced the diodes and all was well. Could the parts that you have be from a suspect vendor?
I recently assembled this PSU using the standard parts as called out in the BOM. My board is v2.2. I placed jumpers as indicated on the PCB for 120v use. I used the standard AMGIS L01-6354 transformer. I adjusted the 5K trim pots on both rails for +-15VDC. All things checked out on my bench, with no magic smoke.
Here's my issue: After a few minutes in my application, I noticed an extra amount of bass that I had not heard before. Almost a boomy sound. I decided to touch one of the rectifier diodes (D4). I almost burned myself it was so hot. I took my infrared thermometer and measured the temp. D4 was reading ~155 degrees fahrenheit! I measured the other 3 diodes.
The readings were;
D3 = ~142 degrees
D5 = ~100 degrees
D6 = ~102 degrees
I also felt the regulator heatsinks, and both were cool to the touch. I thought these temps were extremely high and that something must be amiss - either with a bad part, or my assembly.
My question to you is; Are the temps uncharacteristically high? Or, are these temp readings in the normal range? Are you aware of any issue with this PSU, and/or other reports of extreme temps on the rectifier diodes?
I've double-checked my parts layout and everything looks correct. Given the higher temps on D3 and D4, is there a particular part that may be suspect, or causing this if indeed it is abnormal?
Same thing here....
I recently assembled this PSU also using the standard parts as called out in the BOM. Same board v2.2. I have the jumper as indicated on the PCB for 230v use. I used AMGIS L01-6364 (not L01-6354) transformer. The 5K trim pots are adjusted for +-17VDC.
After a few minutes with no load the MBR1045-E3/45 diodes are very hot, more then 60 ºC I Got those from Digikey Month ago.
Of course like usual I tried first with the mains bulb tester and it did not lit.
Reedcat and Heijnsva had the same issue last January, see post #985 to #994 and they swapped the rectifiers with mbr1060's. I don't thing that is right thing to do..that should work with mbr1045. See AndrewT Post #991
i will try the board without the 47nf caps...
Carl, your thought please
Last edited:
Much better to avoid schottkys altogether for higher voltages, the reverse leakage is subject to thermal runaway. Soft recovery fast rectifiers are the natural choice, but if you go with schottky 100V+ parts are a much safer choice - the closer you go to the max reverse voltage the higher the reverse leakage, and its an _exponential_ function. Its also exponential with temperature so schottkys would typically require a heatsink for good thermal control.
Where schottkys are used in anger, like a computer SMPS they are on fan cooled heatsinks, because its necessary.
The other issue with that supply is the schottkys chosen are stupidly over spec'd, and leakage current scales with working current, so the problem is much worse than it would be with more reasonable devices. A preamp power supply can happily use 0.5A diodes, which would, all else being equal, have 20 times less leakage. The MBR1045's have a 150A pulse rating!
Where schottkys are used in anger, like a computer SMPS they are on fan cooled heatsinks, because its necessary.
The other issue with that supply is the schottkys chosen are stupidly over spec'd, and leakage current scales with working current, so the problem is much worse than it would be with more reasonable devices. A preamp power supply can happily use 0.5A diodes, which would, all else being equal, have 20 times less leakage. The MBR1045's have a 150A pulse rating!
Last edited:
Mark makes a valid point. There is no reason that I can think of not to use four 1N4004/ 1N4007 parts. You will have to stand them on end to fit the PCB. Put the band end of the 1N4004/ 1N4007 part in what is now pin #1 (ie: most left leg) of the current part.
Thank you for your contribution Mark.
Last edited:
Mark - Can you recommend a suitable rectifier diode for use with this supply as a better replacement for the MBR1045s - from either Mouser or Digikey? I've already ordered some MBR1060, but it sounds like a better spec'd part is necessary. I'm putting together another 'blue' board and would prefer to use a part I know will work better.
I replaced all four with another set of MBR1045s last night and got the same result. So it would seem there's an issue using that part in the PSU. Multiple bad diodes would be very rare indeed.
Its not a question of a better spec'd part, its choosing one that appropriate to the circuit its going to be used in, ie meets the requirements of the circuit.
Bog standard 100V bridge rectifier would work, soft recovery might be quieter,
these are cheap as you can get parts, get both. UF4002+ seems to be a fraction of the price of MBR1045's, as would any standard bridge rect.
Ok. I'd prefer to keep the TO-220 case style. Assume we want to meet the needs of the of the DS pre-amplifier, and the part could be a drop-in replacement for the MBR1045s - sans the heat issue. Suggestions other than the UF4002+ ?
Anything that satisfies the criteria, there are probably thousands of parts that will do that, use a search tool, select rectifier type, package TO220, select voltage, current, etc. Its just a rectifier diode.
I fitted some MUR820G diodes to DS series regulated PS (blue board) and all is well. Instead of temps of 140F, I'm getting more respectable temps of between 75-85F. Both regulators at around ~78F. And these temps are under load. I've adjusted the trim pots to get +-14volts ouput. Everything is cool to the touch. I'm happy and the sound of my DAC output stage is great!
Thanks Hicoco for suggesting the 820Gs.
Thanks Hicoco for suggesting the 820Gs.