'90s audio amp repair (need help) : Sony F261R

[Vince]

Hello people,



Warning : long post, please come back when you have some time ahead of you... then grab a cup of tea and a packet of your favorite biscuits to go with it.


Just done fixing my childhood 30 year old CD player donated by the old man, that old swing arm Philips :

https://www.eevblog.com/forum/repair/vintage-cd-player-repair-philips-cdc-486-6-disc-changer/msg3366804/#msg3366804

Now need an amplifier and some loudspeakers so I can hear something out of it...

No, I don't have any... since I left the parent's home 20 years ago, I never bought "proper" audio gear, since I am not an audio guy so to speak.
So always had other priorities...
So far I have just been playing music from my desktop computer via half decent little Altec-lansing amplified speakers sitting on the desk either side of the monitor.

I am a simple man. Maybe too simple ?!......

Anyway, have at last a home of my own, so I can now contemplate a more "permanent" installation, and envisage at long last to get some proper audio gear sitting in the living room.
The CD player is dealt with. Now need an amp.

My knowledge of amplifiers is limited to what follows :

- At the back there are lots of connectors. Some of them you connect the CD player to, and some others you shove bare wires into, that would be the loudspeakers I believe.

- At the front there is a big power switch, fancy LED lit buttons to select an input source, and of course that big knob to make the thing sound louder or quieter. " Volume control " I think they call it. Sometimes there are more buttons, to control treble/bass/balance.


That's about it...


To make it cheap, I bought a used amp of course, and to make it even cheaper I bought a broken one. This way I would get some fun (or not...) trying to repair it ?! ... and learn some stuff about audio amps along the way. That's the plan anyhow...

So I sorted by price and just picked the first decent looking amp I could find. So, destiny had me buy this amp: a Sony F261R , Ultra 5000 v2.0 limited Bonus Mega sound edition. Whatever... it's an old amp about the era of the CD player, a tad younger... I figured that whatever I picked, it had to be at least as good as the built-in sound chip on the motherboard of my 15 year old computer, and built-in amplifier in the little speakers.

Paid 10 Euros for it, plus 8.5 for shipping, so no big deal...

Seller said that the problem is that "stereo does not work any more".

ME : cool, maybe just one or two power transistors to replace in the affected channel.

REALITY : Just AFTER having made the deal, not before.... hum.. I downloaded the service manual for it. Happy enough that there was one available to begin with ! A quick look at the schematic revealed that oh.. crap, power stage is not made of easily replaceable, discrete trannies, no. It's an integrated power amp IC ! A huge IC, 18 pins in line. An STK4182-Mk2.

So... if the chip is bad then good luck finding a new one. Best I could find were used ones in China for a few bucks. But well, at least chip is not complete unobtainium, so it could be worse.


So I received the amp yesterday evening, and started on it right away. Was so impatient to discover my first audio amp and start playing with it.

However since it's my first one and I know nil about audio stuff, I figured I would bother the forum with this repair... so that people in the know can give advise and share their knowledge and experience that's bound to be infinitely superior to mine... anything greater than zero is infinite, after all...


So let's get started ! 


Popped the hood. Looks very clean and tidy, Japanese style. Looks very basic too. No big toroîdal transformer. Other than the power supply and front panel board, the business end of the thing is a single board construction, a small one at that. Single side, large through-hole components, easy access to everything. Simple construction. That will make it easy to probe around, just what I need for a first repair... happy with that.

So before touching anything, I gave the thing a test ride to confirm the seller's symptoms. I don't have speakers just yet, haven't had time to start hunting for some yet. I just have my big headphones. So I plugged them on the front jack : works just fine ?! Sounds good, I have sound in both ears no problems. Balance and tone controls work fine, main volume knob too....
Hmmm... strange... too good to be true. So I thought it's good and bad news : probably means that the headphones being low power maybe are not fed by the big power amp STK chip, but either directly by the input signal (not knowing any better...), or maybe a small low power chip somewhere. On the plus side, it meant that 90% of the amp was good : power supply, input selection mechanism, tone, balance and volume control... so a lot of the unit was clearly working.

