NAD 7020 weak channel on FM stereo

[qualityten]

I recently acquired a NAD 7020 in 'non-working' and original condition. After cleaning the pots and switches it's working fine, except that on FM stereo one of the stereo channels is weak and distorted. FM mono is loud and clear. And the phono input and the Aux inputs both produce good clear signals in both of channels.

I have downloaded the schematic and service manual from HIFI Engine. This is pretty full, with an inside view showing the physical locations of areas of the circuit, a block diagram, full parts list, schematic and wiring diagram.

But as I don't understand the FM multiplex and buffer circuit, I don't know where to look for whichever problem component or dry joint might be causing one FM channel to be weak. Any assistance members can offer towards tracing this fault would be appreciated. I have been testing the tuner with a simple ribbon wire aerial.

[Edward Huggins]

I wonder if when you switch the FM signal to Mono, it may sound cleaner, but you may only be hearing one channel? The build quality on these early-ish NADs had some batch variation and there may be dry joints. It depends on how important FM Stereo is to you as I would leave well alone if all the other functions are fine. It may be possible to remove the original multiplex module and replace it with a later NatSemi chip.

[qualityten]

Originally Posted by Edward Huggins

I wonder if when you switch the FM signal to Mono, it may sound cleaner, but you may only be hearing one channel?

Yes, that has occurred to me.

Originally Posted by Edward Huggins

The build quality on these early-ish NADs had some batch variation and there may be dry joints. It depends on how important FM Stereo is to you as I would leave well alone if all the other functions are fine. It may be possible to remove the original multiplex module and replace it with a later NatSemi chip.

I noted the build quality issue whilst replacing the burned out dial bulb. In view of the age of the unit I also replace the power filter capacitors, and noticed that one of the ceramic capacitors was only half soldered to the board.

All other functions are still fine. I may reflow the solder on the joints. FM Stereo is not particularly important to me, but I'd like to give this to a friend.

If the issue is more than a dry joint, should I consider replacing the IC FM stereo demodulator chip? It is an HA1156, and I see these are still available.

[paulsherwin]

Turn off the muting, disconnect the aerial and tune in a station weak enough to cause the tuner to drop to mono. Is the resultant signal the same with the mono switch in both positions?

It's worth sorting this out as the FM tuner is very good indeed for a mass market product. Distortion on one channel could very well be caused by a dry joint in the audio path.

Try removing the muting transistors.

[qualityten]

Thanks Paul. I've just read your threads about your fault finding work on an HK FM tuner.

I'm way out of my limited knowledge zone here, as my limited competence is in valve equipment, and even there the usual advice is to leave FM MPX circuits well alone.

As you say in your threads, these FM tuners are not of much value today, but this tuner is pretty good and integral to this receiver, so I'd like to see where this goes. Even without an aerial, it pulls in stations clearly in mono.

Following your advice, it does seem that the signal content is the same with the mono switch in and out.

[paulsherwin]

While most separate tuners aren't worth much, this one is because it is effectively a NAD 3020 amplifier as well. The tuner section is most of a NAD 4020, which I'm using as my main tuner at the moment (and very good it is too, despite being relatively simple).

So the channels are at an equal level both when switched to mono and when forced to mono because of a weak signal, but one channel becomes low and distorted when the signal increases and reception switches to stereo?

[qualityten]

Originally Posted by paulsherwin

So the channels are at an equal level both when switched to mono and when forced to mono because of a weak signal, but one channel becomes low and distorted when the signal increases and reception switches to stereo?

Not quite. When forced into mono with a weak signal I just get one channel. When I then tune a little more accurately, the other channel comes in, but weakly.

Pressing the mono switch delivers good sound to both channels, which (as far as I can make out) contains the full content in mono, not just one side of stereo.

But, as I say, I don't understand enough about how the mono signal comes in, as opposed to how the stereo signal is then derived. I've looked at a few sites explaining MPX, but I've not yet grasped it.

[paulsherwin]

The mono switch will be parallelling the channels.

My previous suggestions still stand - look for dry joints after the discriminator, and try removing the muting transistors. This is unlikely to be a chip fault.

[Radio Wrangler]

The zenith multiplex system isn't hard to understand if you've ever messed about with shortwave radio and SSB signals.

