NAD 7020 weak channel on FM stereo - Page 3 - UK Vintage Radio Repair and Restoration Discussion Forum

qualityten · 14th Dec 2016, 11:39 am

Have just downloaded the schematic in my tea break! Are the decoupling caps those rated at 0.1uF, i.e. C633 (and C634) and C627 (and C628)? The only other large one, 47uF, is C625 (and C626). The others in that area are much smaller (1n, 22p, 100p).

Radio Wrangler · 14th Dec 2016, 1:50 pm

I'm away from my home computer which has the schematic on it.

No, there will be an awful lot of supply decoupling capacitors. You have a feedback loop (not an intentional feedback loop, one formed by coupling through ripples on supplies) which is oscillating. The -20dB switch affecting it makes me suspect it is fluctuations from the power amps getting back into preamp stages.

Ceramic and film 0.1uF 22p 100p should be fine. It's old electrolytics I'd go hunting first.

David

qualityten · 14th Dec 2016, 1:54 pm

Lunch-break further thought. Just read about decoupling caps. As they decouple the circuit from the power supply to smooth out the power supply voltage, then I realise that I should check all the caps in the power amp.

Sorry, this post was sent before I read yours above David. I will go electrolytic checking.

paulsherwin · 14th Dec 2016, 2:05 pm

You need to look for big ones - 47uF and greater.

I'm generally very critical of the obsession with changing electrolytics without a good reason in this class of equipment, but this may be one of the cases where a mass recapping is warranted. It's a big job though, particularly if you just plan to give it away to a friend.

qualityten · 14th Dec 2016, 5:30 pm

Thanks Paul. I'll check/change the two 47uF caps for a start.

qualityten · 16th Dec 2016, 4:53 pm

Quote:

Originally Posted by paulsherwin

You need to look for big ones - 47uF and greater.

I'm generally very critical of the obsession with changing electrolytics without a good reason in this class of equipment, but this may be one of the cases where a mass recapping is warranted. It's a big job though, particularly if you just plan to give it away to a friend.

Well, I'm a bit puzzled about where the soft motorboating is coming from. I've replaced almost all the capactors in the power supply, power amp and tone amplifier, and not found any more defective electrolytic caps.

I've deliberately not checked any of the ceramic or film caps. Should I suspect them, or their solder joints? This feels a bit hit and miss. Is there a better way to locate the source of such oscillation?

Apart from this, all functions and parts work fine.

Radio Wrangler · 16th Dec 2016, 7:23 pm

A dc coupled scope could show which rail has a synchronous twitch. It's worth looking at the bias conditions of everything even sections not in use. It could also be a pulsating RF oscillation (squegging)

David

qualityten · 16th Dec 2016, 8:08 pm

Thanks David. I'll check voltages with reference to the circuit diagram.

paulsherwin · 16th Dec 2016, 9:29 pm

You could just live with it if it's not intrusive. Eventually it'll go bad and be much easier to diagnose.

qualityten · 16th Dec 2016, 11:30 pm

Living with it is quite possible, as it disappears when the low level switch is engaged, and I can use a different (and less used) part of the volume pot.

But, as I've replaced most of the electrolytics, I'll probably complete the job. As we know, it's a bit like doing a jigsaw puzzle. Not very taxing and quite relaxing. Also, learning is very much part of this hobby.

Goldie99 · 17th Dec 2016, 7:51 am

Did you already change C601-604 ? and is the issue audible in both channels ? also using lab input ?

qualityten · 17th Dec 2016, 11:11 am

Thanks for your suggestions Goldie99.

Yes, I replaced C601-604 before I started on the FM stereo issue. Yes, the motorboating is audible via Lab In too.

But I can't understand why it disappears when the Low Level switch is engaged and volume is increased.

Goldie99 · 17th Dec 2016, 6:19 pm

And, is it still there if you just disconnect the pre-out / normal- (or lab-) in jumpers ? or use another pre-amp - connected at the normal- (or lab-) inputs ?

If so, I think that would pretty much localise the fault to the power amps / their supplies.

qualityten · 17th Dec 2016, 11:32 pm

If I disconnect the jumpers, it disappears. I guess this localises it to the preamp section and supplies?

Radio Wrangler · 17th Dec 2016, 11:48 pm

'fraid not. This sort of thing is a loop. It can be an influence from the power amp getting back into any of the preamp sections.

Looking through the schematic, I wonder about a power supply regulator instability?

David

qualityten · 17th Dec 2016, 11:59 pm

Is there any clue to the location of the loop in the fact that it disappears when the low level switch is engaged?

The user manual states that this button provides 'optimum signal/noise ratio' for low level listening; 'it minimises the already-low residual noise of the preamp and tone control circuits'.

