Pioneer SX-1050 one channel sound distortion

[markthefixer]

Do NOT disconnect, cut or unwind any wires.

The transistors are socketed and after the two screws are removed, pull straight up from the heat sink.

There are three layers to the assembly, the transistor(front), the heat sink(with 4 holes in it) and the socket(rear) which has two receptacles for pins and two threaded hole receptacles.

You can probably use the thermal mastic, but you are nowhere near replacing the transistors yet.

Leave the other resistors alone. low ohms in circuit are hard to spoof.

 

[CohibaJoe]

Mark...As for my Re-Cap Project..any reason you cannot use:

MULTICOMP - MK3301/S - Insulating Kit
http://www.newark.com/multicomp/mk3...rue&MER=ACC_N_L5_SemiconductorsDiscretes_None

This are Silicone washer that "Heat Sink" compound is not used.

If you what to use "Heat Sink" compound then use:
MULTICOMP - MK3301 - Insulating Kit
http://www.newark.com/multicomp/mk3...T_N_L5_FansHeatSinksAndThermalManagement_None

If ymarko is removing the NEC Power Trans, best to replace the Insulating with New ones.

 

[ymarko]

Good to learn that there power transistors simply pulled out.

I tested four transistors Q24, Q26, Q28, Q30 and they all seem good. I attached the measurements in Excell form (downloaded from AK). My measurements are in blue (top line). Black (bottom line) belongs to AK form (for reference only, show what numbers should be in good transistor).

 

[markthefixer]

Yes, the transistors look good, and you are sure they are from the channel with the fried resistors?

There is a little voice in my head saying that the current that fried the resistors had to go through some transistors somewhere....

These are the resistors:
(mouser.com)
71-CP5J-0.5 0.5 ohms 5 watts $0.49 ea

but we are not yet at the ordering stage. We need to test the transistors and components on the board.

 

[ymarko]

Exactly. After I took heat sink out of receiver, suddenly there is no R & L sides anymore. That was my doubt too if I removed transistors from the R channel. So I reviewed few times and it seems they’re indeed from R channel.

Now I have to test four small transistors on board Q14, Q16, Q18 and Q20. Do I have to un-solder 1, 2 or all 3 legs? Or don’t un-solder any leg (measure them on-board)?

 

[markthefixer]

Exactly. After I took heat sink out of receiver, suddenly there is no R & L sides anymore. That was my doubt too if I removed transistors from the R channel. So I reviewed few times and it seems they’re indeed from R channel.

Now I have to test four small transistors on board Q14, Q16, Q18 and Q20. Do I have to un-solder 1, 2 or all 3 legs? Or don’t un-solder any leg (measure them on-board)?

This is critical enough that unsoldering the leads for isolation is necessary.

If Q18 & Q20 are actually removed, they too have thermal mastic and mica insulators.

 

[ymarko]

Tested Q18 and Q20. Both look fine, result attached.

Do I have to solder them back now or wait?
If solder them back, use thin layer of thermal mastic in both sides of mica insulators, right? The same question for installation of metal power transistors.

Do you suggest to un-solder Q14 and Q16 to test?

 

[markthefixer]

They are easy to remove and test, but the orientation is easier to get wrong on reassembly.

It is very surprising to me that only those two resistors suffered damage. But the results are irrefutable. The transistors could be weakened, but that's the hardest thing to spot.

They are three "layers" back, and if the first two layers survived, these are likely good.

The reason for the testing is the price of shipping, the replacements cost is far less than the cost of a second shipping cost.

Yes, reassemble them with a thin layer of thermal compound, there should be minimal squeeze out, then solder them. Run ohmmeter tests to the heat sink, a problem will show up as 5 ohms or less, other readings are in-circuit effects.

When reassembling the round transistors, make SURE the transistor socket is in it's recess in the heat sink before tightening - it's best to get a feel for the socket's recess into the heat sink before trying to insert a transistor.

In testing for heat sink continuity, I dig in one probe (or find a bare spot) and test for zero ohms to be sure the probe is making contact, then afterwards I re-test for zero to be sure the probe didn't slip and the reading is trustworthy.

 

[ymarko]

Run ohmmeter tests to the heat sink, a problem will show up as 5 ohms or less, other readings are in-circuit effects.

In testing for heat sink continuity, I dig in one probe (or find a bare spot) and test for zero ohms to be sure the probe is making contact, then afterwards I re-test for zero to be sure the probe didn't slip and the reading is trustworthy.

Could you please explain this a little bit more? Between what and what should I test ohms? What ohms it should be?

And between what and what to test heat sink continuity?

 

[markthefixer]

The metal of the heat sink is at ground potential, and NONE of the transistor leads are at ground potential, so if ANY of the transistor leads makes contact with the heat sink - blooie -

The heat sink is covered by paint or anodizing, but this coating cannot be relied on for electrical insulation to prevent a short circuit should a transistor's lead come in contact with the heat sink, coated or not.

The paint or anodizing on the heat sink will impede the test probe from making electrical contact to the heat sink's base metal, giving a false hi-resistance reading in case one of the transistor leads IS touching the heat sink.....

