Sony STR-6800 receiver - stiff tuning, dial lamps and...

[stevehertz]

I've been doing some hunting around on the internet - where incidentally this receiver has a big following - and I found a fairly detailed report of how an American engineer had refurbished a 6800. Part 3 of his 'report' says this:

3. FM Tuner Issues

The FM tuner was found to only play in mono (i.e., didn't decode stereo broadcasts). The MPX circuit had to be adjusted to detect the stereo pilot signal: a frequency counter was connected to the test point and RT205 was adjusted for 19kHz.

Do you think this could be my problem? He doesn't say that the sound was only coming out of one speaker, but would this be the case?

I have an old Thandar frequency meter so I could use that if need be. I guess component drift could cause this kind of thing to happen.

[sp10mk11]

Hmm I see, a different reason for it not working on one channel, as I understand it if the decoder is not switched on you get mono but out of both channels.
That big red thingy I mentioned is the Low Pass Filter for the audio circuits, when I did mine I just used a phono lead with a 10uf cap on it into the aux of my test amp so I could trace the audio signal, I didn't have a service manual for mine either, but I started by tracing the audio l/r back from the function selector switch onto the board and through the components that's how found the resistor that had gone open circuit, sound one side of it but not the other I was surprised but that's what it was.
Gary

[stevehertz]

I can't tackle this level of fault TBH, solid state stereo decoders and MPX filter circuitry, it's beyond my capabilities. Having first carefully marked its position, I have tried adjusting RT205 with no effect. Ok, I wasn't inputting the required test frequency, I was picking up a stereo channel, but as I did my sweep, then surely if the 'correct' position had been passed, the left hand channel would have sprung into life, it didn't.

I'm going to get the freezer spray out and see if that pinpoints anything, crude I know, but I'm stumped otherwise.

[sp10mk11]

If I were closer I would help you out, can you measure resistors?
if you can then it is a matter of testing about 6 to see if it is the fault is the same as mine.
Each three of those resistors are for one channel, left/right so if compare them you might see some thing

If you can take a picture above the Large rectangular RED block (LPF) to include the front to edge of the circuit board I might be able to direct you as to where this resistor is that I am talking about

Gary

I have blown up the picture of the parts I am talking about, in that black square you will see two transistors with a electrolytic cap between them and 3 resistor infront of each of them, from what I remember it was on of those resistors that gone open circuit, worth checking the Transistors aswell, that cluster of components is final audio stage before the audio signal is sent off to the function selector switch

[stevehertz]

Gary I can measure resistors, I'm a qualified electronics engineer, I did 12 years working in a test department of an industrial electronics company and I've been repairing and restoring vintage radios, TVs and videos for nigh on 40 years. But this level of complexity and the highly populated, intense 'modern' build style just has me foxed, you can't get at anything, that's even if I understood what was going on and where to look, measure, whatever. I'll measure those resistors but it's unlikely that another should go open circuit unless there's something in the circuit that has caused it (them) to do so. But in any case, measuring resistors in circuit is pretty meaningless unless it reads near o/c, there's so many parallel (and semiconductor) paths to consider.

[stevehertz]

Those resistors all seem to measure ok, and at least, there's two rows (channels) of them and they all compare like for like, so it does not point to a faulty resistor in that small bunch. I'm out of freezer spray but will get some and see if I can pinpoint anything that way. Not very technical I know, but I have to do something or this receiver's gonna sit half naked gathering dust on my dining room table for ever and a day. Annoying as I bought it as a worker - and it basically was apart from the lamps and the drive cord - I really didn't want a long, drawn out fault finding test on something of this size and complexity.

