Yamaha SY-22 synth issue

[frisc0]

Hi, I could use some help with an issue with my Yamaha SY-22 synthesizer. Recently plugged it in again after some years of storage, used a Yamaha 12V adapter (the synth accepts 9-12V), it shortly produced distorted sound but was silent then. No audio anymore, but everything else still working, display, controls, etc.

I've checked the schematics looking at the power circuit, from the 12V input there is a 5V IC regulated power line and a ~9V power line produced by a zener diode with a transistor. Did some reading on the latter and I understand the transistor in this circuit is to allow for supplying higher currents than the zener would be able to handle. I've did some measuring of voltages here and instead of ~9V (or 8.4V since the circuit decreases the voltage a little as I understand) I measure roughly 3.4V from the emitter of the transistor TR9. First I thought the zener would be defunct but measuring it out of circuit (with a power source and a resister in series) it produces somewhere near 8.6V so looks like it's ok. Replaced the 2 electrolytic caps, no effect.
Could it be the transistor that is defunct? Or should I look for issues further down the line, in the circuits that are using the ~9V power line (+BA in the circuit)?

Btw: I've checked the nominal voltage of the Yamaha PSU which reads 16V without load... that could have caused this issue I assume. Now using a 12V PSU that delivers 12V without load.

Hope someone can point me in the right direction finding the cause of the issue... thanks!

https://www.synthxl.com/wp-content/uploads/2017/12/yamaha_sy22_synthesizer_service_manual.pdf

[Audiorepair]

To the north of IC5 you have a circuit powered by BA that provides battery backup voltage.
Check ZD2 or the cap after D3 isn't shorted, as that will short BA to ground via 470 ohms.

[Vicus]

Also, in some cases you have to reload the "sound banks" if battery dry and no longer maintain the data in the memory. Indon't know if it's the case.

[frisc0]

To the north of IC5 you have a circuit powered by BA that provides battery backup voltage.
Check ZD2 or the cap after D3 isn't shorted, as that will short BA to ground via 470 ohms.

Thanks @Audiorepair, I've checked both the cap and the Zener. The cap (0.47F, 5,5V) when measured outside the circuit with an ohm meter, looks shorted in one direction, and in the other direction shows ~2.7 volts, probably after being charged by the ohm meter, so looks okay?

ZD2 is a 5.6V zener which outside the circuit, in series with a resistor on a power source measures 4.6V one direction and 0.7V the other direction, is this okay? Not sure if 4.6V is too low but it does not look shorted at least.

What can be concluded from the above?

Btw the SY-22 cannot be loaded with patches and does not have a battery to retain patches (only that big 0.47F cap to retain memory as long as the PSU is plugged in.

[Audiorepair]

To the north of IC5 you have a circuit powered by BA that provides battery backup voltage.
Check ZD2 or the cap after D3 isn't shorted, as that will short BA to ground via 470 ohms.

Thanks @Audiorepair, I've checked both the cap and the Zener. The cap (0.47F, 5,5V) when measured outside the circuit with an ohm meter, looks shorted in one direction, and in the other direction shows ~2.7 volts, probably after being charged by the ohm meter, so looks okay?

Unless I'm misunderstanding what you are saying, that sounds very much NOT ok. 
Measuring a cap on ohms setting should give no reading at all in any direction.

When Zeners go bad they very nearly always just go short, you can easily find that scenario without taking them off the board.

4.6v does sound a little low, but they do measure low at low currents anyway.  0.7v forward voltage is bang on what they tend to measure in diode mode on a multimeter.  (rectifier diodes measure less than that)

[frisc0]

Unless I'm misunderstanding what you are saying, that sounds very much NOT ok. 
Measuring a cap on ohms setting should give no reading at all in any direction.

