Kawai CA 65 digital piano amplifier

[shockpoint]

Hi, just wondering if I can get some help with this diagram. The piano symptoms were as follows:
- suddenly not able to hear sound from either the headphones or the speakers (Not related to the Jacks in the following schematic, same circuit board but different jacks)
- digital screen still works
- MIDI output still works
- Fuses to amplifier board blown (which is the board shown here)

I've tested everything across the board and I've come across an oddity that I cannot get my head around. There seems to be a short across a PAIR of capacitors, and I cannot figure out whether this is abnormal or normal.

In the attached schematic, the two red circled capacitors C59 and C60 read as a dead short, 0.001ohms, and the value doesn't change over 5 seconds either, so I don't think it's charging. They both short to each other as well.
I have also taken a photo with some markup to give some physical context - as you can see there is a burn mark circled.
I have marked the pins of the Line In 1/4" jack, and the pins are: https://www.synth-parts.com/media/image/ac/39/07/YKB21-5074-schematic.jpg
As you can see, the "burn mark" straddles the trace between pin T of JA4 and Capacitor 61.

The op-amp is a dual op amp NJM4580L, and I have marked the pins 1-8. The datasheet is here: https://pdf1.alldatasheet.com/datasheet-pdf/view/7454/NJRC/NJM4580L.html

In the third image, you can see the way the 12V and -12V rails are connected on the schematic, all the way to the transformer. Fuses (slow blow) F5 and F4 were found to be blown.

Questions:
1. Can someone please have a look and explain to me how this might cause an overload on F5 and F4? (I tested the bridge rectifier and it was working fine).
I can't figure out how shorting out C59 and C60 would result in an overload; the positive terminals of each of the capacitors connect to a jack INPUT, moreover this terminal does not cause the (-) terminal of the op-amp to short to ground.

2. How is it both capacitors C59 and C60 short TOGETHER? How likely does this occur? I'm not very experienced with circuitry so I don't know how capacitors in this configuration behave but I presume that the likelihood of both failing together is low.
The only thing I can think of is that pin T of JA4 is connected to Pin B of JA3, so an overvoltage would blow both C59 and C60 together.
Incidentally if you look at the circled "?" trace of JA3, you can see a bit of a bubbliness under the PCB top layer suggestive of high temperature exposure, albeit less severe than the JA4 one.

3. What could have caused this issue? What would have failed first in the schematic, in your opinion to have led to the cascade of events and then the symptoms?
**EDIT - I was looking over the schematics again, and I guess with how the op-amp negative feedback resistor R44 is tied to the negative input terminal, there must be a parasitic drain from the +/-12V rails across R44, R41 and C61 to GND?
Noting that C61 is a polarity specific capacitor, how could this current occur?

Impression
- So my impression is that the short caused an excessive draw from the 12V or -12V rails to ground, resulting in overcurrent in fuses F4 and F5 and causing them to blow. It isn't an outright short circuit but more of an overload.

Would someone be able to provide insight, because I think I understand what's going on but I can't fully explain from cause to symptoms

Thank you!

Update 1: Turns out the jack has a clip that when connected, decouples the short, so that was a wild goose chase. I have now turned my attention to the power board - there are two voltage regulators on it - 78M12 and 79M12, which are positive and negative voltage regulators respectively.
- Is there a difference in the way the positive regulators and negative regulators are constructed? I see a very different voltage drop using diode mode on the 78M12 output->GND; This might be measuring diode D27, but my voltage drop is 0.126V, while it is 0.600ish volts for the Output->GND of the positive regulator. Alternatively maybe D27 is a mess and D26 is functioning correctly.

Update 2: I found a capacitor that seems to be stuck in open circuit (doesn't charge up). This is capacitor C12 on the KEP 420 Full schematic. The Capacitor links Analog GND to Frame Ground (FG), which is a heat sink for the voltage regulators and the big bridge rectifier. Would this capacitor being in open cause fuses F4 and F5 to blow? Is there a reason for this capacitor not to behave like a capacitor?

Update 3: I have gone back to test the NJM V+ and V- rails as well and in particular the opamp IC responsible for putting audio out to the speaker and the headphones read like this on a quick and dirty diode mode test in-circuit:
- red lead on V- and black lead cycles through all the pins Ainput+, Ainput-, Aoutput, Binput+,Binput-,BOutput
- the V- and Aoutput, V- and Boutput in diode mode read: 0.827V
- every other NJM4580 (identical biasing circuitry) is reading 1.087V apart from the NJM4556AD (which I have nothing to compare against)

I suspect there is an issue here; would you be able to shed light on what diode mode would be reading between V- and Aoutput given the NJM4580 has the attached circuitry?
- attached datasheet only shows the circuit for half the dual op-amp

[Andy Watson]

They are grounded via the input chokes, L4 and L5, then via the switch contacts on the input jacks. So they will measure short-circuit unless you plug something into the input jacks.