First and above all, the system had to win a beauty contest against some competing systems in order to get chosen for adoption as an American national standard. That meant it had to be fully compatible with old mono FM radios and to not impair their quality to any noticeable extent.

So the modulation seen within the bandwidth of an old mono receiver had to be the same as mono: Left plus Right.

All the added stuff for stereo had to be at modulating frequencies high above audibility, and would be stripped off by the filters of an old mono receiver.

What needed to be added amounts to the difference between the two channels. So left-right or right minus left is needed.

They used a simple analogue difference amplifier to compute L-R and then they needed to shift it up in frequency above the audio range. They used an analogue multiplier or frequency mixer, or balanced modulator (whichever you prefer to call it.) This turner the L-R audio into a Double-Sideband-Suppressed-Carrier (DSBSC) signal. They chose 38kHz to centre it on. DSBSC is easy to demodulate, just mix it with its own the carrier frequency, 38kHz in this case. But that 38kHz was suppressed and hasn't been sent along with the signal (to avoid distortion in the subsequent chain upsetting the mono compatibility by mixing some L-R into audio too soon) so they did the next best thing, they added a 19kHz signal, exactly half the 38kHz required.

So a MPX decoder does some filtering to separate-off the L+R mono audio, the 19kHz reference frequency (called the pilot tone) and the hight frequency difference stuff (about 22 to 54kHz).

The 19kHz tone is fed into a frequency doubler. At first this was a highly distorting amp stage driving a 38kHz filter. Later designs used a phase locked loop to lock a small oscillator onto the 19kHz, using a flip-flop frequency divide by 2.

Now armed with 38kHz, the DSBSC difference signal is mixed with the 38kHz and the L-R signal is re-created at its proper audio frequency range. A bit of filtering is good to remove other mixer products.

So now you have an L+R signal and an L-R signal.

Just add them together in a resistor network or with an opamp.

(L+R) + (L-R) = 2 x L and the R cancelled.

Hey, we've got the left channel

Now subtract them

(L+R) - (L-R) = 2 x R

OK, we've got both. Job done.

So a decoder consists of a frequency doubler, a mixer (= modulator = demodulator = multiplier) and a network adding and subtracting signals.

Notice that right up to the adder-subtractor circuit there is nothing that can spoil one channel without upsetting the other.

Anything doing so has to come downstream of this point.

David

[qualityten]

Thank you Paul and David. David, your explanation is really helpful.

Am I right in thinking that you are both steering me towards the buffer amps after de-emphasis switch and networks, and before the low pass filter?

Paul is the discriminator the HA1156 chip and the buffer amps what you call the muting transistors (BC549s on the schematic)? I wouldn't know how to remove them (bridge the base and collector?), but would know how to replace them.

Am I on the right track?

[Radio Wrangler]

Yes, If the amp plays CDs and records etc equally through both channels, then all the stuff after the selector switching must be OK.

The way Zenith Pilot Tone Multiplex to give it its full name means that everything up to the adding/subtracting 'matrix' must be OK.

So that leaves everything between them, and weeds out an awful lot of the machine. The scale of the job has come down to a few stages, and you have a good channel to compare against.

So what sort of equipment do you have? A multimeter's been mentioned, a small soldering iron will probably be needed. I need to find a schematic and stop talking in general terms.

David

[Radio Wrangler]

OK gotit! Thank-you electrotanya.

Before we start, have you looked around inside the unit and spoted where all the dangerous voltages are ? the mains input, fuses, on/off switch and voltage selector. You need to know where they are so you don't touch them. Some equipment covers and insulates them to protect service people, some doesn't. If you're not happy with this, go no further without help.

The stereo decoder is IC201 audio comes out of pins 4 and 5

W9a and SW9b switch in different de-emphasis timeconstants. It should stay in the 75us position unless you emigrate. But switch it through al the positions several times to clean the contacts. Sometimes these things get dirty and tarnish when left alone. Moving wipes the contacts and cleans them.

Q201 and Q202 are buffer amplifiers powered by a regulator transistor Q901

LPF is a pair of lowpass filters in a box, bought in as a subassembly, probably from TOKO. These sometimes fail. They're only needed if you want to make tape recordings from radio stations. Then it's a few resistors and nto the selector switch.