So it looked like the stereo problem was probably affecting only the loud speaker, not the headphones.
Stuffed the DMM into the speaker connectors at the back... gotta do with what you have eh ?
Readings were strange... 2.5 VAC on one channel, 2V on the other one, which quickly declined to 0.5V while the other stayed at 2+ V. OK I thought... there is indeed a problem with one of the channels. Looks like maybe thermal issue that takes a couple minutes to progressively manifest itself ? Except that I then power cycled the amp and this time the readings were different.. channels were reading the same this time, and then they declined to 0.5V, but in unison. There was no longer one channel behaving differently than the other. Checked DC levels, got 0mV, dead zero. Good news I thought ?! I don't know...
At any rate, there was SOMETHING coming out of the both channels. It's not like there was sometihng on one channel, and the other channel was 100% dead. No.
So I was starting to have some hopes that the STK maybe was not dead. Just maybe.

So I pulled the main board out of the unit so I could give it a good close-up visual inspection. On the component side I didn't see anything obvious. No burn marks. No brown glue going bad/conductive (no glue at all on this board). As for the 34 electrolytic caps on that board, none were bulging, and none were leaking. A third of them are Nichicon and Elna branded so happy with that. Those are the bigger ones, and those located near the power amp and power supply. The remaining 65% are small caps located in the low power parts of the board, and are crappy "Jamicon" brand. They sit high above the board surface so one can easily see if they are leaking. The bigger caps however sit flush on the board so I can't see underneath them.
But, at least at first glance, nothing massively wrong with caps.

Then I flipped the board to inspect the solder side. All the joints looked nice and shiny. Could not find any bad joints at first glance. Wiggled the power amp chip to see if I could "trigger" some joints... but not that I could see. A bit disappointed then ! LOL
Still, since I was in there, and just for peace of mind, I reflowed some joints :

- All 18 joints of the STK power amp chip.
- Connector for the large flat flex cable that comes from the front panel board.
- The relay
- The speaker connectors.

Put the board back into place and gave it another test ride.
Somehow it behave differently :

- Now I get the same voltage on both channels. Obviously it's dancing up and down as the song from the CD player is unfolding, but you could see that both channels behaved very similarly none the less. Also, if I move the Volume knob, voltage goes up and down as you would expect, and affects both channels the same way. So... it looked like at a fundamental level at least, it was doing something promising...
So to go any further I would now have to take it a step further and plug actual speakers to hear what it sounded like in practice, what was actually going on. Didn't feel like waiting weeks to hunt for speakers, so instead I stuffed some bits of wires on the speaker connectors, and used that to hook a couple probes to check with a scope what was coming out of the connectors...
It was looking pretty good : waveforms looked like music to me, dancing around in unison with the music playing from the CD. Also, both channels looked similar in shape and amplitude. So kinda happy with that.Volume knob was able to vary the amplitude of the waveforms from tiny to super large, so looked good.

So I really needed some speakers to try it out, was getting excited that it might be working now !
So... I thought maybe I could hook up my headphones to the speaker outputs ?! At first I thought it might blow them and I would be very unhappy, but well, had to do something ! I thought maybe if I set the volume to zero and increase it only extremely slowly, a tiny little bit, maybe the headphones would be able to cope with it. At any rate, worse case scenario, I paid the headphones 15 Euros 15 years ago, and one of the wire is chewed up, the wire is exposed. so I was not gonna cry if they did blow...

So I searched my ton of random salvaged parts and found a headphone socket. Soldered 50cm of wires to it, just enough so I can bring the socket back to the front of the unit, for convenience. Turned the volume all the way down. Plugged the headphones, powered up the amp... and anxiously, extremely slowly increased the volume....


we have sound !    

And it sounds just fine, and in both ears !


Balance control works fine too. It just... works.
With the volume set 5 to 10% it's fine. At 15% it's very loud and at 20% it's getting ear damaging... So 15% max it is for headphones.

