official GDIY51PSU Help & Support Thread

[Denyle Guitars]

Once that is done I have to decide what to do for cooling.  I will probably drills holes above the heatsink. I don't think I will need them below the heat sink as the heat sink sits right on the bottom with 4 screws.  So the heat should transfer right to the bottom if there is any.

I would think it's good practice to have holes top & bottom for airflow, but it might not be necessary in this case.

I think the heat sink is way over kill anyways, hahahha.

The original design for that heatsink included another 4-40 hole adjacent to the bridge rectifiers so I could connect them via aluminum spacers to the sink.

 

[saxmonster]

Well I now have holes above and below the heatsink.  Air will flow just fine now.  Not the most even of lines but will work for now.  Wish I had a drill press or a cnc machine.  All is finished now.

 

[kante1603]

Well I now have holes above and below the heatsink.  Air will flow just fine now.  Not the most even of lines but will work for now.  Wish I had a drill press or a cnc machine.  All is finished now.

Pics please,

Udo

 

[saxmonster]

Not the straightest lines but it will do.  Same thing on the bottom under the heatsink.

It should draw fresh air up from bottom and exit hot air through the top if needed.

 

[sahib]

Good stuff Scott.

 

[kante1603]

Hi Scott,

Well done,welcome to the club

Udo.

 

[Govinda Doyle]

Hey there!
So i have built the kit, it has been alot of fun, i finally powered it up and it pretty much works! LOL
my 48v rail is perfect and i calibrated voltage, my negative rails all work, but both my positive 16v and 24v rails are not lighting up the leds, i have been having a good look but cant seem to figure out what i have done wrong on both positive voltage rails! thats the puzzling part! its on both positive!

1.the fuses are fine

2 the leds are wired correctly

3 the power tranny seems to be wired correctly

hint or perhaps a question... i have one of the 51x kits with gold front rack and turning red light knob for power on.. when the iec cable is plugged in even when its off its still lit up but alot less brighter than when in the on position is this normal??
my blue wire from jeffs power transformer join together on the mains voltage 120/240 selector tab and a blue wire goes to the on off switch in position 2 bottom..
im a little stumped.
For example the LM350's for +16v rail that is not working right measure from left to right or positive to right on the legs
+      1.9mv
pin 2 1.4mv
pin 3 23.3v
at the 511 connector rail i get 1.0mv

the working negative 16v rail measures
+      -1.23v
pin 2 -0.3mv
pin 3 -8.73v
at the 511 connector rail i get -16v

perhaps i have man handled the LM350 or fried them?
thanks in advance!
Govinda

 

[sahib]

Swap over the phase and neutral  connections on the mains switch (also at the output to keep the colour code correct). The light will be completely off when the switch is off.

On your + 16 and 24V rails. Although it may all seem o.k. check the components and the soldering carefully again. You may have connected the LM350s wrong way round.

 

[Govinda Doyle]

Thanks Sahib!!
I got it working but i have to investigate a little further... i only did two things i swapped the two green ac wires opposite on the 16v+ rail and i isolated the LM350's heatsinks off the case... im not sure which did what but i will try and figure it out and report back... im about to try your advice regarding the mains on off switch!
Thanks again
Govinda

 

[Govinda Doyle]

all working!
I just dont understand how reversing the positive and negative made the switch work!
totally stoked!
thanks mr sahib!

 

[sahib]

You are welcome.

 

[Tajmoben]

Hey everybody, I built my 51x rack a couple years back and I've been using it since w/ no apparent problems. This changed a couple of months back when I tried to add 2 more pre's to the setup.

My problem is on the +16v rail somewhere. The voltage on that rail starts to sag under any load, no matter how small. At 25mA it sags by .02v, at 50mA it sags by .04v. When the total load reaches 250mA the voltage starts a slow creep down until it reaches +7v and then it crashes to +2v and stays there until I lighten the load. At 350mA it crashes fast after about 5 seconds.

It appears that the +16v rail is the only one giving me trouble. I've been emailing back and forth w/ Jeff @ classic API and he suggested that I post here. We've ruled out the Toroidal Transformer, and so far I've replaced every component in the path. I've also checked all the connectors, cables, and the rack for shorts.....and I've found nothing.

I'm all out of ideas.......any suggestions?

Thanks,
Ben

 

[bruce0]

... changed a couple of months back when I tried to add 2 more pre's to the setup. ... the +16v rail ... starts to sag ...

Ben:

Forgive this pretty obvious stuff but:
Are you measuring the actual load on the PSU +16V rail and how are you doing that?
Are you using the rack as the load, or have you tried running the PS with a test load? 
Are you measuring all this with a rack full of modules?

