Preamp module fault. No audio. -18v is low.

[steenamaroo]

Hi all.
I racked eight wbs m490 modules last christmas and then went to uni until now.
When i left they all worked, and now that i'm back only seven of them work.
The one that doesn't work powers on, but does not pass audio.

I've done the usual trouble shooting and proven the module itself to be faulty.

I've also ensured that the switches for input and output selection are not at fault.

I've tested the resistances of the input transformer and they check out fine.

There are no visibly damaged components and the opamps have been tested/replaced and proven to be fine.


Two things that i've discovered.

The module accepts +/-25v from a psu which is tested, and turns it into +/-18v.

Testing the power section of the module showed that i'm getting the +18 as i should, but where i should get -18 i'm getting -12.
I've replaced the relevant resistors/diodes/caps (c51 c53 r148 cr17) for the -18v and it's still showing -12.

The points on q1,q2 and q3 where my other modules show -15v or so is showing -18 on the duff module.
I replace Q1 on a hunch but to no avail.

Full schemo is at www.steenaudio.com/m490.pdf

Can anyone take this problem a little further? I'm a little out of my depth (as usual).

Thanks in advance.

 

[Walrus]

If you only have -12v on C53, then I don't see how you can have -18v on Q1 Q2 Q3 unless you are not measuring with respect to ground.
I'd check Z6 first, as this is connected direct to the -18v line, then Z3/4 and maybe C22.

 

[steenamaroo]

Sorry, that's my mistake, i'm getting 18v at q1,2+3, although the other modules read between 15 and 16v at these points.


The problem still stands though. anywhere on the board that should have -18 only has -12.
-18v is not present at z6. It's the same -12v.

Thanks for the reply.

 

[MarkusBass]

The problem seems to be simple although fixing it may require some soldering. The +/-18 voltages in this module are obtained from +/-24V rails and voltage drops on R147 and R148. It looks like the -18V rail is drawing more current than expected. So I would start with measurement of the voltage drop on these resistors, calculation of the current consumption (from Ohm's law) and comparing the results with other modules. It would be the best if there were a jumpers on +/-18V rails that you could disconnect temporarily and check how this influences the rails. If there are no such jumpers, you have to investigate the problem a little bit more, possibly with external power supply. I assume that ground rails are contacting correctly and you are measuring the voltages in reference to the ground, are you? Are the integrated circuits in sockets, or they are soldered directly?

Mark

 

[steenamaroo]

Thanks for the reply Markus. That all makes sense.

drop across 147 is 23.6 to 18.7 which seems about normal.
drop across 148 is 23.8 to 11.8

both resistors read as being 620 ohms.

I had thought about disconnecting various sections that use +/-18 for process of elimination.
I think i could do it, but i wouldn't know if it's safe; I just don't know enough about it.

The grounds check out fine and the ICs are socketed. I did remove them all and replace them with ones from a working module but this made no difference.


**update. i removed r148 to test. I got the expected -25v at one side, and +5 at the other side.
The other modules display -25 and -11 in this situation.

 

[s2udio]

Isolate the PSU from all the modules and start there,  check all the voltages with load resistors
if nessesary, around  5k 2W should do it.

 

[steenamaroo]

thanks for the reply.
Sadly the psu is fine.

I have one psu running a set of eight modules (one of which is the faulty one)
and a whole separate identical psu that runs a pair of these modules.

Ive moved the duff module to the second psu as it's on my bench and easy to access for testing.

These are both jlm audio psus, and i am testing voltages with reference to ground.


Working modules run fine on both of the psus.

 

[Walrus]

As I said, check Z6 as this is connected direct to the +-18v rails. Then check the volt drop across R55 to see if Z3 or Z4 or C22 are faulty!

 

[steenamaroo]

i've done that.

Z6 is getting -12v- not -18.

the volt drop across r55 is the same as with the other modules. Voltages here are -11.8 and -11.1.
Wrong voltage, normal drop.