If instability in the power supply regulator, what components would be suspect? Sorry that my understanding of the circuit is so limited.

paulsherwin · 18th Dec 2016, 1:51 am

I think it's just an attenuator before the power amp. Are you sure that the noise is actually disappearing with the circuit enabled, rather than just reducing in audibility?

As David says, if this is a barely perceptible instability problem then it's likely to be a pig to sort out. Have you tried scoping the power and signal lines around the volume control and 'low level' circuit?

Radio Wrangler · 18th Dec 2016, 8:43 am

There are some secrets to the business of oscillators, known only to mages of the seventh order and above, though they ought to be obvious enough to anyone who stops to think a little.

Books and books have been written on the frequency control and frequency stability of oscillators and the naming of all the various circuits.

Damn-all has been written on their amplitude control mechanisms. This is odd, because this has to be right or else oscillation won't start, or else it increases until voltages hammer into the power rails.

An oscillator is an unstable circuit, showing positive feedback. If the loop gain is less than 1, the fed-back signal circling the loop gets smaller with each round trip, and it dies out. If something gives it a bit of a kick, it twangs and fades.

If the loop gain is even the tiniest bit greater than 1, the signal grows each round trip it makes. Such a circuit doesn't need a kick start, it'll start off its own noise. The amplitude will grow and grow until it hits a limit.

Good oscillator design means engineering that limit, so that as the oscillation level increases, the gain starts to reduce, and falls below the magic 1.0000000000.. So the oscillator has enough gain to start reliably, and then stabilises itself at a comfortable running level.

With transistors, there is a strong non-linearity with a base-emitter voltage of about 0.7 volts. This sort of sets a scale for the levels of oscillators you can make. Oscillators tend to run with at least 1 volt amplitude somewhere in the circuit, often more. It would be a very difficult trick and require a lot of added hardware to make an oscillator stabilise with only a few millivolts or microvolts of amplitude.

So the point of the long preamble is that:

1) Somewhere in your amplifier, that low frequency oscillation is likely to have a large enough amplitude to spot on a scope

2) If the loop includes preamp and power amp stages, then switching in a 20dB attenuator cuts the loop gain dramatically (voltage gain is reduced tenfold) and oscillation is killed.

If it's a power supply voltage regulator sulking, then as Paul says, that 20dB attenuator may just cut the level of the sound below noticeability.

Even if you find a place with a viewable oscillation, it can be difficult to decide what the oscillation loop is and difficult to deduce which components are probable causes.

Sometimes a perfectly respectable repair tactic is just to turn your brain off and get on with mindlessly swapping things until you hit it. (Just don't admit it to paying customers)

Oh, and one final gotcha, don't assume the thing was competently designed in the first place! They might all be close to instability in this way. Sometimes tracking a fault finds a design blooper. The audiophile stuff which was 'designed' on religious grounds rather than engineering mathematics by people who listen to components but never measure them (they must sound bad if they've been contaminated by scientific method) are especially fruitful grounds for finding schoolboy howlers.

David

qualityten · 18th Dec 2016, 10:48 am

David and Paul, thank you for your kindness in responding to my puzzlement. I'll get my scope out when I have a break after Christmas! Meanwhile, I'll continue the mindless recapping when I have a moment.

Goldie99 · 18th Dec 2016, 2:35 pm

I'd hoped the 'search' could have been narrowed down a bit, but from David & Pauls' comments (both FAR better informed than my own), it sounds like you're in for the long haul with this one - unless the full recap fixes it.

Please let us know how it's looking as/when you get to it - I'd be interested - I'm just starting on a NAD 3120 with a few quirks and cap issues of its own.

David / Paul - thanks also for the explanation of the low level switch effect, that was puzzling me.

14th Dec 2016, 11:39 am	#41
qualityten Octode Join Date: Feb 2009 Location: Wallasey, UK. Posts: 1,314	Re: NAD 7020 weak channel on FM stereo Have just downloaded the schematic in my tea break! Are the decoupling caps those rated at 0.1uF, i.e. C633 (and C634) and C627 (and C628)? The only other large one, 47uF, is C625 (and C626). The others in that area are much smaller (1n, 22p, 100p).

