DC Offset with Cap Coupled Amp

[indianajo]

D1 is connected to the speaker. Speakers can kick back inductively and there is a 2 amp fuse. I'd use some regular diode 2 amp rated at least, unless you run a sim & prove that current is less. No schottky's. I bought diodesinc A6 for this, you may have a MR something. Newark is selling MR756 today, a bit of overkill for 2 amps rating I asked for. The lead might not fit the hole, I don't think the A6 lead would. You probably have scrap of something in a bigger case than a 1n4004 with a fatter lead. If it looks like a 1n4004, use that.
To quibble about Jean-Paul suggestion for hot rodding the power supply with 4700 for C404, in 1965 the AR3 acoustic suspension speaker hadn't hit the market yet. 85 db 1w1m madness hadn't started. I owned a pair of 101 db 1w1m 8 ohm speakers, and they made perfect music at 1 vac or 1/8 watt. SP2-XT's. The HHScott 348 may not have enough heat sink for more power. The dynakit ST120 sure didn't have enough heat sink, they blew output transistors regularly. Mine not since I upgraded the heat sink.

 

[dbxdx5]

What's in there for D1 is black and looks kind of like a bullet or a little rocket, i.e. cylindrical with a tapered "nose" on one end. I have some IN4007s, which I know is overkill, but I'll put one in. Still need to lift C114.

 

[dbxdx5]

Ok, I swapped my PNP replacement for the blown Q4 between channels and the bias issue moved. (It was an old GE-21 generic service substitute.) Suggestions for a modern PNP I can use here? The Scott internal coding of "QA-9" is leaving me in the dark.

 

[indianajo]

This will be a TO5 PNP? There aren't any at major distibutors.
surplussales.com of nebraska has some 2n2905, 2n2905a, and 2n2905jantx . click transistors page 1. I've gotten good parts from him, usually USAF surplus.
To save freight combining with other parts, newark usually has a pnp to39 or to18 that you can put a heat sink on. Use selector grid to find what number they have in stock today. Vce -60v up, TO18 or to39 package, pnp, in stock, search. They're ****ese but farnell has their own incoming inspection department, I've found. No counterfeits yet from newark.

 

[dbxdx5]

Before I ordered anything, I wanted to see if I could get the one channel to bias up using a temporary sub for Q4. So I plugged in a new NTE 129, which I thought should at least get this working. Well, same thing as with that other GE sub I used: Whichever channel I put the Q4 replacement in won't bias. So there's some parameter of the original QA-9 that these two generic PNP subs aren't delivering.

 

[indianajo]

Any transistor will collapse Vce if too much current goes into/outa the base lead. C leakage? Flux leakage?
You need to do a Iceo test of the original Q4 at 12 v or 30 v. You can get 30 v PS salvage from copier/fax/printers. Through the 47 k resistor.
All that replacing TO5 transistors tore off the PCB land on my dynakit PC14 board. I'd rather test than solder.
You realize TO126 or TO220 transistors have BCE pinout instead of EBC don't you?
I found the NTE 49/50 transistor as VAS or driver sounded like ****, no highs. Probably really a TIP31/2 with 6 mhz Ft. The original rca TO5 transistors were 50 mhz ft, I have one surviving pair. Ended up making a new board to use MJE15028/29, which sound fine with 30 mhz Ft.

 

[dbxdx5]

Thanks. I'll read up on how to do an Iceo test. I'm fairly confident that the circuits in the area of Q4 on both boards are fine, as that one original good QA-9 transistor is able to make each channel bias up correctly. Fortunately, the boards have sockets for Q4 (and several of the other transistors), so no desoldering/soldering has been needed to switch back and forth.

The NTE 129 and GE-21 I've tried at Q4 are both TO5 transistors, so there haven't been any pin-out issues.

 

[dbxdx5]

After much scanning of service manuals and schematics for other Scott SS receivers and amps, I figured out that QA-9 is a germanium PNP: 2N398B. I have to go through my very limited supply to see if I have anything that can sub, but if anyone reading this happens to have one they could spare. . . .

 

[indianajo]

Surplussales has 2n1193 for $3 . Vce 25 v.
The only way to reach the transistor lists is to click semiconductors transistor, then go to top and click in the text semiconductors which gets you a sort page.

 

[dbxdx5]

Resolution is OK for me. I'm viewing it in a small screen.


Please, acclare to me where are you seen the problem. At loudspeaker output or somewhere inside the circuit.
The audible difference between 1Cent cap and a 100U$S is negligible. So don't worry about it. It is only a question of Marketing. Pure marketing as Messi is.

Do you have a 50W light bulb of the filament type? If is the case, wire it at the lousdpeaker's terminals almost in one channel. And tell me what do you see in it.

Finally have both channels working again.

I just wired a 40W bulb to one of the speaker outputs and turned on the unit. The bulb didn't illuminate.

I've also replaced all of the power supply capacitors, and the issue persists.

 

[dbxdx5]

Some more information, fwiw.

I just completed recapping both amp boards--no change.
I noticed that the DC I'm measuring at the speaker terminals (8 ohm dummy load) with my DMM's goes up by several hundred millivolts when I turn the input selector to a different source. Nothing is connected to the inputs.

I swear this thing is possessed.

 

[OldDIY]

In your circuit, the load is connected through an electrolytic capacitor C109 to the ground rail. A small DC comes to the load through the GNFB chain. If С109 is serviceable, there is no significant DC at the output. It is not dangerous for the speakers. Artifacts (clicks) may occur when switching inputs.

 

[dbxdx5]

In your circuit, the load is connected through an electrolytic capacitor C109 to the ground rail. A small DC comes to the load through the GNFB chain. If С109 is serviceable, there is no significant DC at the output. It is not dangerous for the speakers. Artifacts (clicks) may occur when switching inputs.

Thanks. I think I was getting too caught up in DC blips that my meter was measuring, rather than just confirming that it wasn't audible. As for the AC noise I was seeing on my scope earlier, that seems to have been taken care of by recapping the power supply, amp boards, and preamp boards.

 

[dbxdx5]

I thought the AC noise shown in my scope YouTube videos earlier in the thread had been solved by the recap. No luck. It's audible through my test speakers as intermittent rumbles (for lack of a better word), though I don't hear anything when music is playing or between songs.

So still suspecting the bass pot area, I finally just replaced the four .068uF 250V film caps. No change. With the caps eliminated as the cause, I did what I should have done a while back: I kept the speakers connected and hooked my scope to one of the bass pot solder lugs while pushing on and slightly wiggling the shaft of the pot. Finally, I saw on the scope and heard through the speakers a cause and effect from touching the pot. I'll be keeping my eye out for a good used replacement, though I'm sure it'll be a while. Anyway, just wanted to update the thread and thank those who weighed in.

 

[rayma]

Try freeze spray to locate the component. It's most likely to be a transistor or carbon resistor.

 

[dbxdx5]

It's the pot.

 

[OldDIY]

This can be mechanical wear, oxidation of contacts, violation of the soldering of the leads. Sometimes cleaning spray and lubrication of the potentiometer track helps. For diagnosis, you can use a knock-knock.
This can be a time consuming procedure involving dismantling, disassembling, revising and replacing the pot.

 

[dbxdx5]

I did try cleaning it a few times with Faderlube F5. I'm tempted to try flushing it with CRC QD, but I think the issue is mechanical wear. For example, I can feel that there's noticeably more slop in the tolerance between the inner shaft and the bushing, compared to the treble pot. A replacement is in order, I suspect.