There are two ways to troubleshoot this sort of circuit, one is to probe and monitor a signal down the paths checking that you see what is expected. An audio millivoltmeter is used, and many digital multimeters are OK.

The other way is to inject a signal, maybe from an audio signl generator or music from some player, and to move the probe backwards up the chain to see where things go wrong.

You're lucky with a good channel to compare things with.

Access is going to be awkward. It's all in the centre of the board under a big junction of wiring bundles.

David

[qualityten]

Thanks David. I have the service manual, downloaded from the HiFi Engine website (for contents see post #1).

I'm aware of and confident about safety precautions, with about 15 years' hobbyist experience working on valve amplifiers. I fixed a dead Trio KA4006 some years ago with remote help from another forum, but I've no prior experience of radio/tuner circuits.

I have cleaned all the switches and pots with Servisol.

I can see the components you are describing, both on the circuit diagram and on the receiver itself and have posted a few pics so you can see the components on the board. Access doesn't look too bad, as the wires can be moved out of the way. However, they are brittle. One just came away when I was replacing the power capacitors.

Equipment wise, I have two good autoranging digital multimeters, a Tektronix 422 oscilloscope (little used, though I have used it to trace audio frequency signals in the past) and a Levell RC Oscillator TG200 signal generator. Also a Weller WTCP 50 soldering iron with a fine tip.

Given this equipment, which would be the better faultfinding route? Am I looking for dry joints or failed transistors?

[qualityten]

A couple of pics. The second shows Q201 and Q202 at the top, on either side of the pair of 100uF capacitors C220 and C221, and the LPF is on the right, lower middle.

[paulsherwin]

I agree that the board is a bit crowded.

You should be able to remove the muting transistors by just desoldering them. The transistors themselves are unlikely to be faulty, but if you remove them and everything starts working then the transistors or their associated control circuitry will be responsible.

[qualityten]

I've just inspected the solder side of the PCB. There is no evidence that it has been opened since assembly, but the traces in the area look messed up,coming from SW9a and 9b. Here is a pic. I guess it's important because this is the path of the audio channel from pins 4 and 5 of IC201. Is it possible, this stereo FM channel has always been faulty, or is this just typical NAD rough assembly?

[qualityten]

Paul, can I check terminology? Are the buffer amplifiers also the muting transistors, and is their function to mute the FM between stations? Thus, are you pointing to these because you suspect that one of the transistors, or the associated circuitry, is faulty and thus permanently muting one channel? Just checking that I understand the function of this part of the circuit.

Thanks for all the help so far.

[cmjones01]

That looks pretty grotty. It's probably just some soldering flux left over from assembly. It has been known for old flux to become a bit conductive and thus affect the operation of the circuit. I'd start by giving it a good clean. Car brake cleaner and an old toothbrush is my favourite recipe. As far as I can tell it's the same stuff as expensive PCB flux remover but miles cheaper! Others may suggest different solvents and I offer no guarantee.

Chris

[paulsherwin]

I don't have a circuit, but I suspect the 'buffer transistors' will be the muting transistors. Their function is to mute the output when the main chip indicates a low signal which is likely to be noisy. The control circuitry may be more or less complex, but removing the transistors will reveal if there is a problem with the muting circuitry if everything then starts working (apart from the muting, obviously).

[Radio Wrangler]

Hmm, definitely worse than manky. Definitely grotty, possibly as bad as NADgered.

Looks like a very fruitful area for a general clean up, and didn't I see on the topside a pair of those transistors with a ceramic base and a blob of plastic on the top like the ones what the legs drop off?

It sounds like you're properly tooled-up for the job in hand, and well aware of the risks. My conscience feels better already!

If a clean up and a check of those transistors doesn't fix it, then my choice would be to inject a trace of signal.

One good cheat is to use a 10:1 scope probe. plug it onto the BNC output of an audio generator and use the probe tip to explore various places. It's effectively a 9 meg resisor so 9v from your generator injects a microamp which is usually enough to be noticed. This trick works at RF too.

David

Cross-posted with Paul. They're not the usual shunt muting transistors, they are amplifiers carrying the signal path, so pulling them kills the signal, not unmutes it.