Hey guys,
This amp has started misbehaving. I've had for a couple of years and it has worked flawlessly until now. It was recapped with Panasonic caps a couple of years ago, and idle current and dc offset has been set properly.
A few weeks ago I noticed that the RIAA stage was noisy so I went in there and touched up some solder joints, which fixed the problem. While doing this I also replaced the input capacitors in the power amp to film caps from Wima. This may be a cause of this problem, but if so it's kinda odd that it's happening now and not right after.
Anyway, how it first started - one channel started getting "hissy", which was really apparent when listening to calm music. It was not nice on the ears at all.
I checked it with a scope today. With no load, the waveform looks perfect.
With headphones plugged in, I got this almost right away:
This sounds like when the stereo picks up interference from an incoming cellphone signal. I'm sure you know that sound.
I disconnected the headphones and went away for a little while. When I came back I plugged them back in, and now the problem was not just on the left channel any more. Now we have a brand new problem - and on both channels.
This sounds like a slow, calm fishing boat motor on a windless lake. With cellphone interference in the background. Even with no load - the motor sound is there (but not the cellphone noise).
I was originally rather confident in fixing this, but when the new problem appeared - I'm super lost. What could this possibly be?
If you want to take a look at the schematic, here's the power amp part:
If that's not enough, here's the entire service manual: www.hifiengine.com/manual_library/luxman/l-510.shtml
Thanks, man. 🙂
This amp has started misbehaving. I've had for a couple of years and it has worked flawlessly until now. It was recapped with Panasonic caps a couple of years ago, and idle current and dc offset has been set properly.
A few weeks ago I noticed that the RIAA stage was noisy so I went in there and touched up some solder joints, which fixed the problem. While doing this I also replaced the input capacitors in the power amp to film caps from Wima. This may be a cause of this problem, but if so it's kinda odd that it's happening now and not right after.
Anyway, how it first started - one channel started getting "hissy", which was really apparent when listening to calm music. It was not nice on the ears at all.
I checked it with a scope today. With no load, the waveform looks perfect.
With headphones plugged in, I got this almost right away:
This sounds like when the stereo picks up interference from an incoming cellphone signal. I'm sure you know that sound.
I disconnected the headphones and went away for a little while. When I came back I plugged them back in, and now the problem was not just on the left channel any more. Now we have a brand new problem - and on both channels.
This sounds like a slow, calm fishing boat motor on a windless lake. With cellphone interference in the background. Even with no load - the motor sound is there (but not the cellphone noise).
I was originally rather confident in fixing this, but when the new problem appeared - I'm super lost. What could this possibly be?
If you want to take a look at the schematic, here's the power amp part:
If that's not enough, here's the entire service manual: www.hifiengine.com/manual_library/luxman/l-510.shtml
Thanks, man. 🙂
Wow, I need to pay more attention. That electrolytic capacitor I replaced with a film cap wasn't the input cap at all - there is no input cap! What I replaced was sitting in the feedback loop. I guess mucking around with that could cause instability, right? I better put the electrolytic back and see what happens.
The cap in question is C417 on the schematic.
The cap in question is C417 on the schematic.
Some of those waveforms look more like artifacts of the "scope" tbh. They don't look real. Whatever the amp is doing the timeline of the sinewave test signal is constant... and yours isn't.
The spikes look plausible though.
The spikes look plausible though.
Your last scope plots reveal "spontaneous phase reversal" of the signal imho.
Very strange indeed
Very strange indeed
I tested the amp this morning (for about a minute only) and it worked fine, with no scope attached and headphones for audio. I think the scope may be screwing with the amp somehow. The spike/distortion seen on the first pic however, will likely come back after the thing has warmed up a bit.
The last pic was taken with no load connected - just the scope. The amp was then motorboating, or something such. Maybe the scope couldn't correctly display that?
The last pic was taken with no load connected - just the scope. The amp was then motorboating, or something such. Maybe the scope couldn't correctly display that?
I think you need to carefully check all the work that has been done with regard to all the cap changes. To much of a coincidence for both channels to be faulty usnless there is some common PSU issue to both.
To clarity: I only had problems with one channel that was noisy. When I connected the scope and did some testing, eventually the entire amp became unstable. This has never been the case before.
The two capacitors (one on each channel) were replaced some weeks ago, and the amp was in service the entire time since then - and worked flawlessly. Only recently did this happen.
The PSU is only two capacitors and bridge rectifier. Nothing else going on there. There is a more advanced part, but that's for the RIAA stage and speaker protection circuit.
The two capacitors (one on each channel) were replaced some weeks ago, and the amp was in service the entire time since then - and worked flawlessly. Only recently did this happen.
The PSU is only two capacitors and bridge rectifier. Nothing else going on there. There is a more advanced part, but that's for the RIAA stage and speaker protection circuit.
I've no quick fix I'm afraid. I would begin by setting the scope (dual channel) to look at the input and output of one channel simultaneously both under load and not under load. You also need to check the rails on a scope (but make sure it can handle the voltages involved first.
If both channels are faulty then also compare outputs of both simultaneously using a mono source as the input.
If both channels are faulty then also compare outputs of both simultaneously using a mono source as the input.