Sorry I wasn't clear, I first measured on Ohms setting, measuring a short in one direction, few megs in the other. When changing setting to Volts I measured the 2.7V, so it charges, but it's faulty as you say . I've put in a substitute cap (470uF, 16V) and put back the zener, but to no avail, still no sound, still 3.4V output from +BA.

So probably there is another place in the circuit pulling the voltage down. There are lots of locations where +BA is used, many (op)amps for example. What are typical components that would fail after having had a high input voltage?

There are some caps having low voltage limits (10V), I guess the ones that are connected to/near ground are suspect, like the one near D13?

[Audiorepair]

If that 470uF cap has failed, it is likely all other 470uF caps of the same type have failed, or will shortly do so.

That doesn't mean to say ALL the electrolytics are bad, it is usually one type that will fail first, so you might want to replace all those types, and see how you go from there.

[Audiorepair]

Unless I'm misunderstanding what you are saying, that sounds very much NOT ok. 
Measuring a cap on ohms setting should give no reading at all in any direction.

So probably there is another place in the circuit pulling the voltage down. There are lots of locations where +BA is used, many (op)amps for example. What are typical components that would fail after having had a high input voltage?

Yes, many op amps share that voltage. 
A thermal camera can be very useful in determining the one op amp that is faulty, otherwise a finger might also home in on the culprit getting a bit warmer than it should be.
There are also videos on YouTube showing how to locate hot components by smearing with IPA alcohol and watching how quick it evaporates.

Point being, if if something is pulling down that voltage, it should be getting hot.


Edit: well that isn't strictly true - if, say, the Zener was short it would be the 470 ohm resistor that would get hot, but it would at least point you to the problem.

[frisc0]

I've been doing some more digging, and tried to follow the 8V line that is faulty. Btw, the 5V circuit functions fine and all TTL logic that is fed from the regulated 5V is working fine, the synth starts up normally. However, the 8V line gets cut in half to 4V somewhere down the line...

Not sure what the function of the part of the circuit around IC19 is btw, it receives 5V from the left end, and 8V between IC19 and IC35 (which measures 4V at pin 9 of IC35, half of what it should be), see attachment.

Nothing gets hot/warm except for the power amp and the 5V regulator IC.

I've unplugged the aftertouch circuit (since the 8V line is also fed there) but to no avail.

The voltage divider (attached pic 2) does it's work nicely and cuts the 3.3V input in half to 1.6/1.7V (which should be 4V).

Any thoughts?

[frisc0]

I've seen reports of other SY22 users that TR4 that drives the output lines relay sometimes burns out. Could this cause a drop of +8V voltage to half?
See attached. How to test this transistor?

[Audiorepair]

I've seen reports of other SY22 users that TR4 that drives the output lines relay sometimes burns out. Could this cause a drop of +8V voltage to half?
See attached. How to test this transistor?

The transistor is used as a switch.  When on, C-E is for all intents and purposes shorted.  (leaving around 0.2v across it, no sot an exact short)
So if this transistor is shorted, it can only do what it normally does, but slightly more shorted.

If D6 across the relay is shorted, (unlikely in reality but worth a check), then BA would be shorted to ground via the 100 ohms when the relay is switched, or T4 shorted.

[frisc0]

Tested D6 ok, T4 is not shorted.

I found something else that I could not find on the circuit diagram, a diode placed over a capacitor, see attached pic. Looks like something added later, but I see it in other pics of the SY22 PCB as well.

I wish I could 'cut' the circuit 8V line, moving my way from the PSU down to where ever the issue is, but no idea how to do that properly. Any ideas/tips?

Would love to get this thing working again...!

[elecdonia]

However, the 8V line gets cut in half to 4V somewhere down the line...

