official GDIY51PSU Help & Support Thread

[Harpo]

Different LED colour have different forward voltage and/or current requirement. Same type comes with +/-tolerances. Generated heat of insufficient power rated current limiting resistor might rise its resistance value and thereby dropping brightness. Its always something.

 

[New Soul Rebel]

;D

or Cemal, like a complete twat I could have just soldered a 4k7 resistor in there, when it should have been a 2k2!

god, today has been a day of numptiness.

cheers guys

 

[suttlet1]

I have a  51X backplane with fully soldered  36 pin edge connectors. I will sell very reasonably. Change of plans for myself and I've gone with the 30 pin VPR version.
Email me at:  [email protected]

 

[applebeach]

Hello everyone,
first of all i want to thank you for this great place.
I´m lurking around for some time now and finally found some time to jump into the water.

I´ve nearly finished my 51x PSU build but am wondering what (if) i´m doing wrong here.
I´m using the PSU rev 02 in an unused 2U rack-case i had lying around.
I connected everything like shown in chunger´s awesome thread (thank you sir !).

Before firing the PSU i checked for shorts between chassis and the metal of the LM350´s.
I´m getting shorts when testing the LM350´s of the negative rails. I then checked the same
against the PCB-GND (disconnected from STAR-GROUND / CHASSIS) and got the same results.
This only affects the negative rails, and there is no short between the LM350´s and the heatsink.

Then i thought it would be safe enough without anything hooked up to the PCB to switch that
thing on very carefully and doing some measurements.
From the xformer i read (V AC):
51,50    (48V connection)
25,75 and 25,73 (24V connection)
17,15 and 17,17 (16V connection)

I measured the DC Voltages (between GND and one of the holes beneath the trim pots):
+49,12
+24,37  /  -24,43
+16,60  /  -16,65
So that looks ok for me (even though i don´t know yet what those readings should be
without connecting any load).

As far as I understand i MUST NOT connect the metal of the LM350´s with the chassis/heatsinks,
but that would happen if i connect PCB-GND to STAR-GROUND again, right ? Or is the
STAR-GROUND in chunger´s drawing supposed to be lift from the case I´m building this thing in?

Please can anybody point me to what I´m doing or thinking wrong ?
Any help would be appreciated.


(please excuse me if this was covered somewhere else before)

 

[sahib]

Firstly the voltages you are reading are correct. Everything must be in order.

As for the short you mention, the power supply uses only positive type of regulators, therefore  both the positive and negative rails are produced from two positive regulators.  In other words what you have is two positive regulators connected in series.

As a result the negative end of one regulator is connected the the positive end of the other  to form the 0V / Common.

Therefore if you measure the negative end of the positive rail against the 0V on the PCB it will show continuity as usual. But on the negative rail the positive end will show continuity with the 0V point on the PCB.

The 0V/Common on the PCB is connected to the Earth/Safety Ground that comes from the mains on the chassis star point.  As the heatsinks also have contact with the chassis,  you will read continuity between the positive pin of the negative rail regulator and the chassis/heatsink.

 

[applebeach]

Thank you sahib for your explanation.
Now, after having a closer look at the schematics, at the negative rails i see
Vout of the regulators connected to GND.
So with the continuity testing of regulators/heatsink it's all about not shorting
Vin and Vout of the regulator eg. via the chassis, right?
Lesson learned.

 

[sahib]

True.

 

[bruce0]

applebeach...

One more note... just for safety..

There is NEVER a ground lift on a mains power supply *(as far as I know)

(Forgive me if this is obvious, but we amateurs should be careful about such things)

 

[applebeach]

Thank you bruce,
I never really thought about lifting main's ground, although I was
asking about it in my first post.
But nevertheless I'm glad you reminded me, as with high voltages
there is no room for trying stupid things ...
As a sideeffect I'm gradually moving from noob to amateur 

 

[saxmonster]

Does anyone know what the digital key or mouser part number is for the black bezel that keeps the green leds in place on the original 51x power supply with the gold front and 51x logo on it? What is the hole size or led size that was used? What is the thickness of the front plate? I moved to a new place and a few leds got pushed in and now the black bezel / holder is broken. Actually I think this was Jeff's prototype case.
-Scott

 

[saxmonster]

So I measured the LEDs and they are 5mm I just don't know about the thickness of the front panel. These seem like the will work well. They look like they can fit different thickness of panels too. Maybe Jeff or Cemal can remember what they used for the led holders.

https://www.amazon.com/SQXBK-Plasti...5&sprefix=5mm+led+holder,aps,180&sr=8-13&th=1

****. These do not work, Errrrrr

 

[saxmonster]

So here are the parts I ordered from mouser that are just like the original ones. Everything is back up and running and looks great.

