Read the number before you order the part. On this Sony HP-510A the big metal cans on the heatsink say SONY D291. Those are output transistors. The small gray parts with a metal tab are the drivers, and one of them reads A706. Do not order a set of D291s because they are the parts you can see.
The right channel on this set still crackled after DeoxIT, on every source, and the volume knob did not change it. That puts the noise in the power amp. The tests that got to that point are the one-channel guide. The board was reached by lifting the turntable out, which is the case guide.
The noisy part has not been named. Freeze spray was ordered and came back undelivered. Nothing on this page claims a finished repair.
What is actually on this HP-510A board
| What you see | What it is | Replace it when |
|---|---|---|
| Five metal cans on the heatsink, SONY D291 P76 | 2SD291 output transistors | One of them tests bad, or it reacts to freeze spray |
| Small gray square, tab on top, marked A706 | 2SA706 driver | That channel's driver pair is the one freeze spray points at |
| The matching small part whose number starts with C | 2SC1124 driver, the partner | Always with the 2SA706, as a pair |
| A part with a slot you could turn | A trimmer, usually idle current | Do not turn it while you are hunting crackle |
The cans on the heatsink
There are five of them, bolted to a sheet of metal. Every one that could be read says SONY D291 P76. That marking is a 2SD291, the NPN output transistor in a large metal package. Shiny legs are normal. Black or dark gray legs are the oxidation a former Sony technician has described as a crackle source on sets from this period, and that clue applies to the small parts as well as these.
A D291 is the expensive, hard-to-match part. It is the one a missing driver can destroy. It is not the first part to pull.
The small gray ones
The drivers do not look like the cans. On this board they are four small gray squares with a silver tab. One face reads A706 AU31. Another starts with C and faces the other way, so the full number is harder to see. That pair is the 2SA706 and the 2SC1124. Audiokarma's long-running failure list flags both for gain that falls off with age, and it flags Sony's smaller 2SC632 and 2SC458 for noise. Those smaller ones are a different package again. If you cannot see a three-legged part besides the D291s and these gray squares, you have found the drivers and you have not found the small-signal row yet.
Original part and the substitute restorers use
| Original | Job | Current substitute |
|---|---|---|
| 2SA706 / 2SC1124 | Driver pair | Toshiba TTA004B / TTC004B. MJE15033 / MJE15032 if they fit |
| 2SC632, 2SC458 | Small-signal | KSC1845, and only if freeze spray points at one |
| 2SD291 | Output | Only the one that tests bad. Match the package, at least 40 V and 3 A |
The older driver substitute was KSA1220 and KSC2690. KSA1220 is obsolete, which is why the Toshiba pair replaced it on repair benches.
Buy those from Mouser or DigiKey. A marketplace listing at several times the distributor price is the pattern for a remarked fake. A fake in the driver seat can fail and take the 2SD291s with it. This site does not carry an affiliate link for those transistors.
Name the part before it comes off the board
Play the bad channel at low volume. One short burst of freeze spray on one transistor, then listen, then wait about half a minute before the next part. Start with that channel's driver pair. The part that makes the crackle change, stop, or get worse is the suspect.
Pulling the transistors one at a time to see which one stops the noise is how a power amp loses its outputs. A missing transistor in this circuit can swing the output to DC or let the 2SD291s overheat. The speaker voice coil goes with them. Each desoldering pass also lifts pads on a 1970s board.
When a driver does come out, replace both drivers in that channel. The originals put the base in the middle. The Toshiba substitutes do not, so two legs have to cross. Mark emitter, base, and collector from the part that came out, check the new datasheet, and sleeve the crossed legs. Power back up through a dim bulb tester, then check the speaker output for DC before a speaker goes back on that channel.
The service voltages for this chassis are in SAMS Photofact MHF-44. If nothing reacts to cold, compare the good channel's voltages with the bad channel's. The first point where they disagree is the fault.