YES ! It's fixed !!!   

Retrospectively, I think the most probable is that the problem was bad solder joints on the speaker connectors, causing intermittent, unreliable connections, leading to random weird voltage readings. Reflowing them fixed the problem. Knowing that, I might consider pulling the board again, to suck all that old crusty solder and put some fresh solder in there...
So... EVEN if the joints look fine at first glance.. reflow the usual culprits ANYWAY, you never know !!! 



So... the big STK power chip is working in the end, got lucky ! LOL 




HOWEVER.... even though the main issue is fixed, I still have other minor issues that you may be able to help me with...


1)

The headphone jack at the front, no longer works ?! Well, it works strangely at the least : I can hear the music playing, but I need to crank up the volume to 50%. Also, it sounds weird, very hollow... sorry I don't know my sound description terminology...  It sounds like the singer is 2 miles away from me, very faint. The mid-range frequencies are almost absent. I can hear mostly high-frequencies, and even then they sound "metallic".. it all sounds very weird.
Also, the balance control acts bizarrely too ! It doesn't change balance at all : no matter its position fully to the left or right, it always spits out a balanced/normal output. Also, strange too : I noticed that if I turn the balance control all the way to the left or to the right, it changes the sound and makes it sound almost normal. 

But as soon as you disconnect the headphones from the front jack, and plug it to the back of the unit on the speaker outputs... sound is perfect and the balance control now acts as it should !

So the weird sound and balance control weirdness only show when using the headphone output on the front panel.

Also, I noticed while probing the speaker outputs with the scope, that when I plug the headphone at the front, the speaker output, which normally I think are supposed to be disconnected by the relay ( I can hear it clicking whenever I insert or extract the headphone jack into its socket)... actually DO output something ! I see a clean sine wave oscillation ! Eh ? Well it's a sine wave when the volume is set very low. The lower the output the higher the frequency. 275kHz in this case. Then as I increase the volume, the amplitude of the oscillations increases too, and the frequency gets lower and lower. The lower it gets to, is about 35kHz, at which point the amp is clipping and the oscillation looks more like a square wave than a sine. That allowed me to measure the slew rate of the amp : about 0.5V / us ... any good ?
Then if I keep increasing the volume, frequency of the oscillation is now starting to INCREASE again.

Now I understand that any amplifier by nature can oscillate. Maybe it does so because the 1Mohm load of the scope probes is way too high compared to what its designed to drive, obviously. However as I said, from what I can see on the schematics, and from I witness/hear : the relay disconnects the speakers from the power amp output when you insert the headphone jack so.... how come I get ANY signal at all on the speaker output ?! .....

Of course the easy answer would be : "Vince, what do you care, you are not even going to be using the headphone output (indeed it's not likely) ! ".
But obviously I would rather try to understand what's going on and learn something along the way... otherwise not much fun eh....

Given that the headphone output used to work fine when I first tested the amp, it's reasonable to think that I caused the problem. All I did was to disconnect the flat flex cable that links the front panel to the main board, to spray some contact cleaner (some "Deoxit") into the connector.
It's supposed to help, not hurt...
Still, just in case, I inserted a piece of white paper into the connector and "wiped" its inside, several times, to remove excess cleaner.
Did not change anything. Waited 24 hours to let it dry or whatever contact cleaner does, tried again today.. still zero change.
So not too sure that it's the problem... Plus, again, balance control works just fine as soon as you switch from headphone at the front, to the speaker outputs at the back.

So I need to find an explanation that takes this fact into account ! It's not like the sound and balance controls were defective all the time... they misbehave ONLY when using the headphone output !  Any clue ?

Looking at the schematic (attached), it's hard to follow the mess of "wires"  in there, but it looks like the headphone jack is connected to two paths (I highlighted them in blue and yellow, one color per channel) :

- Goes to a "network" of discrete passive components, all intertwined, that handle all the sound "processing" features : treble, bass, loudness, and the balance and main volume control. The STK power amp does nothing but amplify with a fixed gain it seems, it doesn't modify the signal in any other way.

