Audio has to pump up?

[MKwrench]

I have fun posting these, I always hope someone can learn along with me, and maybe geta kick out of seeing these old radios come back to life.
Here's another Sears Roadtalker, it's a 663.38070700 AM Radio. I had recently pulled a few radios off the shelf to make sure they work to put on marketplace.
Check out the video, It describes the issue best.
This happens any time the power is cycled, all of the electrolytics in the audio area have been replaced, the audio is always there at pin 4, but not on pin 10, and behaved the same way with C109 removed, so it must be localized within the 1156 audio chip. Pin 1 always had it's voltage and pin 8 is always had it's ground. The 1156 was swapped out with a known good chip this morning, but the issue remains.
I won't be able to peek at this again until Sunday.
The only thing I can think of is my using a 100uf in C105s spot instead of the original 33uf to help plump up the bottom of the amplified audio...... Thinking, this is a long output delay, a way larger cap would cause.....
As you can see after everything catches up, everything looks and sounds great...
Thanks for having a peek, here's the related SAMs and my favorite part *the video*.... Enjoy!

EDIT: I forgot to mention that the PA speaker output had the same issue.

 

[brandon7861]

If nothing remains but resistors and ceramic capacitors, I would bet that a ceramic cap has an internal fault, maybe a crack with moisture, idk. What happens if you blast them with hot air before turning it on? Or maybe heat them one at a time with your iron checking each time.

 

[MKwrench]

I wasn't going to have time until tomorrow morning, but plans change. I was exited to try Brandons suggestion of pre heating some of the 1156s nearby components. Before I grabbed the soldering iron, I wanted to look if any DC voltages on any of the pins changed during the "pump up", which could steer me to the correct trace to test. They all stayed consistant, even with audio added..... ......*except pin 4*.
On my scope I had what looked like a consistant input wave... In an AC measurement, but if I had been monitoring DC too I would have seen the waves average voltage dropping from 8V down to 3V during the "pump up".
I followed the path C104-C105-R91-C108, if I had 3 shorted caps I guess it could be possible ... but, not likely....
So, I went the other way R89-R88 and *C96* which was originaly a .22uf that I didnt have handy, so I opted for a 1uf hoping to open up some hidden audio.
The issue here is that my dyslexic ass put a 10uf in here. I figured out that I had the whole thing *Backwards*, Pin 4 Was Pumping Down, Not Up, the 10uf was slowing it down and the extra voltage was hitting the roof of the 1156s input capability.
The smallest electrolytic I had handy is a .47uf, which seems to work fine.

Some folks can say that they amaze themselves, then theres me who says I confuse myself... Hahaha!

 

[brandon7861]

edit: I misunderstood. Was it in backwards? Or just too big a cap"

 

[MKwrench]

edit: I misunderstood. Was it in backwards? Or just too big a cap"

I guess it was too big, I had no idea that could happen. It must have been like that for a while, I don't know why I didn't notice before...
I tried a 2.2uf and a 1uf (which I thought was there), they both caused simular delays.

 

[MKwrench]

The .47uf in that spot does cause a small delay that can be seen on the scope. I used a 1M resistor from the ( - ) of C96 to ground to chill the DC that ends up there. But 1st a 10K from the same spot to pin 4 for the slightly flatter response I was looking for. We now have impressive crisp/clean/loud response with a Plain Jane, cheapo DM507.
It got a quick receive, squelch and meter align/touch up, I'll put it in my truck for a few days of real world testing.
Then to marketplace. I dont expect much $$ from these old radios, the exteriors are sometimes pretty beat. I just feel better knowing that what I sell will work great for the person who buys it.