A nice SX-770 was smoke'n from the amp.

zebulon1

Working on my own stuff. Finally
Subscriber
I finely found a free cap coupled set to work on. My first.

A SX-770 with a amp smoking from one side.
The set was on its side so the smoke didn't give a clue to the side, so I am checking both sides of the board.
Power supply voltages check good.
The two amplifier, 2 amp fuses check good.
I removed and checked the outputs. All Good.
Checked the STV-3's. Good.
Lifted one leg and checked all the resistors. Good enough.
Replacing all the E caps.
Checked all the drivers and found only one bad. The Q7, 2SA569.

Does anyone have a good equivalent for this 2SA569? I did a google/ Audiokarma search but nothing really showed up. I'm still looking.
Thanks
 
I found a 990 Thread and a post from avionic with a suitable sub for the 2SA569. He lists a KAS733 as a good equivalent.
 
Would a KSA1220A fit in the space? It's probably overkill, but there's a good chance you have some in the bin.

If that seems excessive a KSA733G is pretty close to the original, with comparable (slightly higher) Ft and power dissipation numbers, and they are plentiful.
 
The 1220's I do have but I still need to order up the caps. Thanks for the conformation. I checked the big 1000uf/25 caps with my ESR checker and found them just below the threshold. I don't have the axial's either. It would be nice to horse it together and see what else is going on for the Mouser order. I might have enough to do just that.
Thanks for the quick reply!
 
Last edited:
I placed the Mouser order for items I didn't have in stock. Like the 1000uf/25v caps. I have gone over the board, checking components and desided to replace the 2SA569's only. Reusing the tested good transistors and replacing all the caps. The 1000uf/25v cap's I ordered are General purpose.

Rebuilt the power supply.

How do you set this little amp? Any links? I did read about checking the DC and the Tuning Fork articles but the Idle settings?
 
Last edited:
VR1/VR2 are offset (L-R) and VR3/VR4 are bias levels (L-R). Setting cap-coupled amps can get a little strange, since the unit probably needs to have a dummy load connected to the speaker terminals and no source/no volume selected. You'll need a dummy load good at 10Ω and at least 1W, more wattage would be safer.

For checking the DC Balance, I would presume the voltage can be read directly at the 2A output protection fuse since it is upstream of the speaker coupling cap. VR1 for the left channel and VR2 for the right. I'd set them as close to 0.00V as possible.

Once offset is adjusted, one method of checking bias might be to measure the voltage across either R31 or R33 in the left channel and R32 or R34 for the right channel. Each of these would carry half the idle current. The voltage reading ÷ the 0.5Ω resistance should indicate half the idle current. In other words, a reading of 10mV should indicate 20mA. I'm guessing 50-80mA would be a limit, but I'm chicken.

Check the DC offset after adjusting bias to make sure it hasn't drifted, since it appears that the bias change may affect the offset, and thus would require multiple adjustments to "zero in" on the desired levels.

EDIT I located some additional comments in THIS POST which indicates that VR1 should be adjusted for 20.2V at the speaker protection fuse to provide proper offset and symmetrical clipping once the amp gets to that point. I don't know under what conditions that is adjusted and cannot verify the concept.

The post goes on to indicate the final amp bias is set as I described, but remarks that the schematic shows 350mA at the resistors, and that seems beyond excessive. Again, the test conditions are unknown, and if this is to be doen at some prescribed signal level I have no clue what that may be.
 
Last edited:
Watthour. With unit's this old, the way I've read on most of the one's that do have instructions was to connect a siggen or adjustable source and use a 1kHz signal. DUT set for AUX, Volume to MAX, and scope on the channel's RCDR output with VOM on the spkr output. Then bring up signal to get a certain voltage at the output, and adjust to get symmetrical clipping on both ends of the sine wave. Not sure if this is what you're looking for tho, or whether PIONEER used this method.
 
I just bought one of these yesterday;
Ebay

A few years ago I made up a 8 ohm dummy load from 8, 100 watt heatsinked resistors. Good for the load!

This is sounding fun!
 
Talk about overthinking the plumbing!!!!!!!!!!!!! :yikes:

The outputs: set one collector to HALF of the other collector's voltage with VR1(VR2) - easy to hit, it's the output transistor's CASE.
You want to play around, double checking for symmetrical clipping, go ahead, when you are done the result won't be much different than that thumbnail.
We are trying to avoid MAJOR imbalances that would reduce output swing by 50%, not less than 1%.

The scat shows 50mA idle to both main amps.
thus 25mA draw per channel.

Rather than fool around trying to figure quiescent draw of the rest of the circuit, set each channel to 25 mA which is 0.0125v across one of it's 0.5 ohm resistors using VR3(VR4).

IF you want to be picky, then measure the total B1 draw current, see how much over the current is in millivolts, divide by 2 then by 2 again, and subtract the millivolts from the starting 12.5 millivolts. It won't be much... Then dial back VR3(VR4) to the new millivolts.

OR

You could dial VR3(VR4) down to NO idle current and measure the B1 draw,
calculate IT'S difference from 50mA in amps, divide by 2 (channels) and by 2 again (0.5 ohms) and set THAT voltage using VR3(VR4).

Remember, the ONLY penalty is a bit too much idle current is some heat when NO audio is coming through.
 
Last edited:
I rebuilt the amp board with new caps and transistors.
Q1, Q2 KSC1845's
Q3, Q4 KSC1815's
Q5, Q6 KSC2690's
Q7, Q8 KSA1220's

Right channel is adjusting to instruction's.

Left channel is showing voltage (22.0v plus) on the E3 (GRD3) ground pin on the amp board.
No voltage is showing on the output.
Scoured the traces for bridges, none found.
I have been every which way but fixed on this problem.
After lifting one at a time, the resistors and caps connected to the ground trace, I realized it was coming from both the +B1 and +B2 supplies.
Now I am thinking the ground needs to be balanced and not knowing how the heck this simple amp works, makes any more troubleshooting a waste of time.

Help. :tears:
 
Last edited:
Pete, you are awesome! :thmbsp:
I was on a wacky adventure starting when I was getting a bright DBT. This was after restoring the board. Your 12 words to put me back on track.
I reconnected my mess and low and behold the board dials in per MTF's instructions. No smoke or bright bulb. When messing with the board I had it out of the set, connected to the power supply with jumpers. So now putting it together is the next objective. :yes:
(The problem was I didn't have the grounds connected, T/S'ing a non-issue) :D
 
What a nice set!!
I first wanted to say "A nice little set" but this 770 packs a sweet punch! Its burning in, pushing some recaped Sansuis SP-2000. Very nice! Quarter throttle is beyond conversation level.
This was my first foray into the cap coupled amp design. A great lesson thanks to the members here on AK.

I have spare SX-770 for parts. It was the donor for hardware and buttons. "Buddy pills" if you will! PM if you need anything other than a transformer and power cord.
I will post it in "Barter Parts" later this week.
 
it appears the ground is broken and allowing the voltage to rise .
I am working on a 770 and when I got it both amp fuses were blown. Somebody seemed to have put in a wrong transistor and that may be it, but I’m not sure how to test the output transistors on the heatsink . What do I test for ?
 
Back
Top Bottom