- it also goes straight to the output of the STK power amp chip ?! So.. unlike what I assumed at the very beginning of the repair, the power amp DOES seem to drive even the headphone jack in the end... 

Then let's look at what happens when I insert the headphone jack : the socket connects two pins together, highlighted in blue and red. From what I understand, it goes like this : chip IC601 , a uPC1237HA, monitors the two outputs sof the power amp, and if it thinks that something is fishy, it protects the power amp and speakers, by operating the relay to disconnect the speakers. One side of the relay is tied to a positive supply, and the other side goes through the contacts on the headphones socket, then back to pin 6 of that "protector" chip as it's labeled on the schematic. Looking at the data sheet of that chip, and although it's not explicitly noted anywhere, it's most likely just an open collector output. The contacts on the headphone socket are normally closed so that in normal operation, they don't interfere with the protector chip, and it can operate the relay as it sees fit. Then, when you plug in the headphones, it forces the relay to open no matter what.

So... here I am trying to make sense of all that...

Problem kicks in when I plug the headphones. All that it does is basically cut power to the relay coil so that it opens up its contacts and disconnects the loudspeakers. That's it. Nothing more. How is that supposed to cause my problems ? How the hell can the act of powering off the relay... cause havoc with the balance control, and how the sound... "sounds". Weird. As far as I can see, the headphones are driven directly by the amp, the path is uninterrupted. Looks like the headphones are ALWAYS fed by the power amp, no matter what, and that it's only the loudspeakers that get turned on and off by the relay. So even if the relay was dodgy, how could it affect the sound that goes to the headphones, since the relay contacts are DOWN stream of the headphones.

I am... puzzled....

Still, the fact that I can see the speaker output oscillate when the headphones are plugged in even though the relay is supposed to disconnect the speaker output is... rather strange to say the least. So I guess I should hang on to that cold undeniable fact, and start from there ?!...
So I could try removing the relay altogether, see what that does...




2)

Problem #2 : in the absence of an input signal, it I turn the volume up, then from as low as 20%, I can hear a loud "buzz", monotone sound, maybe the thing referred to as "hum" ? Don't know, maybe.  Then, superimposed on this, there is also a less powerful but very distinct "random" crackling noise ? White noise ? I hear talking about "static" sometimes, maybe that could be it ? Either way, I fear it would be a problem when I get round to connecting loudspeakers and need to pump up the volume to 20% or more. So again I would like to understand where these two noises come from and how to fix them.

The main humming noise, IIRC that could be excessive 100Hz ripple from tired filter caps in the power supply somewhere ? I guess I could easily probe for that, see how much ripple I get on the main power rails.

As for the static, don't know.... bad connection somewhere ? bad solder joint or dying electrolytic somewhere ? There are 34 of them on the board... might be quite a challenge to figure out which one(s) is at fault... any particular method to quickly find the faulty cap, if cap it is ?

3) Problem  #3 : well not really that big a problem, but I noticed something "funny" : while I am listening to the CD player, if I then switch to another input, say Tuner or whatever, then I can STILL here the music from the CD player ! LOL  Well of course it's very faint, but the lab is quiet and if you pay attention, you can hear the music. So there is some cross talk shall we say ! Now I don't know how much of it is tolerable / within spec, or if here too something is wrong and I ought to find where that cross talk comes from ? Tired electrolytic caps again, me thinks !   


OK, so that's it for now... hoping some kind soul might look at the schematics, my story, and offer some insight... any help or advise appreciated, as always...
In the meantime, will keep working on the thing and post updates as I go...

[Vince]

Schematic.....

[Vince]

Sorry for spreading the images over several posts, but recently the forum started randomly refusing to upload some pics. Don't fancy spending time uploading many pics then having the process fail at the end, not even knowing WHICH pic or pics made the whole process fail..
So until this is officially resolved (might never be...), I am reluctant posting more than a handful of pics at a time...