[sp10mk11]

Hello Steve I sympathise with jobs that turn out to be long ones, they end up being part of the furniture.
Sorry I wasn't being condescending when I asked if you can measure resistors, I would be the last person to take on any of that sort of attitude, I am totally unskilled in this field but I have years and years of experience with this sort of gear, and as I mentioned before the same fault my own STR-6800 turned out to be one of the resistors in the set of 3 going o/c.
Good Luck
Gary

[vidjoman]

Do you think the sound from the working channel is mono or one chan of stereo. If the latter then the decoder should be working. My first checks would be the switches - the radio selector, the mono/stereo switch or tape monitor etc. If the decoder isn't giving stereo it would normally give mono on both channels unless there is a component failure in the latter stage of the decoder.

[Radio Wrangler]

Hi Steve,

You've said enough to isolate the fault to a small area of circuitry. That there is the rest of a large and intimidating beastie around it is of no importance.

The heart of this machine is the input switch. It selects amongst all the external inputs, the AM tuner and the FM tuner.

I assume that both left and right inputs work normally with signals applied to the external inputs?

I assume that you get sound out of both left and right channels when AM is selected?

If so, then everything after the input switch is OK.

Now, if you select FM and tune around with a bit of antenna connected, you receive stations, but you only get sound out of one channel? Can you get the 'stereo' light to come on? If not, have you checked the bulb?

You don't need the bulb to be intact to have the stereo decoder working, but if you get it to light, it is an indicator that the phase locked loop in the decoder has locked successfully and that a lot of the decoder working.

We know that the tuner up to the decoder chip IC202 is working because it is a single path and carries all the information that turns into both channels.

The decoder chip should be giving audio on both outputs (pins 4 and 5 on IC202) whether ot not it is decoding a stereo signal. If you turn muting off, then you should get noise on both outputs if you are tuned off of any station. If you have an oscilloscope, you should be able to see noise waveforms on both these pins, or audio waveforms if you tune a station. Don't worry about stereo or decoding yet. There should just be audio on these pins.

These outputs feed a pair of single transistor amplifiers Q210 and Q211 it looks like and these feed the multiplex filter (green block). Now I've known these filters fail on their internal PCB, and open circuit one channel. The filter removes 19kHz and 38kHz contentso that if you tape record from the radio, you don't hear beats with the recorder's bias oscillator. There are two filters, left and right in the box.

Next off, the signals are passed through a pair of Dolby processor chips IC 301 and IC3??

These are very unimportant. I don't think any station now transmits Dolby encoded material, so they have no use, and they could be bridged-out if one of them has failed.

After the Dolby chips, there are two single transistor buffers Q301 and Q33? and these feed the selector switch board.

If you have a scope, you can follow the signal from, say C231 where it should be visible and work your way down the signal flow until the signal vanishes on one channel.

If you don't have a scope, you can tune away from any station and engage muting to get silence. wind up the volume control to 50% and scratch one end of each of C315 and C365? to check that you hear a scratchnig noise in each channel. IF OK, try C301 and C351? to see if anything is getting through each Dolby chip. If so, then try C250 and C251 at the inputs to the dolby chips. to see if anything is getting through. If not, you either bridge-out the chips, or find out what's wrong.

If both wee OK, then work back via the inputs to the green filter, then Q210 Q211 and hthen pins 4 and 5 of the IF discriminator chip IC202.

David

[stevehertz]

Thanks guys for your input and encouragement. I have not delved into, or actioned any of the suggestions yet, but I can confirm:

On an FM broadcast, the stereo lamp lights up. The sound coming out of the right hand channel is the right hand side only (I'm a musician, I can hear things like guitar solos are clearly missing or very quiet). When I adjust the balance pot with a stereo channel playing, the right hand 'half' of the swing is all the same volume (constant), it's only when you start to move past the indent to the left that the volume gradually goes down to zero - well, you can hear a very faint channel. If I press the mono switch, the set behaves normally as you would expect it to do with it switched as such; both speakers working, channel balance smoothly swings etc. On AM, sound comes out of both speakers. The two wires going to the Dolby FM lamp have been disconnected and joined together; the lamp is bypassed with a link..