14th Dec 2016, 1:50 pm	#42
Radio Wrangler Moderator Join Date: Mar 2012 Location: Fife, Scotland, UK. Posts: 22,902	Re: NAD 7020 weak channel on FM stereo I'm away from my home computer which has the schematic on it. No, there will be an awful lot of supply decoupling capacitors. You have a feedback loop (not an intentional feedback loop, one formed by coupling through ripples on supplies) which is oscillating. The -20dB switch affecting it makes me suspect it is fluctuations from the power amps getting back into preamp stages. Ceramic and film 0.1uF 22p 100p should be fine. It's old electrolytics I'd go hunting first. David __________________ Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done

14th Dec 2016, 1:54 pm	#43
qualityten Octode Join Date: Feb 2009 Location: Wallasey, UK. Posts: 1,314	Re: NAD 7020 weak channel on FM stereo Lunch-break further thought. Just read about decoupling caps. As they decouple the circuit from the power supply to smooth out the power supply voltage, then I realise that I should check all the caps in the power amp. Sorry, this post was sent before I read yours above David. I will go electrolytic checking.

14th Dec 2016, 2:05 pm	#44
paulsherwin Moderator Join Date: Jun 2003 Location: Oxford, UK Posts: 27,970	Re: NAD 7020 weak channel on FM stereo You need to look for big ones - 47uF and greater. I'm generally very critical of the obsession with changing electrolytics without a good reason in this class of equipment, but this may be one of the cases where a mass recapping is warranted. It's a big job though, particularly if you just plan to give it away to a friend.

14th Dec 2016, 5:30 pm	#45
qualityten Octode Join Date: Feb 2009 Location: Wallasey, UK. Posts: 1,314	Re: NAD 7020 weak channel on FM stereo Thanks Paul. I'll check/change the two 47uF caps for a start.

16th Dec 2016, 7:23 pm	#47
Radio Wrangler Moderator Join Date: Mar 2012 Location: Fife, Scotland, UK. Posts: 22,902	Re: NAD 7020 weak channel on FM stereo A dc coupled scope could show which rail has a synchronous twitch. It's worth looking at the bias conditions of everything even sections not in use. It could also be a pulsating RF oscillation (squegging) David __________________ Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done

16th Dec 2016, 8:08 pm	#48
qualityten Octode Join Date: Feb 2009 Location: Wallasey, UK. Posts: 1,314	Re: NAD 7020 weak channel on FM stereo Thanks David. I'll check voltages with reference to the circuit diagram.

16th Dec 2016, 9:29 pm	#49
paulsherwin Moderator Join Date: Jun 2003 Location: Oxford, UK Posts: 27,970	Re: NAD 7020 weak channel on FM stereo You could just live with it if it's not intrusive. Eventually it'll go bad and be much easier to diagnose.

16th Dec 2016, 11:30 pm	#50
qualityten Octode Join Date: Feb 2009 Location: Wallasey, UK. Posts: 1,314	Re: NAD 7020 weak channel on FM stereo Living with it is quite possible, as it disappears when the low level switch is engaged, and I can use a different (and less used) part of the volume pot. But, as I've replaced most of the electrolytics, I'll probably complete the job. As we know, it's a bit like doing a jigsaw puzzle. Not very taxing and quite relaxing. Also, learning is very much part of this hobby.

17th Dec 2016, 7:51 am	#51
Goldie99 Heptode Join Date: Jul 2014 Location: Boston, Lincolnshire, UK. Posts: 995	Re: NAD 7020 weak channel on FM stereo Did you already change C601-604 ? and is the issue audible in both channels ? also using lab input ?

17th Dec 2016, 11:11 am	#52
qualityten Octode Join Date: Feb 2009 Location: Wallasey, UK. Posts: 1,314	Re: NAD 7020 weak channel on FM stereo Thanks for your suggestions Goldie99. Yes, I replaced C601-604 before I started on the FM stereo issue. Yes, the motorboating is audible via Lab In too. But I can't understand why it disappears when the Low Level switch is engaged and volume is increased.

17th Dec 2016, 6:19 pm	#53
Goldie99 Heptode Join Date: Jul 2014 Location: Boston, Lincolnshire, UK. Posts: 995	Re: NAD 7020 weak channel on FM stereo And, is it still there if you just disconnect the pre-out / normal- (or lab-) in jumpers ? or use another pre-amp - connected at the normal- (or lab-) inputs ? If so, I think that would pretty much localise the fault to the power amps / their supplies.

17th Dec 2016, 11:32 pm	#54
qualityten Octode Join Date: Feb 2009 Location: Wallasey, UK. Posts: 1,314	Re: NAD 7020 weak channel on FM stereo If I disconnect the jumpers, it disappears. I guess this localises it to the preamp section and supplies?