These are the leds and holders.


https://www.mouser.com/ProductDetail/Kingbright/WP483GDT?qs=rAVhgu1ixJNlKQJ6pk%2BFvw==



https://www.mouser.com/ProductDetail/Kingbright/RTF5010?qs=y3UkhjaxUTagMl45k1/VQQ==

 

[bobbriessinck]

Hello there!

I already started another thread on this issue since I couldn't find this one, but @fallout was so kind to already help some over there. I'll copy paste the initial post with udpates below:


up until recently both my 51X racks & PSU worked just fine, but ever since I took one to another studio with me they've been acting weird, and now they're not working anymore. If I power up the PSU without any rack attached all the voltages measure fine (a bit above the voltages they should be at but I imagine they drop a V with load?), but once I attach a rack and power it up all the LEDs go out except for + and - 24V.
I've measured the voltages with 1 rack attached, results below:

+48V: 121mV
-16V: -304mV
+16V: 520mV
+24V: 24,96V
-24V: -24,87V
+48V: 123mV

The relays of the 500 unit click when I power the PSU on, but the LEDs don’t work, the VU of the LA500A (JLM) lights up but others don't show any sign of life besides a possible relay click.
Any idea what could be happening here? I'd like to check if there's a simple solution for this before I take the racks apart. It's just odd that they both do it with both power cables, but only 1 traveled with me outside of the studio.


I have to mention that in the other studio some of the 500s didn't work eventhough I slided them in with highest caution en checked if they were in the correct position. We were under big high pressure working against the clock and I didn't take the time to remove them from the chassis, I should've done that immediately. Then again, only one of the racks left the studio and they both experience the power drop, so my guess is that the problem lies within the PSU? At least I hope so. In coming days I'll check the 500s in a regular Fredenstein Bento to see if they're alright.

-UPDATE-
@fallout asked to check if removing and putting the fuses back in one by one would show any result and as expected, the 24V LEDs lighted up when the lower 2 fuses went back in, the 16V & 48V stayed off when the 2 upper fuses went back in. If I remove the 48V fuse only positive 24V LED goes on. I didn't assemble this PSU and I don't know enough about the whole thing to be able to draw conclusions here.
Fallout also asked me to check if any of the voltage regulators were significantly hotter than others, and indeed the 2 upper ones, above the 16V fuses so I'm guessing those are the 16V regulators (?), are significantly hotter than the lower 2.
Should I conclude that these 2 died and I should replace them? Or can the problem lie somewhere else still?
-END UPDATE-


Thanks for checking this out!
Bob

 

[fallout]

If you remove only the 16V fuses, do the +/- 24V and 48V LEDs remain lit? I'm not sure what the relation between the 48V and +/- 16V would be.

Are the regulators secured to the heatsink properly?

I'd still try removing some of your 500 modules and see what happens. Maybe a thermal issue under load and the vregs are shutting down?

 

[bobbriessinck]

Okay so I removed all units (6 regular 1 space EQs and 2 2space EQs) and all the LEDs were lit. I then put them back in, one by one while waiting long enough after powering down the PSU, and all the LEDs stayed lit. Now both 11u racks are plugged in again, rack 1 full and rack 2 with 10 units, and all the LEDs are still lit. I'd think that there was one not plugged in properly causing something to happen in the PSU, but I didn't touch any of the 500s in the full 11u rack, and it did have the same problem.
I kept the PSU on for a good while now and they're still all lit. I ran pink noise through all 19 units individually and everything works/sounds fine...
I guess I shouldn't be concerned now but should I be? I just don't like not knowing what the problem was...

Anyhow thanks for helping!