goldenreference low voltage power supply

[chinsettawong]

How can I get a pair, Kerry? 

[Kerry]

Once I’ve tested everything, we can run a group buy.

[Pars]

same size board, up to 35mm caps, had to move the pass transistors, needs checking silicon carbide rectifiers 100mm x 109mm
goldenreferencelargecap.zip

Anyone ever build this version? This is the one that I would be up for in a group buy...


Sent from my iPhone using Tapatalk

[looke]

12 hours ago, Pars said:

Anyone ever build this version? This is the one that I would be up for in a group buy...


Sent from my iPhone using Tapatalk

Me too. Do you have the group buy link?

[Kerry]

I've tested the positive rail, but not the negative.  I'll try to get to that shortly.

[Pars]

Thanks Kerry.

Kevin: would it be possible to design this board with dual rectifiers using the SiC diodes? Looks like lack of board space might preclude that?


Sent from my iPhone using Tapatalk

[kevin gilmore]

no room for anything else...

[Pars]

@looke: this wasn’t the GRLV version done for the group buy, mainly because we didn’t know if anyone had built it (though I wouldn’t think the board would have any problems).

 

 

Sent from my iPhone using Tapatalk

 

[ti5002000]

Can i use 2sa1930/2sc5171 instead of mjf15030/15031?

Thanks

[kevin gilmore]

Same pinout so they should work

[UFN]

FWIW I've use them as output transistors on Dynahis and CFAs without problems. Got significantly better N-P matching on them compared to the MJFs as well by the way.

[cspirou]

How's testing on the mini GRLV and GRHV? Any revisions needed?

[Pars]

I think this is right (and already had some run), but if someone wouldn't mind checking the schematic, particularly around the rectifiers, I'd appreciate it.

I had corresponded with ang728 back when he did a multi-output version of the GRLV, and he sent me his layout files. I made some changes to these around:

1. Added dual rectifier banks, supporting TO220 packages so I could use the SiC schottkys as Kevin did on his 35mm cap board
2. Supports up to 35mm main filter caps
3. Outputs will take either 5.08mm LS or 2.54mm LS connectors (i.e. Molex SPOX or KK)
4. Added positions for xfrmr secondary snubbers
5. Removed the caps across the rectifier diodes as they aren't a real snubber and don't do anything useful

[Pars]

These are the boards; I had seeed run 10 of them. Guess I'll have to build one up and see if it works...

The board is a bit larger at 116 x 107mm, so new build only. Mounting holes are inset 5mm (x,y).

 

[sbelyo]

looks good

[Pars]

Got one put together, but having a strange problem with it. The output voltages seem fine (and stable). However on startup, D1 flashes and then goes dim. D9 is brighter than either of the negative channel LEDs (D2 or D8). The V drop across the 150R are 1.51V for R1 and 1.62V for R2. Other voltages appear to be approximately the same. I'll have to remeasure the drops across the LEDs. The zeners appear to be similar (~8.9V). Any initial thoughts? Components appear to be in the proper spots.

Since I was in a rush to get these made (seeed 30% off), I also hadn't noticed some placement changes vs. the layouts from Kevin, particularly the LED placement is for D1 and D9 were reversed from Kevin's. I have made some changes in my diptrace files since I had these boards run to correct for that and label placements, etc.

EDIT: Found/fixed. Somehow I had put a BC559 in in place of an MPSA92 (Q3). Must have gotten mixed in my bag of MPSA92s somehow and I didn't read the labels when building it. Works just fine now (DOH!)

EDIT2: This one is for Diptrace users. When I went back into the schematic for this, after making some changes for the LED placement, and to add a pair of 220uf with 5mm LS in addition to the 7.5mm LS caps that were there (Silmics?), I found that the V- net had connected itself to the AGND net (no V- net present). Fortunately, this was not present in the board file I had run. How does this happen? And it wasn't easy/straight forward to fix.

Edited January 9, 2019 by Pars

[Ariamella]

Apart from the big caps (using 50v ones) and r8/r9 to 3kR, do I need to change any other parts assuming a 30v output? I assume the tants too, but would 50v for the 10uF and 35 for the 47uF ones be fine?

[Pars]

You should be good. I’ll have to look at the one in my DynaFET as it is +/-30V. I think 35V is as high as you can get for the 47uf. I’ll take a look in a sim and see what voltage these are on a 30V supply.


Sent from my iPhone using Tapatalk

[Ariamella]

Thanks my dude. Also, I'm aware that 28-30vac is a good range for the traffo for 30vdc but if I want to power a CFA2, what va is recommended?

[jose]

For headphones I used 180VA (24Vac x 2)

 

[Ariamella]

Holy crap, it takes that much power? 

[Pars]

50VA would be fine, though I usually use the 100VA Antek transformers. This is team overkill, after all

EDIT: I also ran the sim I have with 35VAC input, and used 4K/2K for R7/R8 for 30V out. The voltages across the 47uf caps was 9V for the first one and 10V for the second. The 10uf had ~16V on it. This is a sim, which may not match real life, but I would think it would be close to actual.

Edited January 12, 2019 by Pars

[Ariamella]

Alright, I'll think I'll use a 100VA traffo and keep it at 35V for 10uF and 25V for 47uF for the tants then. Thank you for the help!

[jose]

overkill? not overkill ... just cautious. ?

[Pars]

Here's a pic of the 35mm multioutput boards I had run. DImensions are 116 x 107mm, with the mounting holes on 106 x 97mm centers. On Semi schottky rectifiers, 30mm caps installed.

I'm working on shrinking this down to match the size and mounting pattern of Kevin's boards (110 x 93 with mtg. holes 100 x 83), and so far it looks good, though I'll probably have to lose the following:

1. power LED and resistors
2. the current board allows 220uf caps that are 18mm diam. with 7.5mm lead spacing in the upper corners. Not sure what cap was the target here, as a Silmic is 12.5mm d x 5mm ls (for 35V). This will shrink to Silmic size (though I don't use them).
3. This doesn't have the rail balance pot and stuff on it (to equalize the +/- rails). I'll take a look at that and see if it would fit.
4. I'll probably ditch the 1uf film caps following the main filter caps. Not sure they are really of any benefit.

These boards do support R-C snubbers. You could also do Mark Johnson's (diyaudio-Quasimodo thread) favored R-C-R snubber by soldering the 0.01uf film caps across the 2 AC inputs on the bottom. I may try to accommodate these on the board itself.