Edit: Oops, that was the answer to your second question. I don't think the caps are shorted - and even if they were, that would not cause the problem with the blown fuses.

The Blown fuses are due to an overload. It would appear that the power supply section can be un-plugged (plug/socket CN3) from the rest of device. I would start by isolating the power supply and bringing that to a functiong state.

What fault-find facilities do you have? Meter ? Bench power supply ?

[shockpoint]

Thanks! I'll try it out now and see if the short circuit goes away.

[shockpoint]

They are grounded via the input chokes, L4 and L5, then via the switch contacts on the input jacks. So they will measure short-circuit unless you plug something into the input jacks.

Thanks for the help! They did go away with a 1/4" plug inserted.

What a bungle! I thought I found the cause of the symptoms!

[shockpoint]

They are grounded via the input chokes, L4 and L5, then via the switch contacts on the input jacks. So they will measure short-circuit unless you plug something into the input jacks.

Edit: Oops, that was the answer to your second question. I don't think the caps are shorted - and even if they were, that would not cause the problem with the blown fuses.

The Blown fuses are due to an overload. It would appear that the power supply section can be un-plugged (plug/socket CN3) from the rest of device. I would start by isolating the power supply and bringing that to a functiong state.

What fault-find facilities do you have? Meter ? Bench power supply ?

Yes that's correct, I have isolated the PSU board and I am waiting for the fuses to arrive in the mail.

I have a couple of meters, oscilloscope and also a bench power supply (although I don't believe it is current limited, only the voltage is changeable).

I also have an ESR meter (Bob Parker design)

I had a question regarding the power supply board schematic - if a pair of fuses like F4 and F5 fail, is the problem typically overcurrent upstream or downstream?
- I checked the bridge rectifier and it seems to be conducting normally in diode mode testing, appropriate forward voltages across the pins.
- I'm not quite sure how to check the transformer, which happens to be a KT-033, a proprietary Kawai Transformer, although I'd imagine the design would be similar to other transformers. The AC lines feeding F4 and F5 seem to be an independent secondary winding separated from the other secondary windings. I'll double check for any shorts across the AC lines, but none of the other circuitry supplied by the other secondary windings have any issues, just the circuitry supplying the amplifier board as shown.
- I heard that the trasnformer can be a weak point in these but theres not a lot of documentation on how they actually fail!

[magic]

If both fuses blew then it may have been a short between the ±12V rails. In such case the most likely culprits are the NJM4580 chips, unless this PSU also supplies other circuits not shown above.

Before the fuses arrive, see if there is a short between the ±12V rails or between any rail and GND. If so, disconnect the boards and check which one is shorted, then disconnect individual components and so on.

[shockpoint]

Hi, thanks for your input

I did that but could not find any shorts. There are 5 NJM4580Ls on the board, with the same arrangement of +-12V rails and ground.
I used the resistance mode, not the beep mode, and most of the resistances read high or in the kiloohm range.

Is it possible for a circuit component to have an abnormally low resistance that doesn't look like a dead short to sink enough current to ground to blow a fuse? I have only been looking for dead shorts using resistance mode and also cursory checks on the capacitors to check for any odd behaviours.

The PSU supplies other things, but the fuses that blew are limited to the NJM4580L circuitry. Other fuses on the same power board (to the digital piano mainboard and speakers) remain intact.

I will run another check for 12V rail shorts and see if there is abnormally low resistance. I've been looking for dead shorts on the order of 0.001-3 ohms

[Audiorepair]

The op-amps are supplied by the regulators.  If you short the regulator outputs, they will current limit/shut down/protect etc.
This may or may not draw enough power that the fuses blow, but it is quite possible the fuses will survive under these conditions.

So it may actually be a short on the V++ line to the amplifier that took out the fuses.

[shockpoint]

The op-amps are supplied by the regulators.  If you short the regulator outputs, they will current limit/shut down/protect etc.
This may or may not draw enough power that the fuses blow, but it is quite possible the fuses will survive under these conditions.

So it may actually be a short on the V++ line to the amplifier that took out the fuses.

Thanks for your input.

