i'm on a roll... the kgsshv

[kevin gilmore]

ah well-played. 52mA looks good to me as well, but it's tracking this value back to RMS current before the diode that might be tricky to nail down exactly.

And the problem is that the current waveform before the diodes is nothing like a sine wave. So the RMS value which is what

all transformer manufacturers want you to specify does not exist. Why these companies cannot understand full wave diode

bridge into capacitor of specific value, and that into a resistive load of specific value i don't know. And even if you tell them

the number of turns on the primary, and wire size, and the number of turns on the secondary and wire size, they insist

on arguing with you.

[wink]

"And even if you tell them

the number of turns on the primary, and wire size, and the number of turns on the secondary and wire size, they insist

on arguing with you. "

That's because they're the ""experts""....

[rille]

How do I build the Power Supply v8 as simple version (only zener diodes)? I guess there is a connection between the Cathode of D5 and the Gates of Q1/Q2 missing. In version 4 was there a 250kOhm resistor (R23). Should I add it also in v8?

[rille]

Nobody any idea?

[livewire]

Ididnt answer earlier because I really don't know.

KG or maybe Kerry would know.

The version 4 PSU was only on paper, never built as far as I know.

I really don't think that the 250K resistor would be needed on the version 8 psu.

As for modifying the version 8 PSU, here again you may want to PM KG to discuss this.

FWIW, my version two PSU on the version 3 PCB is working just fine. It doesnt get much simpler than this.

[Kerry]

I should have the boards in the next few of days and I'll take a look.

[kevin gilmore]

i'm running way behind for various reasons.

To do the simple zener version, make the zener string a total of 450 volts. (3 x 150v zeners)

Do not populate the following (this is for the top supply, the bottom supply is the same)

Q5,Q6,Q3,Q4,Q7, lt1021

R10,R4,R7,R6,R5,R29

C16

make R25 10k to 100k. (need to test this for lowest noise)

put a jumper wire between the emitter and collector of the empty Q7

Which you can also do between the 2 ends of the missing 240k resistors

like this

http://gilmore.chem.northwestern.edu/pszeneronly.jpg

next board run i'll add a 2 pin jumper.

Edited September 11, 2011 by kevin gilmore

[rille]

Thanks, this answers my question

[rille]

The printing on the power supply board for BR1 is wrong. The "+" is at the wrong side. I guess it's a board from the current group buy (Production Rev 0.6). So be carefull if you have one of these boards.

[spritzer]

Yup, I just checked and the bridge for the +/-15V supply is indeed reversed. Just flip it over and you will be fine (- where + is marked).

[kevin gilmore]

actually the + mark is the mark for the capacitor

since once you stuff it, the + mark under the cap is

now hidden.

the plus mark on the diode bridge is towards the edge

of the board, where the angle mark is.

I will move the + mark for the cap. and add a + mark

for the diode bridge.

[spritzer]

Ahh ok, clearly not idiot proof enough for me...

[kevin gilmore]

new version of board should be idiot proof for next time.

[spritzer]

I'll test it then just to me sure...

[rille]

The power supply is working now. I used the BUZ80 instead of the Ixys 10M90S with some modifications:

But the amp makes problems. The problem seems to be also the Ixys parts. I measured ~9V across R12/R19 only at one transistor. I guess this is ok? The other 3 seems to be defective (~0V across the R12/R19). I guess nearly all the Ixys parts from Littlediode were defective. Or they are fake parts. I will complain about the parts.

[kevin gilmore]

I can't even tell you how pissed i'm getting at all the fake parts.

And i've been recently bit by $200 in fake parts myself. None of

the distributors stand by them either.

Dalbani seems to be the worst of the bunch, complete thiefs.

[spritzer]

Fake parts piss me off no end...

Here is my first HV board ready for testing. I have some issues that need to be sorted but at least nothing blew up. I have the same problem as Kevin, I never follow the BOM...

[eggil]

All the 2SA1486's I tested were good, with a VBR in the 700 range.

The 2SA970BL's however have numbers that jump all over. The may test initially at 40-60, then go up to 170-190 for a fraction of a second and then go to 1. Not sure why.

EDIT: my bad, if I go to 1000V they (2SA970BL) test fine at 200-210. I guess testing them at 200V is not enough....

I got both from bdent. Everything I have tested from them has been good so far.

[ujamerstand]

bdent seems to have their shit together. Here's a comparison of the C3675 from bdent to utsource:

top bdent, bottom utsource:

You can see that the utsource devices are not as cleanly made as the bdent devices. In particular, bdent devices have a circle on the front when utsource device is a dot, and the edges of the bdent devices are much redder than the utsource devices. Both have voltage breakdown around 1110V. If these are fakes these are quite advanced fakes. If someone would like to test them under a curve tracer I'd gladly drop a few in the mail to you for free. album here.

[Horio]

Some progress photos:

I'm building the 500V version (hence the missing voltage multiplier parts).

Anything special I need to know/do before I try powering these bad boys up tomorrow?

[livewire]

Yes. High voltage can be lethal.

But you already know that. Right?

Please be careful.

[blubliss]

I thought the onboard heatsink version needed to be the lower voltage, like 450?

Edit: NVM, why do have three PSs, are you building for other people?

Edited October 9, 2011 by blubliss

[Horio]

Yes. High voltage can be lethal.

But you already know that. Right?

Please be careful.

That's what people keep telling me... I will definitely be taking the appropriate precautions.

Edit: NVM, why do have three PSs, are you building for other people?

Yea, I'm building three amps.

[spritzer]

bdent seems to have their shit together. Here's a comparison of the C3675 from bdent to utsource:

top bdent, bottom utsource:

You can see that the utsource devices are not as cleanly made as the bdent devices. In particular, bdent devices have a circle on the front when utsource device is a dot, and the edges of the bdent devices are much redder than the utsource devices. Both have voltage breakdown around 1110V. If these are fakes these are quite advanced fakes. If someone would like to test them under a curve tracer I'd gladly drop a few in the mail to you for free. album here.

The bottom ones look like every C3675 I've ever use so they are real.

[Kerry]

Nice work on the PS's Horio!

I llike to duck under the table with a voltage meter tied to the outputs when I first turn a new build on

EDIT: Just looking at the photos again, I can't tell if you have insulators mounted behind the output fets.

Edited October 9, 2011 by Kerry