Power Supply problems and power reg questions

Hi

I'm building a SSL9k PSU from the neeno's pdf for a Studer 169 Eq, using a 15V transformer and 78-7915 for the regulators. I'm not populating the +48V part of the pcb so it's quite simple. But in the first run one of the electrolytics was wrong oriented with his consequent explosion and bad smell. It was the one before the 7915. I replaced it with the correct orientation and now the positive rail is giving me +20.1V (I think it's a bit high, but maybe it's just because it has no load, anyway, it won't blow the 5534's in the EQ) but the negative rail is giving me 0.1V.

No one of the other caps looks wrong, the board is properly grounded (or I think so...) the earth mains is going to chassis, no shorts or something in the board, the fuse don't blow, well, the reg hasn't the best dissipation (I have no heatsinks at hand so I used aluminum paper to make some kind of heatsinks  ;D)

Could the 7915 be damaged by the bad position of the cap?

That makes me think how many times I replaced 78-79xx in my builds... Should I take a look into lm317 power supplies and start avoiding 78-79xx? Is the lm317 really better or I'm just having bad luck?

It's probably just bad luck. I always take extras care with power supplies. First get it working without the regulators so you don't do any damage with silly mistakes - and we all make them - and then add the regulators.

Cheers

Ian

the 7915 is famous for start up problems

I have built loads of power supplies with 7815/7915 and never had problems with them. Both very reliable peices of silicon. Like any other part, if you abuse it you will have problems.

do a search in the GSSL thread for 7915 ...
Latch-Up problems everywhere ...

Thanks for the replies

[silent:arts] said:

do a search in the GSSL thread for 7915 ...
Latch-Up problems everywhere ...

Click to expand...

Yes, I've built a GSSL a few months ago with lot of problems on the startup, but also had problems with other builds like SSL9k....

Back in the issues in my psu, maybe the bad polarization of the cap damaged the other cap and the 7915? I attach a picture showing wich was the bad cap

look for a min load current for operation on the data sheet,

you may need a dummy load on the minus 15,

they are short circuit proof so the exploding cap should not matter,

what is the brand of the regulator?

i had a whole batch of bad ones, not fun, pop pop pop, one at a time,

over voltage on tube heaters, wtf, over? 

my garndma just got run over by a 2 toed sloth,

Thanks For the reply CJ

I'll try the psu with a load, I've some boards with some ic's on it  wich doesn't really care me if explode (5532's & 34's, I've got some of them in stock)

The brand looks like ST Electronics (just for the logo on the reg)


BTW Merry Christmas, Happy Saturnalia or whatever you want to celebrate this days to everyone

Hint: I would test/measure the PSU with half-Watt, cca. 680-ohm (or similar) resistors as load.

Simply, if the voltage isn't right under such conditions, I would swap regulators.

ymmv

Well, I tried it with load, (one finished board of the EQ with the IC's plugged in) with the same results.

I replaced the regs and installed proper heatsinks, the same results, tried replacing the electrolitic caps and the rectifier with no luck.

Multimeter measures 15V AC in the transformer secondaries so it looks ok. But when I turn it on (with a 500mA fuse) i can hear some soft "humm" coming from it for a few seconds, and then stops humming but it keeps working,it doesn't seem to get too hot or smell bad (opposed to the 7915 heatsink which gets so hot).

I tried changing the 1000uF caps for 4700uF (googling 7815 psu i found some schematics and them showed  between 2200 and 4700uF) And the results are still the same, +20V and around -0.7V

Just built another psu p2p in a perf board, following the same scheme (pretty basic), with new comps and guess what? same voltages. Maybe a bad batch of 7915's in my local store?

F*ck it, I'm gonna eat something, have some break and think again about it later, or tomorrow.

dirtyhanfri said:

F*ck it, I'm gonna eat something, have some break and think again about it later, or tomorrow.

Click to expand...

This has worked for me in the past. Good luck!

I always found it irritating that 78xx and 79xx VR had different pin outs...  Power supplies are generally pretty straight forward to troubleshoot.  Only 3 pins to measure, ground, unreg and reg output. 

If the un-reg voltage and ground is OK, the regulator is not (unless it is in protection-shut down mode, but then it will usually be too hot to touch).

JR

It is too hot to touch and unreg V is around 20V. The transformer is 15v. Strange?

thermal shut down...

Look for short on output.

What is reg out voltage?

If positive may be latch up.


Not too many options... regs are simple.

JR

dirtyhanfri said:

It is too hot to touch and unreg V is around 20V. The transformer is 15v. Strange?

Click to expand...

That seems OK if it's after the bridge rectifier.

warpie said:

That seems OK if it's after the bridge rectifier.

Click to expand...

True, I've been reading basic theory about PSU's and found that mentioned

JohnRoberts said:

thermal shut down...

Look for short on output.

What is reg out voltage?