Does that narrow it down further? I have a scope.

One of the first things I did when I got it was clean the pots and switches as they were playing up - as is normal for these type of things - and all seems ok in that respect, but you can't be sure if there's still a fault that could be attributable to such a problem.

[Radio Wrangler]

It narrows it a LOT.

The stereo light shows the decoder chip is running, is finding a 19kHz subcarrier in the broadcast and is locking onto it.

Knowing that the right channel is playing right-only material shows that the L-R difference signal is being properly demodulated and that the add-subtract matrix circuit is doing its job too.

AM coming out of both speakers proves that everything is OK after the selector switch.

So on the dead left hand channel you have the output pin of the decoder IC202, a single transistor buffer, one Dolby chip (with trimmings) another single transistor buffer and then the input selector switch.

Following the signal through with a scope and scope probe is straight forward. The buffers are eminently fixable if necessary. The type number makes the dolby chip look like a Sony special s that could be a problem, but as no-one transmits Dolby-processed FM (and Dolby definitely isn't needed on decent signals) then if there's aproblem in this area, you lose nothing by bridging out the dolby de-processor section all together.

Out of wanting to keep the beastie original you might fancy having a go at fixing the dolby area if it happens to be in that area, and as long as it isn't the chip itself you have a fighting chance.

Probably you're looking for a bad solder joint, a cracked track, a dead transistor. Nothing too wild.

David

[stevehertz]

Thanks David. Later today I'll scope both channels and compare them. WRT the Dolby FM circuitry, I am not in the least bit bothered about originality since Dolby FM is as redundant now as it was then, as it is for always, so if I have to remove or bridge it that's fine by me. I'm just concerned that this is done 'properly' though so that both channels work the same and as per spec. Anyway, let's see what I can find. Oh, wrt bad soldered joints and PCB tracks etc, there lies a problem. This large board is joined to the chassis via the drive cord (it would have to be removed - after having just battled to replace it), and there are also many stiff connecting wires on three of the four sides, so how is one supposed to flip it over to see the PCB tracks?!

[stevehertz]

Working backwards from the output of the Dolby chip IC301 on the faulty LH channel, there's no AC signal until the base of Q210 where (AC wise) it is as the RH side - well as far as I can see looking at a complex AC waveform jumping around with every beat of the music! There's actually also a tiny signal at the emitter. But DC wise there's some strange anomalies around this transistor:

Q210 Q211 Cct spec
c 12v 7.5v 7.2v
b 0v 1.7v 1.61v
e 0.05v 1.5v 1.5v

I've swapped the transistor for an equivalent new one and there's no difference. Unable to access the underside of the PCB, I snipped off the trannie leaving long leads and quickly soldered the new one to the leads.

[Radio Wrangler]

Good-good-good-good-good in my best Neddy Seagoon voice!

The signal dies before the dobbly bits, and there are bad DC levels to explain it.

But a working transistor in its linear region should have something between 0.6 and 0.7 volts between its base and emitter. Q210 looks like it's biasing network has failed. Q211 shows there may be a bit of misunderstanding in your measurements, or it's bias network might be troubled.

The misunderstanding: A working silicon transistor in its linear region should have something between 0.6 and 0.7 volts between its base and its emitter. Sony's diagram shows 1.5v on the emitter and acurved arrow depicting 0.6v between the base and emitter, which would add to put the base at 2.1v above ground/chassis.

R257 is 330k and R(illegible) from the base to ground should set up the 2.1v from the 14.9v supply. so the illegible resistor should be 54.14k by calculation, not allowing for loading, so they'll have fitted 56k or 68k (either should work)

Now 1.5v across the emitter resistor of 680 Ohms means an emitter current of 1.5/680 = 2.2mA we can say the collector current is pretty much the same because Q210 should have plenty of gain, making the base current much smaller than the collector current. So 2.2mA in the 3,3k collector resistor gives 7.279 volts drop in the collector resistor... which from a 14.9v supply gives 7.6v nominally on the collector. This all fits as well as you get with nearest value resistors.