If you are testing the amplifier with the input disconnected - perhaps as it appears on the schematic with the pre-out/main-in switch open, the amplifier would be exposed to RFI pickup as there are no filters at the input and any leads connected, even the scope could be a source of noise. I'd look at making up a standard input filter arrangement before paying much attention to the low level noise, as it seems to me.
The scope traces are unclear, they don't show a clear fault exactly.
The lower trace shows signs of a bias issue, and of course the glitch.
But what that comes from isn't clear since you also show some sort of noise riding the signal. But since it is a digital scope you have so many artifacts that low level stuff like noise riding a signal is hard to see.
First idea is to take the input signal and the output signal, sum and invert one channel, then you have the difference (once the levels are matched in the scope). That will show if there is something specific going on, perhaps.
Sometimes transistors will break down by first getting noisy, then later doing something akin to "arcing" wherein they will work and then not work, this gives a pulse for the output. They easy way to diagnose this, is to get the amp to "do it" and then hit each transistor selectively with "freeze spray". When you find the bad transistor, the problem stops for the moment, until it gets hot again.
_-_-
The lower trace shows signs of a bias issue, and of course the glitch.
But what that comes from isn't clear since you also show some sort of noise riding the signal. But since it is a digital scope you have so many artifacts that low level stuff like noise riding a signal is hard to see.
First idea is to take the input signal and the output signal, sum and invert one channel, then you have the difference (once the levels are matched in the scope). That will show if there is something specific going on, perhaps.
Sometimes transistors will break down by first getting noisy, then later doing something akin to "arcing" wherein they will work and then not work, this gives a pulse for the output. They easy way to diagnose this, is to get the amp to "do it" and then hit each transistor selectively with "freeze spray". When you find the bad transistor, the problem stops for the moment, until it gets hot again.
_-_-
Ian: I had input, I was using my phone as a signal generator (which I've had great success with many times in the past).
Bear: Bias issue you say? How so? I think I'll try setting it again (even though I did that not a year ago).
Some news: I did some listening today (with speakers) and of course it works almost perfectly now. I DID get that distortion thing seen on the first pic for a short, short while upon startup, but then everything became normal again. I could replicate this once by turning it off and then on again.
EDIT: To be clear, this is not noise I'm showing. I was playing a 200Hz sine wave through the amp. The low voltages are simply because I wasn't playing very loud. 🙂 I DID hear that 200Hz sine wave through the headphones - but it was slightly broken on one channel (first pic), and then later on turned into motorboating (pic 2 and 3).
EDIT2: Ah, got it again. It's exactly like in the first pic. Some horrible noise on one channel on the negative part of the waveform. It only occurs when it feels like it. Usually a little while after restarting the thing. I'm going to go in and check out some transistors I think!
Bear: Bias issue you say? How so? I think I'll try setting it again (even though I did that not a year ago).
Some news: I did some listening today (with speakers) and of course it works almost perfectly now. I DID get that distortion thing seen on the first pic for a short, short while upon startup, but then everything became normal again. I could replicate this once by turning it off and then on again.
EDIT: To be clear, this is not noise I'm showing. I was playing a 200Hz sine wave through the amp. The low voltages are simply because I wasn't playing very loud. 🙂 I DID hear that 200Hz sine wave through the headphones - but it was slightly broken on one channel (first pic), and then later on turned into motorboating (pic 2 and 3).
EDIT2: Ah, got it again. It's exactly like in the first pic. Some horrible noise on one channel on the negative part of the waveform. It only occurs when it feels like it. Usually a little while after restarting the thing. I'm going to go in and check out some transistors I think!
Last edited:
You aren't connecting your scope probe ground to the input ground side are you? This would short out R444 56R via a long earth loop and may cause noises/instability.
No, I had the scope probes on the speaker terminals.
The scope's two negative probes are connected together internally, but so are the two negative speaker terminals (at least it seems like it by doing a continuity test with a multimeter - and the schematic indicates that as well).
The scope's two negative probes are connected together internally, but so are the two negative speaker terminals (at least it seems like it by doing a continuity test with a multimeter - and the schematic indicates that as well).
Potentially fixed it.
I didn't have any cooling spray, so I used a hairdryer set to cold and blew around in there. Almost immediately got the problem started - by directing the flow towards Q401/402. Cooling those down did the trick. It would then fade away after a little while. Setting the dryer to heat took away the problem instantly.
So I replaced them. I had exact replacements on hand - they were 2SA992, I had KSA992 - a clone part from Fairchild. Can't replicate the problem any more.
Of course, it'll probably be back as soon as I stitch the thing back together... 😉
I didn't have any cooling spray, so I used a hairdryer set to cold and blew around in there. Almost immediately got the problem started - by directing the flow towards Q401/402. Cooling those down did the trick. It would then fade away after a little while. Setting the dryer to heat took away the problem instantly.
So I replaced them. I had exact replacements on hand - they were 2SA992, I had KSA992 - a clone part from Fairchild. Can't replicate the problem any more.
Of course, it'll probably be back as soon as I stitch the thing back together... 😉
- Status
- Not open for further replies.