If positive may be latch up.


Not too many options... regs are simple.

JR

Click to expand...

I'm pretty sure there's no shorts on the output or in the whole circuit. reg Out Voltage is around -0.7. I know this is quite simple and there are not many options to look.

I'm not sure if this is something to care about, but measuring AC in the + output of the rectifier gaves me 40V, and 0 in the negative output...

Yesterday, when i went to my local store to buy 7915's and some caps, the guy told me "¿What are you doing with 7915's? you use many of them" And the I told him I think it's a bad batch or something, he was saying it is impossible, he told me that problems in the 7915's are usual and tried to sell me a pre assembled adjustable psu for 30euros, based on lm317 I think.

Anyway, time to refresh, and try a new PSU circuit. I'll go with lm317 & 337, I found a pcb design wich looks fine and with the regs far of the caps, (In the design I was building, the regs are pretty close to the Caps, heating them). The bad side is I have no boards to etch it right now and wouldn't like to go point 2 point.

dirtyhanfri said:

JohnRoberts said:

thermal shut down...

Look for short on output.

What is reg out voltage?

If positive may be latch up.


Not too many options... regs are simple.

JR

Click to expand...

I'm pretty sure there's no shorts on the output or in the whole circuit. reg Out Voltage is around -0.7. I know this is quite simple and there are not many options to look.

Click to expand...

OK, again this is pretty simple stuff... 

First things first... the 7815 putting out +20V is not right. Even a cheap VOM should not be that far off, and the 7815 will be within a few tenths of a volt from +15V.


A- what is voltage of unreg supply going into 7815?
B- what is voltage at 7815 ground pin (better be 0V).
C-confirm +20v on output that should be +15V. If confirmed and no external source of current coming "into" the +15V output,  that VR is faulty.

D- what is voltage of unreg  going into 7915?
E- What is ground pin of 7915 at, again should be 0V
F- confirm -0.7V at 7915 output.

What is connected to 7915 output? A backwards electrolytic cap or IC could load down the -15V rail.

If the 7915 is heating up, that suggests something si drawing current. Where is the current going. IF the 7915 is heating up all by itself with valid input voltage, it too is bad.

The odds of getting a bunch of bad VR sold as new good parts should be low.   

I'm not sure if this is something to care about, but measuring AC in the + output of the rectifier gaves me 40V, and 0 in the negative output...


I'm not sure if this is something to care about, but measuring AC in the + output of the rectifier gaves me 40V, and 0 in the negative output...

Click to expand...

? ? ? AC in the + output of the rectifier? The rectifier should be AC in, DC out.

40V AC sounds high for input, 40V AC on +DC output sounds impossible. I ASSume the ground lead of your VOM is connected to a proper PS ground?

Yesterday, when i went to my local store to buy 7915's and some caps, the guy told me "¿What are you doing with 7915's? you use many of them" And the I told him I think it's a bad batch or something, he was saying it is impossible, he told me that problems in the 7915's are usual and tried to sell me a pre assembled adjustable psu for 30euros, based on lm317 I think.

Click to expand...

I've used far more 7915s than 317s... the parts are not that twitchy...

Anyway, time to refresh, and try a new PSU circuit. I'll go with lm317 & 337, I found a pcb design wich looks fine and with the
Yesterday, when i went to my local store to buy 7915's and some caps, the guy told me "¿What are you doing with 7915's? you use many of them" And the I told him I think it's a bad batch or something, he was saying it is impossible, he told me that problems in the 7915's are usual and tried to sell me a pre assembled adjustable psu for 30euros, based on lm317 I think.

Click to expand...

Until we get the rectifier straight, the VREGs are not going to matter. But hopefully those are just bad measurements and not reality.

Anyway, time to refresh, and try a new PSU circuit. I'll go with lm317 & 337, I found a pcb design wich looks fine and with the regs far of the caps, (In the design I was building, the regs are pretty close to the Caps, heating them). The bad side is I have no boards to etch it right now and wouldn't like to go point 2 point.

Click to expand...

The 317 may be a better reg, but don't ever let a simple circuit get the best of you...  circuitry is your slave, to do your bidding.

JR

JohnRoberts said:

OK, again this is pretty simple stuff... 

First things first... the 7815 putting out +20V is not right. Even a cheap VOM should not be that far off, and the 7815 will be within a few tenths of a volt from +15V.


A- what is voltage of unreg supply going into 7815?
B- what is voltage at 7815 ground pin (better be 0V).
C-confirm +20v on output that should be +15V. If confirmed and no external source of current coming "into" the +15V output,  that VR is faulty.

Click to expand...

A.- +18V DC
B.- 0V DC
C.- +20.1V DC either with one 5534 as load (I know it's working, just tried in another circuit) or with a 680r 1/2 Watt resistor

JohnRoberts said:

D- what is voltage of unreg  going into 7915?
E- What is ground pin of 7915 at, again should be 0V
F- confirm -0.7V at 7915 output.