This is a good start. Next:

1. Identify every load which receives its power from the 8V rail
2. Determine what the expected total load current should be for the 8V rail
3. There are 2 likely faults, but only 1 of these is causing the problem:
      A. The 8V regulation circuitry has failed. In this case the output will stay at 4V even after disconnecting all loads from the 8V rail.
      B. One of the loads is drawing excessive current, thus causing the 8V regulation circuitry to enter a current-limited state.
4. Is it feasible to isolate all loads from the 8V regulation circuitry? Then measure the output voltage without any load. If unloaded voltage is considerably less than 8V then the fault is likely to be inside the 8V regulation circuitry.
5. If unloaded voltage is 8V then measure the current drawn by the 8V rail when connected. There is likely a much higher current than expected. This is what drops the 8V to 4V.

I’ll take a look at the SM and schematic and provide more suggestions. I haven’t personally worked on this model but I do have plenty of experience troubleshooting other Yamaha products.

[wasedadoc]

What is the voltage on TR9 base?

[frisc0]

I'm on holiday for 2wks so not able to check right away, but I believe it was around 2.2V if I remember correctly; will check once back!

[elecdonia]

I've seen reports of other SY22 users that TR4 that drives the output lines relay sometimes burns out. Could this cause a drop of +8V voltage to half?
See attached. How to test this transistor?

TR4 won’t directly affect the 8V power supply rail.
If TR4 is open circuit then the relay will never pull in, with the result being that there won’t be any audio signals at the line output jacks. If TR4 is shorted then the relay will stay pulled in all the time.
However, the relay itself is powered by the 8V rail. But if there is only 4V there, it is unlikely that the relay will pull in regardless of whether TR4 is on or off.

You can test TR4 in-circuit with a multimeter which has “diode-check” feature:
First disconnect  the external 12V power supply
E-B should measure about 0.5V with positive test probe on B
C-B should also measure ~0.5V with positive test probe on B
E-C should not measure 0.0V (but you might get a reading which isn’t “open circuit”)

[elecdonia]

Let’s focus on TR9 (2SA881). This transistor regulates the 8V rail.

If there is an excessive load on the 8V rail then TR9 will become quite warm or too hot to touch soon after applying power. According to the user manual the total power consumption of the SY22 is <8W, which corresponds to drawing <600mA total from the 12V power supply. It is safe to assume that most of this power is used by the 5V rail which powers all of the logic chips. I estimate the total current flowing through TR9 should be <100mA. With a 4V drop from C-E this gives TR9 a power dissipation of ~400mW. This will make Tr9 "warm," but not "too hot to touch."

The circuit around TR9 is simple. TR9 acts as an “emitter follower.” Its base voltage is supplied by 9V zener ZD1. The emitter will be 0.6V lower than base, ~8.4V.

To diagnose this circuit:

First measure and post these 3 voltages at TR9, measured with - probe connected to circuit ground:
C               (Should be 12V or same voltage as external power supply)
B               (Should be 9V)
E               (Should be between 8V and 8.5V)

Next, with external power supply disconnected, measure the in-circuit resistance of the 2 resistors connected to cathode of zener diode ZD1. One resistor should be 270 ohms, the other resistor 100 ohms.

To test TR9 in-circuit with multimeter having “diode-check” feature:
First disconnect external 12V power supply
E-B should measure about 0.5V with positive test probe on B
C-B should also measure ~0.5V with positive test probe on B
E-C should not measure 0.0V (but you might get a reading which isn’t “open circuit”)

If you have the capability to unsolder and remove TR9 from the PC board then some additional tests can be done.
First take a photo or make a drawing to help remember the correct orientation for reinstalling TR9.

Please post photos of TR9 on the PC board prior to removing it.

To desolder without damaging PC board I recommend brushing a small amount rosin flux on the 3 PC board pads for TR9. Also dip your desoldering wick in the rosin flux (a tip I learned only recently - even though desoldering braid already contains some flux, it helps greatly to add more flux). Put the desoldering wick against the pad and heat with soldering iron until all the solder flows into the desoldering wick. Remove all of the solder so that TR9 can be moved easily. The goal is to be able to gently pull TR9 out without damaging the PC board.

