F4 Channel Problem - Running Cold

[vdi_nenna]

Hi All,

One channel of a F4 ran cold. Heat sink would not run hot like other channel. Originally thought problem was due to using a single turn pot for P2. Once replaced with multi-turn, channel ran fine. Then a few hours later noticed the stereo image was off again and the bad channel was running cooler than the functioning channel.

Not sure if it's a bad part or cold solder or other issue.
Here's some voltages of bad channel. Will get more.
Any recommendations where to look would help.


Supply voltage at board: +/-23v

P-Ch

G: -2.62V
D: -23V
S: -81mV

N-Ch

G: 2.96V
D: 23V
S: 81.0 to 81.4mV

D1: -2.7
D2: -2.7
D3: -22.8v
D4: -22.9v


TL431 (facing flat side, left to right)

Ref: -115.9mV
Anode: -2.62V
cathode: 2.9v

Thanks,

Vince

 

[Zen Mod]

if other thingies are proper (soldering etc.), just replace TL431

 

[vdi_nenna]

Thank you ZM. There's no burned parts
Will try replacing the TL431 among other basics.

Thanks.

 

[grimberg]

Vince,

Even though you are an experienced builder, double check if you have soldered ALL pins and if there is any cold ones.

 

[2 picoDumbs]

Yeah looks like the issue is definitely located around TL431.

 

[grimberg]

That's vague. There are many components around it.

🙂

 

[2 picoDumbs]

Number 1 suspect is TL431, followed by all solder joints on resistors used to control voltage of tl431

 

[grimberg]

I knew you could do better.

😉

 

[2 picoDumbs]

Maybe worth mentioning.
Texas instruments makes TL431 in 3 different operating temperature grades
0 C to 70 C
-40C to 85 C
-40C to 125C

It might be worth using the 125C graded parts.

 

[vdi_nenna]

I have another F4 built from same TL431 bulk tape.
It's been running for quite some time without issue.
Bad channel is in new amp and other channels is fine.

Using TI TL431AILPM -40C to +125C

Going to try a re-flow first.

Thanks,

Vince

 

[vdi_nenna]

Re-flowing the joints didn't help, but replacing the TL431 did.
Looks stable, but it was once before when biasing, then failed later.
We'll see.

Thanks everyone for your help and comments.

Vince

 

[vdi_nenna]

Problem is back after hooking up to speakers.
Placed on bench and saw the bias only went to 68mV across the source resistor, then it jumped to 150mV. It's now dropping and at 90mV.

After replacing the TL431, left on the bench for over an hour at 240mV, then dropped it to 200mV. Hooked up the F4 to a completely different front end and speakers.
Noticed after a few minutes the channels were at different temps.

I'm thinking it's a part that is cold soldered and is affected by change in temp, but that doesn't explain it running ok for almost two hours on the bench.

Going to try to reflow all the parts.

Like Mr Pass once said, "If it doesn't below up, you're 90% there." 🙂

 

[2 picoDumbs]

**** Happens- Forrest Gump Quote - YouTube

 

[2 picoDumbs]

Also check the joints on the electrolytic caps

 

[vdi_nenna]

Will make sure to start with the caps. thx.

 

[2 picoDumbs]

I remember Wayne or Papa saying more recently that they were having Elna Silmic cap failures after manufacturing moved to China.
Not sure if that might be applicable to your setup.

Just something else to consider, if you start pulling your hair out.

 

[vdi_nenna]

Interesting, because I'm using Slimic II in all positions across all 4 channels. 🙄

 

[2 picoDumbs]

I would try replacing a new TL431 (maybe one from a different batch) one more time, assuming everything else looks good.
Check soldering temp etc.

If it still fails while everything else looks good.
Then replace caps with panasonic FC or Nichicon.

 

[vdi_nenna]

I have Panasonic FM series rated at 108C.

 

[2 picoDumbs]

That will be good enough, at least to test.

Can you check if the FM has iron leads?
I think I had some but they had iron leads (magnetic) and my OCD got the better of me. The Panasonic FC definitely has copper leads.

You should use those to test regardless of what materials are used for the lead wire.