My DIY Sanken SP-40W replacement (on the cheap)

[elnaldo]

Hello. I'm working on a Pioneer SX-650 with burned output ICs, and decided the easiest route is to build discrete replacements for the Sanken SP-40W.

Still a work in progress, I had a bag with PCBs inherited from some friend's father, and started to sketch a layout.

Drivers are BD139 BD140, outputs are MJE15032 MJE15033, bias diode is 3x 1n4148 . I'll add a variable resistor in series with the diode to adjust the bias.

Emitter resistors are metal film 2 W (5W ceramic are too large to fit)

Some picts of the "design" work:

 

[elnaldo]

One question: I see "R2" and "R3" are not present in this diagram. I don't see them inside the IC, I see 5 resistors.

 

[Lavane]

That's odd. Is the first drawing from the schematic in the service manual?
It sure doesn't match the data sheet. I guess one or the other could be an updated version. I have a SX-780 that uses power packs. Some replace them with a discrete modules like the ones your building. If they aren't inside the power pack then the second drawing must be right?

 

[Lavane]

Or, the first drawing IS an updated version. Maybe added for better longevity and stability?

 

[gslikker]

Nice, this will definately work out!


I made an alike solution for my Setton RS440, where I put in TO-3 darlingtons and decided on a transistorized bias circuit. My experience was using the TO-3 was more work than expected because of the metalwork.....

 

[elnaldo]

Surfing the web, found this datasheet with the resistor values.

https://datasheet.octopart.com/S60W-Sanken-datasheet-105005.pdf

"R2" and "R3" are present only in the 60W 80W and 100W versions, they are 330 ohm. I understand they are not installed in the 40W. If it's an extra insurance, I'd fit them, not a big deal to cut a trace and jump 2 resistors in that board.

"R1" is approx 100ohm. I see my 1st idea of using a 150 ohm with a larger trimpot in parallel was not too far from that.

R4 and R5 are 330 ohm +-30% (seems not critical), I've used 220 ohm in my "prototype", before finding this document

 

[gslikker]

I think extra insurance indeed, they slow down the amp also (I GUESS it will be less prone to oscillations since your new transistors have probably much better specs than the old ones on the circuit).

 

[elnaldo]

I think extra insurance indeed, they slow down the amp also (I GUESS it will be less prone to oscillations since your new transistors have probably much better specs than the old ones on the circuit).

I agree, MJE15032 are 30MHz min fT, I was thinking to add a 100pF capacitor from C to E at the drivers, copying this from other amps.

 

[merlynski]

 

[elnaldo]

Ready to connect !

( I wanted to follow resistor color codes for the wiring, but couldn't find all 7 colors needed)

In the next days I'll try it. I'm a bit busy with other stuff, but step by step this project keeps going.









(Edited, I see the connection from Driver transistor Emitter to output transistor base is missing.Also, signal input is miss-wired, I've soldered the wires "after" R2 and R3, right at the transistor base)

 

[Lavane]

I love your work.arounds and hacks. Looking forword to seeing how this one turns out.

 

[elnaldo]

Thanks. I think it will work, but this SX-650 is not in very good shape, I suspect some other problems will develop.

Living down here with limited resources and low budgets wakes up the imagination ! Think that every part you can buy there, costs at least 5 or 6 times for us (or more with recent currency devaluations), so in most of this repairs we need to find a way to keep the budget low, to make the repair possible. That's why you don't see "professional PCBs" for example. To make 2 small boards could cost 1/2 of the SX-650.

I just finished yesterday another SX-650 with a severe HUM problem (unusable) caused by a previous "full restoring" work. Somebody did a full recap using chinese capacitors, and in an effort to make it "better", added wires connected to chassis from every board, even jumper wires interconnecting all the grounds in the same and in different boards, creating a VERY LOUD hum at the output. And to tame that, he added 2 resistors to the bass tone control, to cut the bass and make the hum less evident. A lot of effort to make the unit to sound worst. Fortunately I had this SX-650 in original state, and managed to remove all the added grounds, and restored the original ground wiring. Hum gone, bass powerful again.

 

[Lavane]

Gotta love those mods! Why would someone intentionally add a ground loop? At leased now it won't be as fragile since you are using discrete outputs.

 

[elnaldo]

First tests, only one channel, was a partial success: Bias adjusted perfectly and reacts to the trimmer I have set it to 15mA. DC offset is almost zero.

Without a load, a 1KHz signal looked fine for the first minutes, but then, zooming on the crossover section, I reached to see a notch, a bit of "crossover distortion" that doesn't change when increasing the bias. But the unit has very old capacitors, even one bulging cap in the other channel, and 2 probably leaky Sanyo OSCON capacitors.

Next step: to recap the power amp board (just 6 electrolytics) and the PS board, to replace the 2SA798 differential pair, jump R2 and R3, and try again.

I'm tempted to reduce R4 and R5 (driver emitters) to 220ohm instead of 330. BTW, looking at the SX-550 schematic, discrete output, similar design, I see those resistors are 150 ohm.

The 13V supply to the tuner (a known problem in this Units) is OK, 13.5V going to the tuner board.

The 100pF cap I was thinking to add, is already fit in the power amp board (180pF actually, from PIN 1 to PIN 2 and from PIN 5 to 6)

 

[gslikker]

Maybe play with the R2 and R3 and lower them. Mine are 91 ohms I see on my own picture.

 

[Lavane]

Any progress?

 

[elnaldo]

Finished recapping the PS and power amp boards, and replaced R2 and R3 with 150 ohms. I didn't install the module yet. Next week for sure.

 

[elnaldo]

Cleaning up the layout, I'll build the 2 boards and do the definitive install. I trust it will work at this 2nd try. I don't see a reason why it shouldn't.

 

[elnaldo]

Almost ready to connect and power on for the 1st time !! (I still need to install a new relay in the power amp board, old relay came damaged)

 

[elnaldo]

I had to mount the trimpots at the solder side to fit the boards in the slim slot the SX650 has.

1st try didn't work. I have like 15v at the output but seems to be from a previous stage. I'll Disconnect the preamp and check everything again.