APT 1 power amp – undeservedly forgotten

If its quiet with the preamp connected I'd leave it alone. One thing you might try is inserting a 10 ohm resistor from the RCA ground on each RCA to the circuit board trace where your getting your ground now. Make sure your RCA's are issolated from the chassis.

The manual says "The input jack chassis ground is the ONLY chassis ground in the amplifier". So they specifically mean that that one screw connecting negative end of RCA connectors to the chassis is the only ground connection.

What worries me is that right channel buzzes much louder than left when not connected. If the screw connection between chassis and the board/RCA is disconnected, both channels buzz loudly and evenly. When the screw is connected, left channel buzz is barely audible, but right channel buzzing is exactly the same as if there was no screw connected to chassis.
 
The manual says "The input jack chassis ground is the ONLY chassis ground in the amplifier". So they specifically mean that that one screw connecting negative end of RCA connectors to the chassis is the only ground connection.

What worries me is that right channel buzzes much louder than left when not connected. If the screw connection between chassis and the board/RCA is disconnected, both channels buzz loudly and evenly. When the screw is connected, left channel buzz is barely audible, but right channel buzzing is exactly the same as if there was no screw connected to chassis.

I'm sorry its been 20+ years since I saw a Apt power amp. I don't believe I even have a schematic for one anymore. Are the RCA jacks a molded one piece assembly or are they individual jacks that are insulated from the chassis?

There seems to be a slight ground problem in my opinion. I've seen this same problem with Gas Son of Ampzilla's. I can remember one years ago that I ended up insulating the RCA's from the chassis and soldering a 10 ohm resistor from the RCA jack to ground. The problem was solved and I had a amp that was 100% quiet.

The fact that there is no noise with the input of the amp is terminated to the output of a preamp makes me say simply leave it alone unless you have the experience and gear to check out the problem.
 
It's a single piece, which has a negative portion all around the block (soldered into the board and to the mounting bracket - mounting bracket is in turn connected to the chassis with a mounting screw.

I have gear and experience, but at this point I don't think it's an actual problem. It looks like it's the way the wires are laid out inside the amplifier.

Most wires come out on the 'left' side of power board, so they are much closer to the left channel. All right channel wires are longer and have to go right beside transformer.

There is no real way to readjust those wires. The chassis is packed when the boards and heatsinks are screwed up in their final positions. I tried moving wires around with a wooden pick, and I can hear the change in buzzing for sure when the wires are being moved. There is just not much space to reroute those wires.

Ideally I may have to take out the right channel assembly, the power supply board, and reroute the wires as from the transformer as possible, which will move them by a few mm at most.
 
This is what it looks like inside between right channel, transformer, power board, and filter caps under the power board.
 

Attachments

  • IMG_20170911_223719.jpg
    IMG_20170911_223719.jpg
    179.4 KB · Views: 449
APT/Holman Amplifier Bias/Offset ?

Hello there, I am an old friend of Tom H.

Resuscitating my 3 pcs APT Holman power amps, the usual oxidized relay contacts and discovered an intermittent RCA jack, resoldered the PCB traces. Works fine.

Need to trim the adjustments, not touched since they were built.
QUESTION: Anyone know the recommended procedure to set offset and bias trimpots?

First preheat at modest load?
Adj offset to 0 DC out +/- 10...20 mv?

Adj Bias:

Set 1 W (or 10W?) output power at 1 kHz?

Use THD analyzer, observe residual THD products.

Trim bias control for lowest zero crossing distortion?


MANY THANKS

Jon PAUL
Crypto-Museum
 
Couple Questions

Couple questions about the amp. My left channel started humming, so took it apart and found C26 had gone bad. Left a mark on the board so easy to spot. Ended up pulling all the caps on both boards and am re-capping.

Things I noticed. C16, the 100uF cap on the + input of the diff amp seems like it should be non-polar. I am replacing with a non-polar.

R6/R13 were 39 ohms not 360 as specified in the schematic. I assume 39 is correct.

Q7 on the left channel with the bad cap had a beta of 60 and I was expecting closer to 200. I am replacing it with a MPS-A42 just to be safe.The zener was ok.

While I was looking around I noticed Q2/Q3 on the right channel had a beta imbalance. One was 80 and one was 200. I ordered about a dozen A42's, and am thinking I will find a close matching pair and replace the right channel Q2/Q3. Seems like a diff pair should be pretty tightly matched.

The only other oddity was the NPN output devices were all around beta of 50. But the PNP's were wildly different. Each channel had one device with a beta of around 80 and one with a beta of around 35. Is this a problem?


Thanks if anyone is listening
 
So after recapping both channels and replacing Q7 and replacing most of the resistors with metal films while I was in there, no change. Disappointed is an understatement. Output was still running -1.5V and humming. So next I replaced the Q2/Q3 diff pair, the current mirror and the cascode pairs. After that, things got better, output offset is now -.5V, but still hums. So next, I added a 2nd 47K resistor across the input because the base of Q1 on the left channel was around -.2V while the right was -.144. By adding the resistor, I got the base voltage to -.1V. Hum stopped, output was around -50mV. So I think what was happening was the DC servo op amp was saturated (it was -9V) and was letting ripple thru causing the hum, instigated by low beta of Q1. Although measuring Q1 with a transistor tester came back around 300. Q4 was hotter and measured 380, so maybe it is the mismatch.


So now the next problem, what to replace the obsolete 2SC1345 with. Digikey appears to have active the KSC1845 which many quote as a good replacement. They have no stock though and 6 weeks to get. Using just ft, beta, and a few other parameters, it looks like an MPSA18 might be a decent match. Decisions decisions. If I go with the KSC1845, I may order 2000 units. Price is 100 for 2K units, and the way discretes are going obsolete, having a cache might not be a bad thing.
 
Mouser is showing about 32,000 KSC1845FTA in stock at this moment. Should be in stock for a while, but they have run out before, with several weeks between shipments.


Thanks, that will save me a couple weeks wait. And to the other poster, correct you are, thru hole is going away, not smd discretes.


One more thing I may do when I replace Q1/Q4 is reduce the OpAmp drive resistor from 47K to around 30K so the OpAmp can push the offset harder. The input side has a 47K to ground for the bias current whereas the other side of the pair has only 870 or so to ground. I had not really looked at it before, as I always assumed the OpAmp was just to push the typical diff amp offset, but there is a baked in offset from the mismatched input resistance to ground between Q1/Q4.