[Vince]

...

[tautech]

Great start Vince. 
Nice small mainboard that looks easy to work on. 

In this pic:


Is that a Phillips sig gen on the shelf ?
Will it do a 1KHz sine wave at low amplitude which is what's usually used for signal tracing through an amp ?
Depending on which amp input you use you'll need to keep its amplitude right down.

[Gyro]

Congratulations on the lucky save with the Sanken chip. 


For the headphone problem, my guess would be a broken or corroded ground connection on the headphone socket. If that is the case, then the headphone elements will be operating in series (out of phase) between the two channel outputs. That would tie in with the sound being tinny and strange sounding, and also improving when the balance control is fully one way or the other (more actual signal rather than channel difference).

The other thing I noticed is that the headphone attenuator is very primitive - just a pair of series 680R resistors. Assuming that you are using 600R headphones, this gives sort of OK damping factor. If using 8R or 32R headphones, the damping factor is much worse. This would give less than optimal sound quality, poor bass etc. Better amplifiers used either a separate headphone stage, or more commonly, a resistive divider to drive them with something like 10R as the bottom resistor (to maintain damping factor into low-Z phones) and the top resistor adjusted to give sensible volume (maybe 47-220R).

Regarding the hum and crosstalk. If listening to the speaker outputs on headphones, you can probably hear all sorts of low level crosstalk and noise. The reservoir caps may be a little dry but it depends whether you can hear any hum with an active source, and at normal listening levels. Source switching is performed by IC103, so there aren't any switch contacts that might need attention, it's probably just crosstalk in the IC.

EDIT: Also note that the inputs are high impedance when there is nothing connected - apart from some series 1k resistors on the inputs of IC103, I can't see any resistors to ground to define the input resistance (presumably this is defined in the datasheet for IC103. Open high impedance inputs are going to pick up all sorts of rubbish from inside the amp casing (including maybe random noise from the front panel micro. The situation changed drastically when the inputs are connected to the low impedance output of the CD player or other sources.

[Vince]

Hi guys, thanks for stopping by ! 

Update : I pulled the relay, no change whatsoever.

Tautech, yep it's my old '70's Philips sig gen, indeed ! A bit temperamental but still alive. I did think of using it to inject a tone in the CD input instead of the CD player... but that was just one thought out of a million that crossed my mind as I was working on the thing yesterday. I was afraid it would blow the input, assuming audio inputs used super low signal levels, but not knowing specific numbers...
So while the CD was playing, connected to the amp, I measured with the DMM the RMS value of the signal, at the input connector (from behind then, of course), and IIRC it read something like 2 Volts. 2V not 2mV ! So much, much higher than what I feared.
At any rate, the old Philips sig gen can attenuate down to 60db which would have made it in any case I think....
But at 2Volts for sure I would have no trouble using that sig gen. 

Gyro : wow... thanks a million for your thoughts ! Sounds like a good plan indeed ! Will dismantle the unit some more to get at the headphone board, and also the front panel board to check those 600+ ohms resistors.
Yes, headphone are low Z. I measured them at DC with the DMM, I get about 30ohms. One side is a bit more than 30, the other a bit less. Go figure why they aren't identical... just old age maybe...

Anyway, going to check around the head phone jack and then report back, as it looks like it might wel solve the main issue. I will do another pass later for dry caps and noise issues etc.

Stay tune ! 

[tautech]

Tautech, yep it's my old '70's Philips sig gen, indeed ! A bit temperamental but still alive. I did think of using it to inject a tone in the CD input instead of the CD player... but that was just one thought out of a million that crossed my mind as I was working on the thing yesterday. I was afraid it would blow the input, assuming audio inputs used super low signal levels, but not knowing specific numbers...
So while the CD was playing, connected to the amp, I measured with the DMM the RMS value of the signal, at the input connector (from behind then, of course), and IIRC it read something like 2 Volts. 2V not 2mV ! So much, much higher than what I feared.
At any rate, the old Philips sig gen can attenuate down to 60db which would have made it in any case I think....
But at 2Volts for sure I would have no trouble using that sig gen. 