Click to expand...

D.- 18V DC
E.- 0 V DC
F.- Sadly I confirm it

JohnRoberts said:

What is connected to 7915 output? A backwards electrolytic cap or IC could load down the -15V rail.

If the 7915 is heating up, that suggests something si drawing current. Where is the current going. IF the 7915 is heating up all by itself with valid input voltage, it too is bad.

The odds of getting a bunch of bad VR sold as new good parts should be low.   

I tried connecting the EQ Circuit (just with an IC as a load), and also with just an 680r 1/2 W resistor as suggested above. Nothing looks wrong on the circuit, I checked it. Other times, when building psu's I've been able to read voltages without load, I've also built a pair of this psu a year ago and didn't remember so faulty. But i can remember changing one, maybe two regs.

JohnRoberts said:

? ? ? AC in the + output of the rectifier? The rectifier should be AC in, DC out.

40V AC sounds high for input, 40V AC on +DC output sounds impossible. I ASSume the ground lead of your VOM is connected to a proper PS ground?

Yes I know it looks weird (it sounds weird to me as a noobie, if it surprises you, some of the big players here, i get more confused now) But I think sometimes my DMM acts strange when working with AC, sometimes the display is showing 166.6 until I push "Hold" button a pair of times and it goes back to zero (when not measuring anything) Not a expensive one but not the cheapest one, it looks fine.

About the ground, maybe I'm making some mistake here... I got the board assembled to the bottom plate of a rack enclosure, making electrical connection with ground of the board in the spacers . And also, the mains earth connected to the plate. I tried measuring with COM of the DMM in the 0V point of the board, and in the bottom plate. Also tried to measure with the board not connected to the chassis, same results.

JohnRoberts said:

I've used far more 7915s than 317s... the parts are not that twitchy...

Not trying to blame any part, maybe I'm just having bad luck or misunderstanding something

JohnRoberts said:

Until we get the rectifier straight, the VREGs are not going to matter. But hopefully those are just bad measurements and not reality.

The 317 may be a better reg, but don't ever let a simple circuit get the best of you...  circuitry is your slave, to do your bidding.

JR

Yes I know, it looks like I'm doing something wrong but I find it super simple so I don't know...

I'm not giving up, I will make it work, for sure. But i don't like to get stuck on things, I have patience, but when I feel I'm walking in circles I like to get fresh air, I will comeback to this psu in a few days, reassemble it and surely it will be working. A spare psu is something nice to have around.

OTOH I'm thinking in build the lm317 power supply and let it as bench psu, check the EQ, and then go back to the faulty psu.

Thanks for your time

Click to expand...

Click to expand...

Click to expand...

Click to expand...

Click to expand...

Click to expand...

Click to expand...

Click to expand...

dirtyhanfri said:

JohnRoberts said:

OK, again this is pretty simple stuff... 

First things first... the 7815 putting out +20V is not right. Even a cheap VOM should not be that far off, and the 7815 will be within a few tenths of a volt from +15V.


A- what is voltage of unreg supply going into 7815?
B- what is voltage at 7815 ground pin (better be 0V).
C-confirm +20v on output that should be +15V. If confirmed and no external source of current coming "into" the +15V output,  that VR is faulty.

Click to expand...

A.- +18V DC
B.- 0V DC
C.- +20.1V DC either with one 5534 as load (I know it's working, just tried in another circuit) or with a 680r 1/2 Watt resistor

Click to expand...

Stop press...  For a 7815 to output more voltage than it's input is not reasonable. Either one of two things is going on,
A) Something externally is dumping current/voltage into that PS output.
B) Your measurement is not reflecting accurately what is going on.  +18V on the unregulated input is marginally low. IIRC the basic 7815 had something like a 3V drop out so 18V is about the minimum to support 15V output, but this does not remotely explain an output higher than the input...  You may have stumbled upon free energy, you can use to make a perpetual motion machine.  8) 8)

It has been many moons since I looked at the circuitry inside a 7815 but it is safe to ASSume that the output there should not be  5V higher than the nominal 15V rated voltage.

JohnRoberts said:

D- what is voltage of unreg  going into 7915?
E- What is ground pin of 7915 at, again should be 0V
F- confirm -0.7V at 7915 output.

Click to expand...

D.- 18V DC
E.- 0 V DC
F.- Sadly I confirm it

Click to expand...

The mystery here is what is holding that voltage down. Something is drawing a lot of current, and since it only has .7V across it, it won't generate as much heat as the 7915 so will be harder to find.


I don't trust your measurements especially around the 7815.

Also when debugging any circuit that you don't understand why it is misbehaving you need to divide and conquer.

How does the PS act by itself with no load? with simple resistor load? etc.  After you get a circuit block working by itself, integrate it into the larger system. 

Good luck.

JR