Replace STKs with rcs16's Design - Technics SA-300

rjd3000

New Member
Hi All,

I'm pretty sure I need to replace both STK-0039s in a beautiful SA-300 I recently acquired. All I get is buzzing on headphones or speakers, but the audio from the tape out sounds great.

I started to suspect the STKs and sure enough the voltages on several of the pins appear to indicate shorts opening the -36v to pins 1, 3 and 8.

I would love to upgrade them using rcs16's design, but I don't know the first thing about submitting designs to a PCB manufacturer and probably wouldn't be able to answer any questions they have. Does anyone have any extra boards or know where I can get one from a place that's made these before?

Thanks!
 
That Receiver is a nice 35 watt RMS unit. But re-engineering the power paks may be an expensive and long procedure for a $150 receiver.
With your statement above, I am not sure what you mean that you have a short "opening the -36v to pins 1, 3, and 8" . If you are saying that you have -36v reading on pin 1, pin 3, and pin 8, those IC's could be shorted, but you need to confirm that you have +36v on pin 9 and -36v on pin 2. If you do, then the chips are probably shorted

But its strange that both IC's have shorted the same way, and usually shorts internal with these IC's can and usually blow the fuse.
I would suggest you check the 2 big power suppy Caps to see if one is open before you but 2 new IC's. This usually shows as less DC voltage (below +36v) but not always. If you are missing the +36v totally and cant see it on either IC at Pin 9 then you have bad wiring or a break in a circuit board trace, possibly by those big capacitors.
 
Hi Tom. Yes, what I meant is I am reading around -36v on pins 1, 3, and 8 of each of the ICs while they're in the unit (I haven't pulled them yet). I am getting the +37v on pin 9 and -36v on pin 2 on both of the ICs. I checked the voltages coming off of the big caps and I get -36.65v coming off of C701 and +36.4v coming off of C702. I will say that both of those caps do look a little suspect despite the good readings. There is some residue under them that may be a sign of leakage, so those will probably get changed anyway as well.

I do think it's very strange that both ICs appear to have failed in the exact same way. I initially thought it was an issue with the power supply because of that, but I started looking at the ICs after noticing that I get nice audio from the tape out and measuring the voltages from the big caps.

As far as the cost goes, I picked it up for $40. If it does end up being bad ICs, I don't mind putting a little extra time and money into this to get it working as good or better than new given how great of condition it's in.
 
Check with RCS16. If the boards will work on the STK0039's (and I think they will), Zebulon1 has the heatsinks and bareboards. You supply the parts per the list on the STK-0050 thread.
 
Be careful all.
The STK0050 is Pure Complimentary while the STK0039 is Quasi Complimentary. I am not sure if it will work properly. There are different components and connection internal to these IC's. This may require a different layout with the replacement boards/components.
 

Attachments

  • STK multis.pdf
    149.9 KB · Views: 31
Which is why I suggested contact RCS16.
Be careful all.
The STK0050 is Pure Complimentary while the STK0039 is Quasi Complimentary. I am not sure if it will work properly. There are different components and connection internal to these IC's. This may require a different layout with the replacement boards/components.
 
In this thread there is a successful placement of the custom modules in a Pioneer SX-680, which originally had STK0039 modules:
http://audiokarma.org/forums/index....x-780-and-others.721181/page-30#post-12360948
Check it out
I looked at the schematic of the Technics SA-300 and it shows a cap connected to pin 5 of its original module. The custom module does not use pin 5 at all. Also the datasheet I have for the Sanyo STK modules equivalent circuit does not show a pin 5 connection either. I did notice that the Technics' schematic shows the module to be SVISTK0039U, which may be something different than the Sanyo STK0039. The Service Manual does show an internal connection on the equivalent circuit module for pin 5.
 
Hmmm... Thanks for pointing that out merlynski. In the SA-300 there is indeed a cap connected to pin 5, and the other side of that cap connects to pin 1 and everything that is connected to. Info on the STK0039U variant seems to be scarce. The SX-680 schematic doesn't seem to have anything on pin 5, so that is a bit discouraging. Looks like more research is in order...
 
Hmmm... Thanks for pointing that out merlynski. In the SA-300 there is indeed a cap connected to pin 5, and the other side of that cap connects to pin 1 and everything that is connected to. Info on the STK0039U variant seems to be scarce. The SX-680 schematic doesn't seem to have anything on pin 5, so that is a bit discouraging. Looks like more research is in order...
Personally I think it would still work, the biasing voltages look much the same, and that cap provides a bit of negative feedback for high freq stabilization.
If you did try it, at worst you would be out $75-$100 for all the parts and some time. You could probably recoup a large part of your parts cost by reselling the completed modules to someone, if it really did not work. And if it did work it would be a good benefit to others who also have SA-300s with blown modules.
:dunno:
FWIW, I have used the RCS16/hbrown modules in a Pioneer SX-780 with great success.
 
If you search the bay for STK0050 you will get a listing (among others) that has the module circuit board, heatsink plates and transistor mounting screws in one kit. It also lists the BOM for the components from Mouser and a couple other sources. You buy your own components. That listing is from one of the AK members who helped develop this project, and is very affordable. It does take some soldering skill and a bit of mechanical ability to do but is really not very difficult.
 