Somewhere in Sony documentation there should be max input specs. Yes for Line In and such the signal levels will not be low like Phono (turntable) inputs are. Some old amps had inputs for crystal and magnetic pickups where the magnetic pickups needed a further amplifying stage.

[Vince]

OK, just took the thing completely apart. It's guts are now all over the place... well at least everything is accessible now ! 

The headphone socket is mounted on a tiny board of its own, which then plugs into the front panel board.

I was as thorough as humanly possible :

- Contact cleaner in the headphone socket, and in the connector linking the two boards.
- Sucked all the old solder from the socket and connector, and redid all the joints with fresh solder. All brand new.
- On the front panel board, tested the two 0.5W 680 ohms dropper resistors. They are fine. Redid their joints too, just because.
- reflowed all the pins of the large flat flex  that goes to the main board, since obviously the headphone signal path goes through there.

Then tested the amp, luckily I could do so without reassembling anything, didn't put any screws anywhere.

Result ? Nope,not a sausage ! Not the slightest improvement, 100% the same as before. Depressing, sounded so promising...

It's midnight so will leave it at that for now. Tomorrow need to do some driving to visit the old man 100 miles away. He will give me 3 more items to fix ! Another CD player, 5 disc carousel thing. Also a thermostat for his fancy modern gas heater that cost him a fortune. Then also an little industrial programmable controller, a " PLC ". A small one that you can plug on a standard rail in your electrical panel at home. Said that a 24V part of it, accidentally got inflincted 240V / mains... ouch... said he had a quick look at it. Found a 2.2 ohms resistor that blew. Replaced it.. no joy. PLC still dead or so it seems : LCD screen doesn't come to life. I am not too hopeful on that one... 

Anyway. Gyro's hypothesis is so interesting that I can't give up as easily. Maybe I missed a joint somewhere... or maybe the jack socket ITSELF is bad, internally ?! After all this amp is just a piece of " junk ", bottom of the barrel design, let's face it... so I doubt Sony would have put money into a top quality durable socket...
So I could try bypassing that socket and replacing it with another of mine, see what that does.

If not, then I guess I will have to go the signal tracer route ! Inject a tone into the CD player input and follow the signal to see where it goes haywire.

Anyway, going to bed, stay tuned !   

[shakalnokturn]

As far as I can tell it's mostly working.

I think Gyro's right on the ground to headphone socket.
Try a continuity check between ground ring (outer contact)  at the phones socket and middle point at the transformer or capacitors on main board.
This amp appears to use the metal frame to get the ground to the phones so check on the socket that there is electrical link between it and the mounting bracket.
On the other side (mainboard) the frame connection will usually be made by a screw and metal tab (sometimes soldered wire) if you've missed a screw or left one fitted but loose it can be that simple.
It would also fit in well with your unexplained oscillations on the scope if your probe ground clip was on the metal frame when you noticed that.

Is the relay a sealed one?

[Vince]

OK I am back from the old man. I now have 3 more items to fix ! ... but he also donated me a pair of loudspeakers ! Crappy ones, small Toshiba 2 way thingies that came with a cheap audio gear we had when I was a kid. Was a crappy all-in-one thing, not portable boombox style, but still, was crap. Had a vinyl turntable but no CD player, so that's why daddy bought the PHilips CD player soon after.
Anyway, they look crap and believe it or not sound worse than my small amplifed speakers on my computer... but for now it will do ! Good enough to help me in fixing the amp. Will hunt for better speakers later.