I'm now checking the board that supplies the op amp board. The power board has 2 regulators that go to the op amp board. I'm not able to demonstrate a short circuit on either the negative nor the positive regulator output to its respective GND pin. Each of these pins supply the +12V and -12V respectively.
V++ actually goes to -12V depending on whether a headphone mute button is pressed.
I'll investigate the mute signal circuit and see if any transistors are blown. I remember testing them but none of them were.
Do you think an overload on this would cause both the fuses to blow? Or just one?

[Audiorepair]

Do you think an overload on this would cause both the fuses to blow? Or just one?

Both rails and VCC+ are supplied via both fuses via the 4 diodes in the Bridge Rectifier. 
If there was a short, one fuse would blow first, the other would still be supplying both rails via the 2 diodes still in circuit, and would blow shortly after.

If you wanted to experiment, you could always deliberately short out the 2 regulator outputs, and see if it blows the fuses or not.
You shouldn't damage anything, the regulators will just do what they are designed to do and shut down.
But they may still get quite hot and draw considerable power.  Dunno, never tried this myself.

[shockpoint]

Do you think an overload on this would cause both the fuses to blow? Or just one?

Both rails and VCC+ are supplied via both fuses via the 4 diodes in the Bridge Rectifier. 
If there was a short, one fuse would blow first, the other would still be supplying both rails via the 2 diodes still in circuit, and would blow shortly after.

If you wanted to experiment, you could always deliberately short out the 2 regulator outputs, and see if it blows the fuses or not.
You shouldn't damage anything, the regulators will just do what they are designed to do and shut down.
But they may still get quite hot and draw considerable power.  Dunno, never tried this myself.

I see, OK. I'll give that a go when my dim light tester is completed, I don't want to blow anything unnecessarily!
I went and tested all the 12+ and 12- rails to all the components and again there were no obvious shorts.

I read your line about the bridge rectifier again, and I see your point. Have you ever had any experience of slow blow fuses blowing out of nowhere in isolation to another issue/short in the board?

Some updates:
- I found that the voltage regulator GND pins do not ground to the heat sink. So Pin 2 and heat sink have no continuity. However, other parts of the circuit including the ANALOG GROUND from the transformer DO in fact ground to the heat sink
- There is a capacitor there (as you can see in diagram 4 in the OP, marked C12); I highly suspect this is blown based on a quick and dirty test - it's not charging up from the multimeter leads voltage, just reads OL.
- However I don't think this would contribute to a short circuit that would blow a fuse, but I'm not that experienced. What do you think?
- Photos of the components are attached

I have gone back to test the NJM V+ and V- rails as well and in particular the opamp IC responsible for putting audio out to the speaker and the headphones read like this on a quick and dirty diode mode test in-circuit:
- red lead on V- and black lead cycles through all the pins Ainput+, Ainput-, Aoutput, Binput+,Binput-,BOutput
- the V- and Aoutput, V- and Boutput in diode mode read: 0.827V
- every other NJM4580 (identical biasing circuitry) is reading 1.087V apart from the NJM4556AD (which I have nothing to compare against)

I suspect there is an issue here; would you be able to shed light on what diode mode would be reading between V- and Aoutput given the NJM4580 has the attached circuitry?
- attached datasheet only shows the circuit for half the dual op-amp

[shockpoint]

An update!

I think I may have found the problem. It was not located on the amplifier board after all!

After some systematic connection of wires, I believe when my amplifier board is connected to the mother board, the -12V and the +12 rails are connected to a 125ohm resistance (I am still trying to find out where exactly this is).
On the mainboard there are 5 op amps of the LM833 variety. I believe these are connected in parallel based on DMM readings. You can see the schematic (but there is no circuit diagram, only an assembly diagram unfortunately. Kawai is a bit stingy)
I have circled the area on the board that is powered by the -12V,+12V rails
When the entire keyboard circuitry is connected together, there is a 125ohm resistance between -12V and +12V.
It is at this point when the power is turned on the fuse blows again

So I replaced the fuse, depinned the connector to the motherboard that had the -12v and +12V rails and what do you know, the fuses stayed intact

Based on the opamps on the amplifier board also running -12v and +12v NJM4580L, but NOT having any 125ohm resistance, I can only deduce that the 125ohms possibly represents an abnormality, or a short circuit. This certainly is in line with the fact that once the -12v and +12v were depinned, the short did not occur.

Now I need to run through all the capacitors and look at the ESRs
But my suspicion is on one of the LM833s because diode mode on the pins reads differently on one out of the five, whereas the other 4 read the same.
- my dirty test was putting the positive DMM lead on -VCC rail and the negative DMM lead on the output leads - this read a different voltage in diode mode compared to the other LM833s.