Thread: Sheesh
View Single Post
  #9   Report Post  
Posted to
[email protected] is offline
external usenet poster
Posts: 18
Default Sheesh

On Tuesday, March 2, 2021 at 9:47:32 AM UTC-5, ohgwrote:
On Tuesday, March 2, 2021 at 9:09:12 AM UTC-5, legg wrote:
On Sat, 27 Feb 2021 12:27:16 -0800 (PST), "ohg..."
ohg... wrote:
On Saturday, February 27, 2021 at 12:42:20 PM UTC-5, jurb wrote:
Lately I work on audio and have long thought that I get the **** that has already been gone through but I think I have ALMOST seen it all.

This amp in a receiver, not the greatest but is vintage, whatever it needed outputs though I don't know why.

So it is offset bigtime. The problem ? A 27K going to the emitters of the diff pair. So with that replaced it centers the voltage and all it supposeedly well but it has distortion. Wracked my brain or a bit and I decided to go inoto detail mode. What do I find ?

An 2.2K ohm resistor where an 18 ohm belongs. What's more it looked like it had never been changed or if so it was a very long time ago because the connections had the same dust on them as the rest of it.

**** like this is not normal, and I am dealing with more and more of it.

Then there is this quad unit, outputs good but it had more bad transistors than you could sweep off the floor. get the amp fixed and a channel is cutting out. Bad connection ? Nope, a transistor. Got that fixed and now off to another board that is no doubt full of bad transistors.

Life is trip, think of the destination.

The only time I really remember transistors being a red flag (other than early Germaniums) were some small Hitachi signal transistors from the 70s-80s - the TO-92 ones with the beveled top. Those things would go leaky or open or intermittent (their choice). It got to the point back then that I just automatically changed them whenever I saw one in anything I worked on. Pretty much eliminated any rechecks.

We had an issue with parts in TO98, which was a common package for
early GE plastic signal transistors. 2N339x 2n34xx. These were
encapsulated off-shore, possibly even a simple rebrand.

At the time, there was a suspicion that they weren't normal
supply chain product, as the bad parts were purchased through

Although conforming to TO98 outline, they were a filled-epoxy
package, vs a molded epoxy. Anything showing up with the
filled-epoxy construction was simply dumped at incoming
inspection. Rubbish.


We've all seen flat out counterfeits. When I used to repair an amplifier with blown outputs, I used to return them along with a copy of my do's and don'ts - things like checking for wayward strands of speaker wires, looking for speaker wires punctured under carpet strips, not doubling up speakers on one terminal block, leaving sufficient ventilation on top of the amplifier, etc. I remember when I had amplifiers coming back with reblown outs, and I blamed the customer. Turns out the semis I was buying were counterfeits, and it wasn't just audio transistors. We ran into this problem with television vertical output ICs, power supply ICs, horiz (line) outputs, STK convergence modules, and even Sony jungle ICs.. I remember going through my parts supply and tossing a good portion of them out and making sure to buy from only reputable suppliers or directly from the OEM.

You ain't ****ting. What I don't see is how they can go to all the trouble making counterfeits that actually work but are like underrated for the job. If you are going to make ICs why not just make them right ?

So now I got this Pioneer QX-747. It had so many bad transistors it ain't funny. Three channels of the power amp and then one other board they call "equalizer" and now the SQ decoder. Not one power transistor or capacitor bad.. It ran me out of elcheapo small signal NPNs ? I don't even remember how many I had, but I just ordered 25 more. (hope that does it lo, naw it will) I guess it is my fault, the guy said quad is not that important but I told him you want that RM function at least, sometimes it sounds good, and sometimes playing movies through it is really nice.

The manufacturers want us dead. They know how warm a vertical IC will get and the put the boost cap right near the heat sink. So this is not that conclusive but on the other POC somewhere I got part of a schematic of a Mitsubishi power supply that is. It has a connection in the feedback loop of a 5 volt or whatever that is made specifically to make the Vcc(ee, whatever) fluctuate to corrupt the data in the IC. Another nail in the coffin ? They sent out a bulletin that the customers were to be told where to download a file to a USB drive and boot the TV with it is themselves, cutting us completely out of the procedure.

The manufacturers waged war on us, I took it that way. I figured I was a general.