Relay, sealed ? Do'nt know... I did'nt pay that much attention to it when I had it removed. ISTR seeing black resin at its bottom, maybe... so yeha maybe sealed. However if you look at its datasheet (somehow Farnell still sells that exact part number, still available !), it just says it's a general purpose relay for audio or home appliances use, nothing special. Doesn't pretend anywhere that it's sealed.

https://fr.farnell.com/oeg-te-connectivity/osa-ss-224dm3-000/relais-dpst-no-240vac-30vdc-3a/dp/1891796


Ground : played some more in this direction... definitely looks like a ground problem indeed !  Not solved yet but the few experiments I just did kinda validate Gyro's scenario !

What I did :

- Checked ground continuity between the ground pin of the socket, to the track, then the bracket that hold the socket to the front panel chassis = GOOD

- Then I used a test lead with crocodile clips to grab the bracket on one side, and with the other clip I grabbed various parts of the main chassis, as well as the 2 screws that hold the main board to the chassis, and also secure the big heatsink of the power amp, the board and chassis. That's the only electrical path I can see between the chassis, and the board. The two heatsink screws...  Yes they are tight.
That made no difference at all, still crap sound in the headphones.

- Then I thought OK maybe the headphone socket itself is bad, so what I did was pull slightly the jack from the socket, so that I can expose the ground part of the jack, just a mm or two,just enough that I can touch it with a solid copper wire while at the same time still hear sound in both ears. This way, I had the signal coming to both ears, and I cold make sure the jack/headphones were getting a ground, because I was in control of the ground wire !  So, again, I put my ground wire on various parts of the chassis, and the heat sink screws... still no joy, zero change.  So the socket itself seems OK, and the board holding it, too !

- Was starting to get depressed at that point...  So I switched gear, and decided to get rid of the headphone jack board altogether ! I pulled it off of the front panel board. So now all I have is the front panel board with the 4 pin socket, that's now unpopulated so to speak.
This way I could plus my home made headphone socket with the wires soldered to it.. and be in complete control of the ground, both channels, and connect them where ever I wanted on that board ! Or even on the main board itself if so I wished. I was also de facto in control also of the output relay : just need to short 2 of the pins on the connector, to simulate the old headphone socket, and I could at will turn the relay on or off. So... in complete control I am, now.

So... I shoved the wires for the left and right channels, into the connector, and put my ground wire in contact with the chassis. REsult ? Still no joy ! So indeed the original socket and its board were not at fault.

Then I did to more experiments...

1) Still with the 2 channels connects, I REMOVED my ground wire.... NO CHANGE !  So... it means Gyro was right, there is no ground there ! 

2) I reattached the ground wire to the chassis, and this time connected only ONE channel at a time. Result : NO SOUND at all !

So looks like that confirms again Gyro's idea. ground is plain missing, and I get a signal only when both channels are connected.


Seems to me that the above experimetns are enough to prove that Gyro was right. So now I will not chase this or that red herring, and concentrate 100% of trying to figure out exactly what is this ground problem I am having...

I guess there must be several different grounds in this amp, I need to look at the schematic carefully to figure out how many different grounds there are, what part of the circuit each of them encompasses etc...

Clearly the headphone ground is supposed to be connected to chassis ground, OK, but then that chassis ground connect to WHAT exact point on the main board...   It connects to the power amp heatsink, sure, but other than that...

Anyway, thanks for the help, looks like you put me on the right track... will keep investigating ground related stuff until it's finally fixed !

[Gyro]

Thanks for the follow-up, I like being right. 

Looking at the schematic, it just shows the headphone socket ground going to a chassis symbol. That doesn't sound like a great idea, ever. Looking at the picture of the headphone PCB. there are only 4 pins on the header (all the signals that you highlighted on the schematic). As the front panel looks like plastic, it must be relying on that metal bracket for the ground return to the main chassis. Hopefully it attaches (to the main chassis) via a solid mounting screw rather than just a sliding fit!

I would be tempted to isolate the bracket from the main chassis (thick tape or whatever) to prevent possible ground loops and run a proper ground lead back to a convenient point on the common speaker -Ve connection. That should take the chassis right out of the equation as far as the headphone output is concerned.