For reference. (from SA-300 SM) My 2 cents - the cap is probably to improve stability of the Quasi complimentary circuit, and the full complimentary circuit shouldn't need it. In fact, the 0050 or replacement supply would be a better choice than a plain stk0039. (My opinion - YMMV)
upload_2019-1-25_22-20-26.png
 
For reference. (from SA-300 SM) My 2 cents - the cap is probably to improve stability of the Quasi complimentary circuit, and the full complimentary circuit shouldn't need it. In fact, the 0050 or replacement supply would be a better choice than a plain stk0039. (My opinion - YMMV)
The same thoughts went through my head :beerchug:
 
Bit of an update, I replaced the STK0039Us with STK0039s from Ebay and between that and a whole lot of De-oxit I was able to get both channels up and running and the receiver sounding quite nice. I am getting some DC bias on the L and R speaker outputs that is higher than I'd like (about 400mv), but it doesn't seem to affect the sound a ton.

Interesting note, the original STK0039Us must have both failed in different ways because different pins were shorted on the L and R units. Go figure.

I'm still going to attempt the upgrade to the STK0050s. The boards and components just came in, so once they're assembled I'll give them a go and report back on how it works.
 
Bit of an update, I replaced the STK0039Us with STK0039s from Ebay and between that and a whole lot of De-oxit I was able to get both channels up and running and the receiver sounding quite nice. I am getting some DC bias on the L and R speaker outputs that is higher than I'd like (about 400mv), but it doesn't seem to affect the sound a ton.

Glad to hear you have it working. Just be wary of those aftermarket STK modules. I won't use them anymore as a few months back l replaced a bad STK with a brand new aftermarket part, the DC offset was perfect after fitment and l ran unit for about 3 hours before it must have had a massive DC spike in one channel and took out a woofer in a pair of my favourite speakers :(.
 
Another update: I was able to confirm that the DIY STK-0050s do work in place of the STK-0039Us. However, I am still getting a lot of DC bias (around 320-380mv) on both channels, so I don't want to connect any good speakers to the receiver yet. The bias is way more than the variable resistors can compensate for.

I noticed the voltages coming off of the power supply are a bit higher than spec, so I checked all of the pins and here is what I got:

IC601
1: -.68v (-1.3v)
2: -38.3 (-36.5v)
3: .39v (0v)

8: .39v (0v)
9: 38.3 (36.5v)
10: 1.7 (1.3v)

IC602
1. -.58 (-1.3v)
2. -38.1 (-36.5v)
3. .326 (0)

8. .326 (0)
9. 38.1 (36.5v)
10. 1.61 (1.3v)

The voltages on pins 1 seem to be pretty far off, and the voltages on pins 3 and 8 seem to match the DC bias I'm seeing on the speaker terminals. Also of note, the dc bias difference between the two is probably due to my playing with the variable resistors to try to get the bias down. Otherwise I believe they'd match pretty much exactly.

Any suggestions as to what to do to get the DC bias down? And should I be concerned about the voltages coming off of the power supply?
 
Another update: I was able to confirm that the DIY STK-0050s do work in place of the STK-0039Us. However, I am still getting a lot of DC bias (around 320-380mv) on both channels, so I don't want to connect any good speakers to the receiver yet. The bias is way more than the variable resistors can compensate for.

I noticed the voltages coming off of the power supply are a bit higher than spec, so I checked all of the pins and here is what I got:

IC601
1: -.68v (-1.3v)
2: -38.3 (-36.5v)
3: .39v (0v)

8: .39v (0v)
9: 38.3 (36.5v)
10: 1.7 (1.3v)

IC602
1. -.58 (-1.3v)
2. -38.1 (-36.5v)
3. .326 (0)

8. .326 (0)
9. 38.1 (36.5v)
10. 1.61 (1.3v)

The voltages on pins 1 seem to be pretty far off, and the voltages on pins 3 and 8 seem to match the DC bias I'm seeing on the speaker terminals. Also of note, the dc bias difference between the two is probably due to my playing with the variable resistors to try to get the bias down. Otherwise I believe they'd match pretty much exactly.

Any suggestions as to what to do to get the DC bias down? And should I be concerned about the voltages coming off of the power supply?

For the power supplies, If you are in the US, the mains voltage is a bit higher than it was when amps of this age were designed, so a couple of volts higher unregulated voltages is to be expected. As they seem to be well balanced I think they are ok.

You have mixed up your terms in this post by saying "DC bias" Bias is the idle current flowing positive supply to the negative supply through the modules' output transistors and emitter resistors with no input signal, and is measured by reading the small dc voltage across the emitter resistors on module pins 3 and 8 and calculated by using Ohm's Law. If, as your readings show, there is no voltage across pins 3 and 8 there is no bias current flowing. This should be adjustable by the variable resistor on the new STK-0050 module. If you adjust it too far (high) you may fry the output transistors, be careful. Adjusting the bias should not greatly affect the DC offset. DC offset is the dc voltage measured on the output of the amp referenced to ground. There does not appear to be any way to adjust DC offset on this amp. The dc offset is affected by the balance of the differential amp formed by Q601 (Left channel) and Q602 (Right channel) through the negative feedback path from the junction of the emitter resistors to the right side base of the diff amp. For Q601 that is R615 and R617. Are the bases of Q601 and Q602 0.0vdc?
Check that the cathode of zener diode D603 is +14.0 vdc, and the anode is 0.0 vdc referenced to ground.
 
Merylnski, thank you for the clear explanation. I was indeed getting the concepts of DC Offset and Bias mixed up. This is very helpful.

Measurements at D603 seem fine as I am getting 14.6vdc on the cathode and 0vdc on the anode.

However the bases of Q601 and 602 are not 0vdc. B1 and B2 for Q601 are both right around 470mV and Q602 B1 is at 332mV and B2 is at 388mV. So it looks like you may be onto something. What do you think could be causing voltage on the bases of both Q601 and Q602??
 
Back
Top Bottom