With TR9 removed it can again be tested with multimeter “diode-check.” It can also be tested with one of these “$20 transistor tester” devices:
     https://www.eevblog.com/forum/testgear/$20-lcr-esr-transistor-checker-project/

Before reinstalling TR9 I recommend applying power and measuring the voltage across zener diode ZD1. This should measure about 9V. Then check voltage from circuit ground to the B pad for TR9. The measured voltage should be exactly the same as the voltage you previously measured across ZD1 (9V).

[elecdonia]

Yes, many op amps share that voltage. 
…a finger might also home in on the culprit getting a bit warmer than it should be.

Point being, if if something is pulling down that voltage, it should be getting hot.

The opamps should not get above room temperature. IC14 (D-A converter) and IC31 (VCA) may operate somewhat above room temperature, but should not get “hot” to a level where you wouldn’t want to keep touching them.

[Audiorepair]

I can assure you that any working op-amp WILL be above room temperature, any thermal imaging camera will show you that, and is a very useful tool for searching out the duff op-amp which may well be significantly warmer than the others.

A finger or some alcohol would do in the old days, and could still actually be used today.

[frisc0]

Great stuff! Thanks @Elecdonia and @Audiorepair for taking the time to check the schematics and help diagnose this board! Valuable explanation and hints to further investigate which I will do once back from my holiday; can't wait to check..!

[frisc0]

Back from holiday, continued the SY22 project...

Many thanks for the clear instructions@Elecdonia! I followed them in order, with the following results:

TR9 in circuit, with power connected:

C = 11,21V
B = 4,02V
E = 3,31V

Resistors: 100ohm and 270ohm as it should

TR9 in circuit, without power, diode check on multimeter, with positive test probe on B:

E-B: 300mV
C-B: 300mV
E-C: 529mV other way around: 555mV

I removed TR9 (great tip with the flux on the desoldering wick!) and without TR9 installed, ZD1 = 8.7V and so is the B pad of TR9!

Can we conclude from this that TR9 is faulty?

Btw, there are misalignments in schematics, PCB printout and actual components installed regarding TR9.
TR9 is a D880 transistor, on the PCB '2SD890' is printed and according to the schematics TR9 is a 2SA881Q...

What should I get as a replacement? I assume a D880 or substitute?

Thanks for your suggestions!

[frisc0]

I can't find the D880 locally, but the BD241 should be equivalent, could I use this one as a replacement? The one I found locally is a BD241C, don't know if that makes a huge difference?

[Vicus]

I don't think it tr9. It behaves as it should. At the emitter there il the voltage at the base minus the Vbe of tr9. Now the question is why the voltage at the base of tr9 is so low?

Well maybe tr9 is the colprite. Consider it not a critical part. I think that any transistor with enough  IC max and the same package should work.

[helius]

A 0.47F 5.5V component is a supercapacitor, and behaves very differently from a capacitor. It is used more like a battery to preserve memory settings. You can't substitute it with some other capacitor. It also has nothing to do with the power supply, it is used only for standby current to power SRAM and nothing else.

[elecdonia]

I agree that TR9 is quite likely to be faulty. It isn’t providing enough current gain. Perhaps none at all. As a result the zener diode voltage drops from 8.7V to 4V. This shouldn’t happen. With a good transistor the voltage across the zener won’t change when the transistor is in circuit vs. out circuit. The base current drawn by a good transistor should be quite small compared to the current flowing through the zener.

Those E-B and B-C diode drop voltages of only 300mV could indicate a damaged transistor. However this measurement should be repeated with the transistor out of circuit to confirm. Ideally testing it out of circuit with one of the “$20 transistor testers” would confirm whether it still has sufficient gain (HFE).

I suggest rigging up some clip leads so that you can experiment with different transistors in place of the original one. This circuit doesn’t require an exact replacement transistor. Any NPN transistor in a TO220 package and with roughly similar specs should work fine.