17th Dec 2016, 11:48 pm	#55
Radio Wrangler Moderator Join Date: Mar 2012 Location: Fife, Scotland, UK. Posts: 22,902	Re: NAD 7020 weak channel on FM stereo 'fraid not. This sort of thing is a loop. It can be an influence from the power amp getting back into any of the preamp sections. Looking through the schematic, I wonder about a power supply regulator instability? David __________________ Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done

17th Dec 2016, 11:59 pm	#56
qualityten Octode Join Date: Feb 2009 Location: Wallasey, UK. Posts: 1,314	Re: NAD 7020 weak channel on FM stereo Is there any clue to the location of the loop in the fact that it disappears when the low level switch is engaged? The user manual states that this button provides 'optimum signal/noise ratio' for low level listening; 'it minimises the already-low residual noise of the preamp and tone control circuits'. If instability in the power supply regulator, what components would be suspect? Sorry that my understanding of the circuit is so limited.

18th Dec 2016, 1:51 am	#57
paulsherwin Moderator Join Date: Jun 2003 Location: Oxford, UK Posts: 27,970	Re: NAD 7020 weak channel on FM stereo I think it's just an attenuator before the power amp. Are you sure that the noise is actually disappearing with the circuit enabled, rather than just reducing in audibility? As David says, if this is a barely perceptible instability problem then it's likely to be a pig to sort out. Have you tried scoping the power and signal lines around the volume control and 'low level' circuit?

18th Dec 2016, 8:43 am	#58
Radio Wrangler Moderator Join Date: Mar 2012 Location: Fife, Scotland, UK. Posts: 22,902	Re: NAD 7020 weak channel on FM stereo There are some secrets to the business of oscillators, known only to mages of the seventh order and above, though they ought to be obvious enough to anyone who stops to think a little. Books and books have been written on the frequency control and frequency stability of oscillators and the naming of all the various circuits. Damn-all has been written on their amplitude control mechanisms. This is odd, because this has to be right or else oscillation won't start, or else it increases until voltages hammer into the power rails. An oscillator is an unstable circuit, showing positive feedback. If the loop gain is less than 1, the fed-back signal circling the loop gets smaller with each round trip, and it dies out. If something gives it a bit of a kick, it twangs and fades. If the loop gain is even the tiniest bit greater than 1, the signal grows each round trip it makes. Such a circuit doesn't need a kick start, it'll start off its own noise. The amplitude will grow and grow until it hits a limit. Good oscillator design means engineering that limit, so that as the oscillation level increases, the gain starts to reduce, and falls below the magic 1.0000000000.. So the oscillator has enough gain to start reliably, and then stabilises itself at a comfortable running level. With transistors, there is a strong non-linearity with a base-emitter voltage of about 0.7 volts. This sort of sets a scale for the levels of oscillators you can make. Oscillators tend to run with at least 1 volt amplitude somewhere in the circuit, often more. It would be a very difficult trick and require a lot of added hardware to make an oscillator stabilise with only a few millivolts or microvolts of amplitude. So the point of the long preamble is that: 1) Somewhere in your amplifier, that low frequency oscillation is likely to have a large enough amplitude to spot on a scope 2) If the loop includes preamp and power amp stages, then switching in a 20dB attenuator cuts the loop gain dramatically (voltage gain is reduced tenfold) and oscillation is killed. If it's a power supply voltage regulator sulking, then as Paul says, that 20dB attenuator may just cut the level of the sound below noticeability. Even if you find a place with a viewable oscillation, it can be difficult to decide what the oscillation loop is and difficult to deduce which components are probable causes. Sometimes a perfectly respectable repair tactic is just to turn your brain off and get on with mindlessly swapping things until you hit it. (Just don't admit it to paying customers) Oh, and one final gotcha, don't assume the thing was competently designed in the first place! They might all be close to instability in this way. Sometimes tracking a fault finds a design blooper. The audiophile stuff which was 'designed' on religious grounds rather than engineering mathematics by people who listen to components but never measure them (they must sound bad if they've been contaminated by scientific method) are especially fruitful grounds for finding schoolboy howlers. David __________________ Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done

18th Dec 2016, 10:48 am	#59
qualityten Octode Join Date: Feb 2009 Location: Wallasey, UK. Posts: 1,314	Re: NAD 7020 weak channel on FM stereo David and Paul, thank you for your kindness in responding to my puzzlement. I'll get my scope out when I have a break after Christmas! Meanwhile, I'll continue the mindless recapping when I have a moment.

18th Dec 2016, 2:35 pm	#60
Goldie99 Heptode Join Date: Jul 2014 Location: Boston, Lincolnshire, UK. Posts: 995	Re: NAD 7020 weak channel on FM stereo I'd hoped the 'search' could have been narrowed down a bit, but from David & Pauls' comments (both FAR better informed than my own), it sounds like you're in for the long haul with this one - unless the full recap fixes it. Please let us know how it's looking as/when you get to it - I'd be interested - I'm just starting on a NAD 3120 with a few quirks and cap issues of its own. David / Paul - thanks also for the explanation of the low level switch effect, that was puzzling me.