So Q210 seems to be turned almost all the way off. The emitter voltage is damned all, and that looks to be caused by no bias on the base. I suspect thar the top resistor, R257, 330k has either gone high or has a duff solder joint at either end.

Q211 has 1.5v at the emitter , and that gives 7.2 at the collector, so all seems well on this side.

I have a few Sony ES series items whose boards would be terrible to have to remove but Mr Sony helpfully put a removable panel in the bottom of the chassis.

David

[Goldie99]

Hi Steve - I've been watching your thread with interest, and to improve my own fault finding skills (or lack thereof, which is probably more accurate) - I suspect both David & Gary may be correct:

The illegible resistor David couldn't read seems to be R255, which the manual indeed shows as 68k, and which together with R257, sets the base voltage of Q210 - the hifiengine copy of the manual is clearer in that area.

If R257 is open, you lose the base voltage - and R257 just happens to be one of the 6 resistors (I think it's probably 8 ?) that are highlighted by Gary in his photo in Post #64. Could it be worth checking them again ?

[Radio Wrangler]

Ah, yes that's a lot clearer.

The 3.3k collector load resistors serve a dual purpose by the way, they set a defined source impedance driving the filter unit. Filters are very critically dependent on their terminating impedances. If you had perfect inductors and capacitors you'd expect to be able to make quite good filter. But perfect Ls and Cs are lossless and dissipate no power. So where do those signals go which don't make it to the output? They bounce back out the input. There's nowhere else for them to go.Anther way of looking at it is tha the source and load resistances load the first and last resonators in a bandpass filter and therefore control their Q, thereby setting the bandwidth of the whole filter.

David

[stevehertz]

As an aside, this receiver uses a connecting wire termination method whereby a wire going to a PCB is wound multiple times around a square section upright pin that is soldered into the PCB. I guess the sharp corners of the pin help to make a good joint, but some of these connections look soldered and some not? I suspect they are all soldered but some are 'more full' of solder than others. Any more info or tips re this type of joint/connection? I mean, good quality push on connectors would have been much better for serviceability, but then the words cost and reliability come in to play. But I did say 'good quality'. Never mind, I doubt if the designers of these sets in the 70s thought or even cared if they would be in use 40 years later!

[Goldie99]

Hi Steve - that connection type is called 'wire wrapping', and from what I've read previously it provides a high quality connection, on a par with, or better than soldering, if done properly - there should be no solder present in or on the wire wrap joints, unless someone has added it later.

They are however a pain for repair work - and should ideally be left alone unless you have wire wrap tools & expertise to replace them. If you need to separate a board or something, the wire wrap posts / blocks can usually be desoldered from the back of the board (assuming you can get access) - that's how I worked around them in an early Sony TA-F222ES a while back.

I also just noticed the green / blue resistor in your last photo - it looks suspiciously like a fusible resistor, which I believe Sony used quite often back in the 70's / '80's. I didn't see any reference to 'fusibles' in the (partial) service manual I was looking at, but if that's what it is, they are notorious for failing & going high as they age. The TA-F222ES was full of them.... it might be worth checking if fusibles are mentioned or shown in your manual ?

Alan

[stevehertz]

Thanks Alan, I'll check out the fusibles. Re desoldering the wire wrap posts from the board, there isn't any access to the back of the PCB, that's my basic problem and prime need to access the back of the PCB in the first place, I cannot see what is physically going on in the area of the fault.

[sp10mk11]

If I cannot remove wires easily, providing there is enough slack on them I will cut them leaving a little of the coloured sheathing on them so I can see where they went.

The Japanese in the 1970s where very proud of there workmanship and the quality of there goods. what you say now would make sense but not back then.

Good luck Steve
Gary