Mounting new transistors with heat sink on the SX-737 Main Amp

[pa_200]

I have the power amp AWH-033 in front of me and want to change all the transistors Q1 until Q10. The 4 Output Q11 – Q14 shall stay at this stage.

I use the BOM list for the SX-737 and everything was fine until I came to the Q7 - Q10. Original installed are 2sd357 and 2sb527 and I have ordered the recommend replacement SC2690 and SA1220.

The issue here are the pins. The old ones are BCE pinout and the new ones have ECB pinout. Normally that not a big deal to bring the legs into the right position, but these four transistors are mounted on big heat sink which have to sit at a special position on the board, due to limited space around.

With the changed leg position on the new transistors it is nearly impossible to mount the transistors on the heatsink crossing the legs E and B without risking a contact between all tree pins.

One could mount the transistor backwards, that will bring the legs into the right position, but then you have the metal side not in contact with the heatsink and even worse, it makes it impossible to read the part number.
We talking about the four U-shaped black heat sinks on that picture.



What are your recommendations on these issue and how can I solve it, still having a good heat dissipation.are

There will be a similar challenge on the power supply board with the q3 2sd313, replaced by a SC2073 if it's not a BCE pinout.

 

[ivandezande]

Use MJE15032/3. Also that 2SD313 really should be a TIP31C.

 

[EastPoint]

Yeah, I would not use the 2690/1220 for those. Either the KSA940/KSC2073 or the MJE pair.

 

[markthefixer]

If that picture is the FINAL version of the amp, I see the dreaded 2sc1451 at the Q5 and Q6 positions.

New trimpots are advised, and be VERY VERY careful with the thermal bias diodes wired to the amplifier board, yet mounted to the big power transistor heat sink.
Those leads LOVE to break off flush at their case.

 

[pa_200]

@ markthefixer – the picture was made before unsoldering and shows the original untouched board. New trimpots are already in position, I will/have to use 200 Ohm because 100 Ohm are not at hand, but that shouldn’t be a problem with a careful adjustment of the bias. The 2sc1451 will be replaced by new SC3503 with heat sinks.
The good thing on that board inside the SX-737, I could reach every part from the foil side without taking the board out. So there is at this time no need to touch the thermal bias diodes.
All the old caps will go as well.


@ EastPoint – can you give me a little help why you would choose the KSA940/KSC2073 instead the 2690/1220?
So far I have no experience with any of them, but I prefer replacements that give me an improvement in sound, noise, quality and durability. For the encountered mechanical problem, I’m working on a solution.

 

[larryderouin]

the 2690/1220 pair re TO-126 mounts, (hence the ECB Pinout) while the MJE's and the 940/2073 are TO-220 mounts which correspond with the originals (basically drop in. ) Also heavier duty than the TO-126's type.

 

[pa_200]

@ larryderouin - so the 940/2073 with the TO-220 package will be the better choice instead of the 2690/1220 pair as TO-126 mounts. What about the sound and noise values?

Maybe someone will change the SX-737 list for the AWH-033 or write a remark with respect to the mount.

 

[markthefixer]

940/2073 with the TO-220 package, vs 2690/1220 pair as TO-126 mounts

Those are driver transistors, and have 180 pF caps Base to Collector to tame any tendency to RF oscillate.
They won't have full load, they will have an appropriate loading for a heat sunk to-126 case.

 

[pa_200]

Mark I am a bit confused now, which pairing should I use as a good replacement? The 940/2073 or the 2690/1220? I have both here.
On the picture you see on the left one of the 1220 (ECB pinout, TO-126 case) with my custom fit heat sink and right side the 940 (BCE pinout, TO-220 case) on the factory fit heat sink.


This time with the change visible on the backside

 

[EastPoint]

Use the 940/2073 pair. They're more heavy-duty and have the same BCE pinout as the originals.

 

[kdris10]

Shave the legs off the heatsinks.

 

[larryderouin]

Au contraire. 940/2073 are BCE. Peruse the Front page of the Data Sheet below!

 

[EastPoint]

On new transistors, almost all the metal-backed ones are BCE and almost all the others are ECB. Almost.

 

[pa_200]

I double rechecked the 940/2073 with the paper and the little LCR-T4 Mega328 Transistor Tester. They are both BCE and fit perfect with the orginal heat sink.

But unfortunately, for some reason I kill at least two outputs, so I start a new thread for troubleshooting on that issue.

 

[kdris10]

Au contraire. 940/2073 are BCE. Peruse the Front page of the Data Sheet below!

View attachment 1582729

My bad Larry

 

[ivandezande]

I double rechecked the 940/2073 with the paper and the little LCR-T4 Mega328 Transistor Tester. They are both BCE and fit perfect with the orginal heat sink.

But unfortunately, for some reason I kill at least two outputs, so I start a new thread for troubleshooting on that issue.

You weren't using a DBT it appears. Nice.

 

[pa_200]

I Did use a DBT, but......?!?

 

[larryderouin]

There is a BIG Screw up in the MANUAL and it concerns the operation of the TRIMPOTS. Basically the Offset pots are marked as bias pots and vice versa. See this thread for CORRECT PROCEDURE............... https://audiokarma.org/forums/index.php?threads/notice-to-all-new-sx-737-or-7730-owners.464083/

More than likely why bias wouldn't come down from 200mv, you were adjusting the offset, thinking it was bias.