Replaced z3 z4 and c22 with parts from a working module with no change.

+voltages are fine at z3 and z4, but the negative voltage is 11/12. Same story all over the board.

 

[steenamaroo]

Ok i have news...

sorry folks, i know this must be tedious,but i'm doing my best!! Thanks for everything so far.


I figured -18 only goes to so many places...

z1 via r22 etc

z3+4 (with c22 as potential leak to ground)

z6

so i removed all these opamps and tested.....Still -12 instead of -18.


then i put everything back and disconnected r22 which gave me -18v where it should be.

I guess that points me to testing c4?

 

[PRR]

C4 is shorted.

How can I say that?

What Mark said.

Where does -18V go? I didn't study the full schematic (you should) but I only see bias for the Q1-Q7 circuits.

R148 is dropping 25V-12V= 13V, and is 620 ohms, so must be flowing 20mA.

-18V flows to R32 direct, but R32 is 15K (and unlikely to fail short) so the worst-case fault (top of R32 shorted to +25V) could flow 2mA, not 19mA.

Other -18V load is R22 1K feeding 10+10K and 22K. Same argument: worst-possible short to +25V at top of R13 or R19 would flow just a few mA, not 19mA.

Consider a short at C4. R22 is 1K, has zero one side, observed -12V other side, is flowing 12mA. That's not quite 19mA but close enough to be suspicious.

Also you have high positive voltages. This suggests R147 and R21 are NOT passing their usual currents, but less. Indeed if C4 is stealing all the current up through R13 R20, Q1 Q2 Q3 aren't getting fed enough, are not sucking against R147 and R22.

 

[PRR]

You got there 3 minutes before I did.

> I guess that points me to testing c4?

Don't even "test". Just lift one leg of C4. If that restores proper DC conditions, you found the problem.

You probably will have to get a new C4. In general a missing rail-cap degrades audio performance. Worst-case the system motor-boats at FULL output. In this case I think it will be stable, but with a little more buzz and crosstalk.

 

[steenamaroo]

you my friend are a genius.

I'm a little gutted that i was comin back to the pc to say "c4 is the fault", and you got there first though!!! lol.


Thank you all....the best thing about this was the explanation. It's much appreciated.


I don't have any 4.7uf in "stock" but i have 10uf....that'd be fine right?

 

[MarkusBass]

drop across 147 is 23.6 to 18.7 which seems about normal.
drop across 148 is 23.8 to 11.8
both resistors read as being 620 ohms.

Paul, you forgot to calculate the currents:
R147: (23.6 - 18.7)/620 = 7.9mA
R148: (23.8 - 11,8) / 620 = 19.4mA
This clearly shows that there is a short circuit after R148.
Then you can even calculate after what value resistor is the short circuit: 11.8V/(19.4mA - 7.9mA) = 1 kOhm. And this is exactly the value of R22  8).

I had thought about disconnecting various sections that use +/-18 for process of elimination.
..The grounds check out fine and the ICs are socketed. I did remove them all and replace them with ones from a working module but this made no difference.

Replacing ICs was not needed. You could get information about the short circuit after you removed ICs from sockets. Also replacing Q1 was not needed. The first test in such cases is to remove ICs (if this is possible).
10uF in place of C4 is fine.

Mark

 

[steenamaroo]

You're right Markus. I get carried away,lol...+ i wasn't really sure what to do,

now that i see it it's so friggin simple!!!!

I'm definitely better equipped for next time guys!!

Hey Markus
Where'd ya get Paul from?

 

[MarkusBass]

Afterwards it's alway so simple. The catch is to make it simple for the next time before you start fixing equipment.
I thought that you are Paul from steenaudio studio, aren't you? Sorry if I made a mistake.

Mark

 

[steenamaroo]

yehp you're right. That's me..

How'd ya know?

**edit...blonde moment. Forgot i put a link in the OP.


The pics on that site show a digi 003 in the middle of my desk.
It's now been replaced with the modules in question lovely stuff.