For me the question would be is the problem in the rack or the PS.
Suggest you test the PSU on some sort of test load, and if it is ok, then power up the empty rack, test  are repeat adding one module at a time.

The backplane is very simple, so most likely if the PS is good, you have a problem module.

b

 

[sahib]

To me it sounds like a thermal problem.

1. Check, there are no dust/particles between the regulator/thermal pad and the heatsink.
2. Check, the regulator is screwed onto the heatsink firmly.
2. Check, the thermal pad is a good quality (unless you bought it from Jeff).

 

[Tajmoben]

Forgive this pretty obvious stuff but:
Are you measuring the actual load on the PSU +16V rail and how are you doing that?
Are you using the rack as the load, or have you tried running the PS with a test load? 
Are you measuring all this with a rack full of modules?

Hey Bruce, at the moment I've got my a multimeter wired inline on the +16v rail. I can set it to tell me the amount of current being pulled at any given time. I've tested all the modules individually and they all work fine, so long as the total load is less than 250mA. That's where things start going south. I've got a list of all the modules and their current draw if you'd like to see it.

Unfortunately, I haven't been able to try the PSU with a test load, I'm not sure I have the necessary tools to do so.

 

[Tajmoben]

To me it sounds like a thermal problem.

1. Check, there are no dust/particles between the regulator/thermal pad and the heatsink.
2. Check, the regulator is screwed onto the heatsink firmly.
2. Check, the thermal pad is a good quality (unless you bought it from Jeff).

Hey Sahib, I thought this might be a thermal problem too, so broke out the ir thermometer. Even under a load, nothing seemed to be getting too hot. I'll give it a look tomorrow and see if I notice anything that could be causing a thermal issue.

 

[[silent:arts]]

Just try a new LM350T ...

 

[bruce0]

at the moment I've got my a multimeter wired inline on the +16v rail. I can set it to tell me the amount of current being pulled at any given time.

I think that test is fine, and it sounds like you really do have a power supply problem.

And the regulator is not hot... another thought ...

If the rectifier (BR31) or the connections to that rectifier were messed up, that might explain this.  If you had a half wave rather than a full wave (if one side was not working) you could have a situation where the big capacitor would hold enough charge at low current to give you +16V but then the input to the regulator would NOT be inadequate for full current.  Then as the voltage dropped the current requirements of the opamps (as they locked up) in the modules could cause the the collapse from 7v to 2v (some math whiz could tell you whether the voltage would begin to fall at 250mA)..  I would put a scope on the output of the rectifier to see what you have there (as the voltage collapses it would start to look like a half wave rectifier), if you have a scope.  (by the way..Tell us what you have for equipment to test with).

It would be surprising if BR31 failed... those are big tough things.

As far as the replace the regulator suggestion.. sure... I think replacing the regulator couldn't hurt...  (IC31 the LM350T) they are cheap... but it seems to regulate at low current and collapse under load, so I was looking for a way to explain that.

Let us know.. I am curious

 

[Tajmoben]

I just spent a couple of hours messing w/ (testing) the psu.

Bruce, your idea seemed like a winner to me.....but the only real testing equipment that I have is a couple of multimeters. Based on the info I found online about testing rectifiers w/ a meter, br31 seems to be working fine. I'm also thinking I may have to eat my words about it not being a thermal problem.

I checked the temp on the regulators and they felt much hotter to the touch than my thermometer was reading, so I took the psu out of the case to get a better angle on them and the regulator on the +16v rail was a good bit hotter than the -16v regulator. The +16v regulator was @ 150-160 F when the voltage bottomed out, also, the -16v regulator was somewhere around 110 F. I also, noticed that the -16v rail started to creep down as the temp of the regulator went up over 100 F.

If this really is a thermal problem it wouldn't be much of a surprise as I've never felt like I had adequate heatsinks or ventilation in the psu case. I'd really like to get one of the new PSU floor boxes that Jeff has for sale, they look sweet, but he suggested I try to get this one fixed before buying a new one.

 

[bruce0]

While warm is not good, 160 F is not hot.. and it probably wouldnt shut down... however the LM350T has some sort of internal limiting on dissipation... maybe you do have a bad LM350, and it is shutting itself down early.  I agree with mr. Arts... Swap it out.

While you could get it better cooled, with a good Good contact, heat sink compound, and a good sink but ... I don't think it should shut down at that temp. 

I think you have a bad regulator (or a bad bridge rectifier).  Given that the regulator gets hotter than the good one... it is probably the culprit.

If you don't have a spare... you could always swap the +16 and -16... and see if the problem moves.

bb