ALL tests will be between each of the three connections per transistor and the heat sink's bare metal.

Since the transistors are IN circuit, there WILL be resistance readings, and any reading over 5 ohms of resistance will NOT be a short circuit to the heat sink.

 

[ymarko]

Assembled large power transistors on heat sink. Heat sink has bare metal holes for screws, so I used them to test transistor leads. All leads show OL in both continuity and ohms tests.

Soldered back Q18 & Q20 transistors, mounted on a small metal plate (probably small heat sink). All leads show OL in both continuity and ohms tests.

Un-soldered and tested Q14 and Q16. Both tested bad (although visually look good), result attached. So now, as you predicted, we have faulty transistors (an addition to resistors R64 and R68).

 

[markthefixer]

I suspect we just have a transistor identification definition disjoint....

It looks like you recorded the pnp transistor on the npn grid
and the npn transistor on the pnp grid.

The 2sa and 2sb transistors ( Q14 is 2sa858 : ECB) are PNP
the 2sc and 2sd transistors ( Q16 id 2sc1439 : ECB) are NPN.

a tipoff is the four OL's and two reasonable numbers, in unexpected places in each report.

The lead identification for the two is to have the leads facing down and the flat facing you.

I'm used to this format, thus it takes longer to spot strangeness in different formats. I just add the results at the end:

first the PNP:
black lead to base, red lead to collector, example = 650, my reading: 687
black lead to base, red lead to emitter, example = 688, my reading: 678
red lead to base, black lead to collector, example = OL, my reading: OL
red lead to base, black lead to emitter, example = OL, my reading: OL
black lead to emitter, red lead to collector, example = OL, my reading: OL
red lead to emitter, black lead to collector, example = OL, my reading: OL

next the NPN:
black lead to base, red lead to collector, example = OL, my reading: OL
black lead to base, red lead to emitter, example = OL, my reading: OL
red lead to base, black lead to collector, example = 677, my reading: 656
red lead to base, black lead to emitter, example = 680, my reading: 662
black lead to emitter, red lead to collector, example = OL, my reading: OL
red lead to emitter, black lead to collector, example = OL, my reading: OL

 

[ymarko]

I suspect we just have a transistor identification definition disjoint....

The lead identification for the two is to have the leads facing down and the flat facing you.

Exactly! That was my suspicious too, therefore I tested them twice. However, I looked on them with leads facing up, so as BCE.

Now I re-tested them and they look good, here’s the result.

 

[markthefixer]

Exactly! That was my suspicious too, therefore I tested them twice. However, I looked on them with leads facing up, so as BCE.

Now I re-tested them and they look good, here’s the result.

Well, I guess it's time to order the resistors. Get four. Swap all in the channel. It all depends upon where the current was flowing as what was racing towards destruction. This is cheap insurance.

No reassembly of board into unit, nor installation of big heat sink, until resistors are here. But transistors can go back on all the heat sinks.

 

[ymarko]

They shipped me 4 resistors today. Original resistors visually look good and show:
R64 and R68 – 5.8 kOhm and 14.2 kOhm (bad)
R62 and R66– 0.5 Ohm each (good).
I replaced all four with new resistors.

 

[markthefixer]

Well, since the transistors tested good, it's now time to carefully reassemble - watching very carefully for shorts and broken wires.

After that it's dim bulb time, with some test settings of the offset voltages and idle currents. There's several recent threads that reference the dim bulb tester, and directions for a trial power up and idle current test - which leaves things turned down ** after each channel is tested. If it all tests out, the dbt comes out, it's plugged directly in and the settings done for real.


** settings done on the dbt would be damagingly high when subjected to unrestricted power line voltage.

 

[ymarko]

I re-assembled and connected via DBT. Bulb dims in 1-2 seconds.

VR3 (good L channel) behaves exactly as before. It has 0mV in CCW position, then steady increased as I turn it CW. In about half turn it reached requires 50mV (then I set it back to zero)

VR4 (bad R channel) behaves different then before. It has minus 4.9 mV now (before it was huge 43 Volts!!!) in CCW position. But the rest is the same - turning VR4 CW doesn’t change it. It stays at minus 4.9mV all the way. I re-tested to make sure. All the same, but minus 4.3mV this time.

Then I tested DC speakers offsets (still on DBT). Both channels are steady now. R channel is holding steady now on about minus 36mV, no jump at all even if I increase the volume (it used to jump every split of second in wild range, especially when volume was increasing).

I hooked-up headphones and can say that R channel sounds not bad, but it still probably obscured by DBT, so it's hard to say for sure. L channel is great.

Replaced resistors hold their value 0.5 Ohm.

Do you suggest it’s time to plug it directly (w/o DBT) and adjust VR3 and VR4 for 50mV (or as much as VR4 allows) and test the receiver?

 

[markthefixer]

Not being able to adjust the right channel idle current with vr4 is a problem.

We'll be taking a lot of measurements with the dbt connected, and vr3 (left channel) turned down.

 

[ymarko]

All right

 

[kubotadave]

I have same exact problem with my sx 1050.How did this turnout?