P.S. I was a bit worried that the speaker terminal -Ve connections might not be referenced to chassis but if you follow the thick black trace on the schematic down and across, it shows a chassis symbol near the junction of the main reservoir capacitors C903 / C904. Where it is physically on the main PCB, I don't know, it may be a PCB mounting screw near them that is not making good contact, or is missing.

[Vince]

FIXED !!!!!   

Yes, teh headphone bracket is firmly secured to the main chassis with a metal screw, no worries.

Finally found the "problem"...


1) Checked the schematic and did some probing on the board(s) to confirm. There aren't that many grounds at all in fact !  Just a seperatage ground and supply for the digital +5V, but for all the analog stuff, low power or or high power, there is just one single ground. So, knowing that, I connected the ground wire of the headphone socket straight to a ground on the main board.. easiest to tap is just the back of negative / black speaker connectors. When I do that, problem fixed ! Sound is loud an clear / perfect in the headphones, yes !

2) Then now remained to figutred out why the chassis was not connected to the ground on the board... I notived a mettalic tab hiddent between the PCB and one of the two mountnig screws for the heat sink. Tab was soldered to the board and buggering off to god knows where... but assumed it had to be ground of course. But... I ALREADY put a screw in there in my prior tests, tightened it well, yet it didn't work !  Then I noticed that among teh screws I had removed / laying around, one of them was standing out : same size and type as the other mounting screws for the board but... different surface finish : it looked light grey, like bare metal, rather than all the other screws that feature this yellowish plating. Hmmm.... I though this particular screw must be used to make a good electrical contact for that tab, maybe the yellow screws don't conduct electricity well enough, don't know. Seemed to me like I had nailed it, was excited..... but it did not fix it ! 
So pulled the board again, checked that tab more closely and eventually realized that NO, it is NOT connected to ground ! It goes to one place and one place only : pin 16 of the power chip ! .. throught a 10 ohms resistor. Indeed the schematic shows that pin 16 is connected to the chassis.

3) So... where the hell is the board supposed to connect to the chassis ! ARRRRRRRRRGH !!!
Previously I had ruled out the connectors on the back panel since the screws bite in the plastic body / insulation of the connectors... but since I had nothing left to look at, I looked more closely and finally found that on the big cluster of jacks for the various inputs (see picture), there is a vertical "fin" / tab that's aligned with one of the mounting screws, and it's soldered to the PCB ! YEAH !!!!!!!

That fixed it ! 

OK, so in the end it was stupid OK... but who cares.  Point is, I learned stuff while trying to figure it out, and got some hands on experience on my first audio amp... which is what I wanted 

So I am happy. Bought for 10 Euros only, fixed quickly for absolutely zero Euros, not a cent !

Thanks you both for your suggestions, you were both spot on, I guess your experience shows !   

I am glad I posted this report on the forum then, got lots from it.

Oh and as far as the other problems go : the humming and static... checked all rails for ripple, they are just fine. With the loud speakers connected, I hear ZERO hum or static any more ! Even when I turned the volume to 10% ! 

It's a delight 

I loved tis repair, wnat to fix another amp now !!! 

This one looks and feels a bit cheap.. face plate and knobs are really cheap plastic, not very pleasing to look at, up close.
Maybe I could sell this one now that it's working, for I don't know... 50 Euros ? Sounds ambitious I know... that would make a 30 Euros profit which I could reinvest in another broken amp, but a little more expensive and better quality ? I don't know, some old Yamaha, liked them. Or whatever decent brand of the early  '90's, not too clued. Will see what I can find....

Anyway, that will be for another time... for now, I have more stuff to fix !!!  Then the house to finish since I am getting bored sleeping in the living room for months now... then when the house looks like a house and I have no more dry walls and painting to do... I will search for nice decent old loudspeakers, and a decent quality amp to fix ! 

Thanks guys for your help on my first repair, that was fun !and interesting ! 

[Gyro]

Yay, a good way to start the New Year! 

You can see why I distrust signals being routed through the chassis -  when it comes down to what screw plating...