help diagnose power amp buzzing from one channel

wasabii

Member
2021-11-30 10:48 pm
This intermittent buzzing comes and goes at random. I may catch it for 5 or 10 seconds once in a month, or as often as every hour or two for the entire length of my listening session. It starts then stops and all seems well. The amp sounds ok otherwise.

A few of the symptoms are as follows...

Right channel only
Does not get louder when volume is increased. Its not even noticeable when the volume is up. I can't even tell its occurring until I decrease the volume or catch it between tracks.

It stops and starts at random. Its happening more frequently today, still very sporadically. Yesterday during my listen I never heard it once.

I'm hoping its something simple, like a loose ground or inferior cap inside. I can't afford to pay a tech and will be system less if I'm not able to track down myself. I'm not sure if I'm pushing my luck by using the amp before it can be resolved. Hopfully more damage doesn't result. I can't afford to replace a 4k amp right now. I can't even afford to pay a tech to look at it.

I was hoping someone could possibly offer some kind of check list that might help me narrow it down?

Thanks in advance.
 

wasabii

Member
2021-11-30 10:48 pm
This amp doesn't have speaker relay's. The power goes directly to the speakers as soon as its switched on without delay.

I think the idea is to achieve the purest signal paths so anything that wasn't necessary wasn't incorporated in the design.
 
No brand no model no technology, no production date. No schematic diagram. No help.
You can trace the origin of buzzing noises with an oscilloscope, a sound probe, or an analog Voltmeter. The cheapest is a $20 analog VOM with a .047 uf >100 v capacitor series the negative probe. Connect the negative probe to the speaker ground with alligator clip leads. Use 50 vac scale after the VAS, 20 vac scale before.
The stage where the buzzing starts, that is where you
Warning, use only one hand at a time with the power on. Voltage over 24 across the heart can stop it. Wear no jewelry on hands wrists or neck. 1 v at 30 amps through a ring can burn your flesh to charcoal. Wear safety glasses. Exploding parts can blow metal to the ceiling. Solder splashes especially unsoldering.
 
The problem is in the power amplifier section of the right hand channel. That narrows done the possibilities quite a lot. Assuming the unit is a conventional post 1970 mass produced integrated amplifier, the source of the problem will not be in the wiring or controls, but will be on the circuit board The problem is intermittent. Most intermittent problems are faulty joints. On a circuit board most faulty joints are 'dry joints' AKA failed solder joints, or very occasionally in some '70s and '80s gear oxidised 'wire wrap' posts used for connections to the circuit board.

Dry joints can be identified by their mottled appearance (rather than shiny) or by hairline cracks through the solder - you might need a magnifying glass to see. They are most often either a result of poor soldering at time of manufacture, or by heat from hot components that were not stood off the circuit board far enough. Look for discoloured (toasted) PCB substrate too see if excess heat is involved. Intermittent connection can also result from hairline cracks in the printed circuit, which is often impossible to see.

The simplest and fastest way to approach a problem like yours is to simply re-solder the entire part of the circuit at fault. Identify the power amplifier section of the right hand channel and retouch all of the solder pads. If that fixes it, great! When you do the re-solder the chances are you will see what caused the problem. If the fault was heat related solder breakdown, the other channel is likely not far behind so re-solder that one as well.

If re-soldering doesn't fix it, then the next step is to start prodding around the circuit board with a chopstick (best to use a dim bulb current limiter and crappo speakers when doing this). Flex the circuit board by pressing on it, push against each component standing off the circuit board, and tap the circuit board all over the place and see if you can make the fault come and go, then narrow down to the most sensitive location. Good luck!
 

wasabii

Member
2021-11-30 10:48 pm
The problem is in the power amplifier section of the right hand channel. That narrows done the possibilities quite a lot. Assuming the unit is a conventional post 1970 mass produced integrated amplifier, the source of the problem will not be in the wiring or controls, but will be on the circuit board The problem is intermittent. Most intermittent problems are faulty joints. On a circuit board most faulty joints are 'dry joints' AKA failed solder joints, or very occasionally in some '70s and '80s gear oxidised 'wire wrap' posts used for connections to the circuit board.

Dry joints can be identified by their mottled appearance (rather than shiny) or by hairline cracks through the solder - you might need a magnifying glass to see. They are most often either a result of poor soldering at time of manufacture, or by heat from hot components that were not stood off the circuit board far enough. Look for discoloured (toasted) PCB substrate too see if excess heat is involved. Intermittent connection can also result from hairline cracks in the printed circuit, which is often impossible to see.

The simplest and fastest way to approach a problem like yours is to simply re-solder the entire part of the circuit at fault. Identify the power amplifier section of the right hand channel and retouch all of the solder pads. If that fixes it, great! When you do the re-solder the chances are you will see what caused the problem. If the fault was heat related solder breakdown, the other channel is likely not far behind so re-solder that one as well.

If re-soldering doesn't fix it, then the next step is to start prodding around the circuit board with a chopstick (best to use a dim bulb current limiter and crappo speakers when doing this). Flex the circuit board by pressing on it, push against each component standing off the circuit board, and tap the circuit board all over the place and see if you can make the fault come and go, then narrow down to the most sensitive location. Good luck!
Its a Madrigal designed and built Mark Levinson. True balanced/ dual monaural design. Essentially its a ML in a different case. The only think that suggest its a ML on the outer case is the colours. https://skyfiaudio.com/products/proceed-from-mark-levinson-amp2-amplifier

It sounds stunning otherwise. I haven't managed to find a tech who will touch Mark Levinson in Canada thus far. As it stands its beginning to look like an expensive boat anchor. Unless I use the other channel for a centre or a sub. But then lose out on the mid and high range glory its capable of.

I will include more detail and be more thorough when I feel better. Currently under the weather and finding it hard to concentrate these past few days..
 

wasabii

Member
2021-11-30 10:48 pm
The problem is in the power amplifier section of the right hand channel. That narrows done the possibilities quite a lot. Assuming the unit is a conventional post 1970 mass produced integrated amplifier, the source of the problem will not be in the wiring or controls, but will be on the circuit board The problem is intermittent. Most intermittent problems are faulty joints. On a circuit board most faulty joints are 'dry joints' AKA failed solder joints, or very occasionally in some '70s and '80s gear oxidised 'wire wrap' posts used for connections to the circuit board.

Dry joints can be identified by their mottled appearance (rather than shiny) or by hairline cracks through the solder - you might need a magnifying glass to see. They are most often either a result of poor soldering at time of manufacture, or by heat from hot components that were not stood off the circuit board far enough. Look for discoloured (toasted) PCB substrate too see if excess heat is involved. Intermittent connection can also result from hairline cracks in the printed circuit, which is often impossible to see.

The simplest and fastest way to approach a problem like yours is to simply re-solder the entire part of the circuit at fault. Identify the power amplifier section of the right hand channel and retouch all of the solder pads. If that fixes it, great! When you do the re-solder the chances are you will see what caused the problem. If the fault was heat related solder breakdown, the other channel is likely not far behind so re-solder that one as well.

If re-soldering doesn't fix it, then the next step is to start prodding around the circuit board with a chopstick (best to use a dim bulb current limiter and crappo speakers when doing this). Flex the circuit board by pressing on it, push against each component standing off the circuit board, and tap the circuit board all over the place and see if you can make the fault come and go, then narrow down to the most sensitive location. Good luck!

Thankyou! That gives me hope ^. Not that others were not helpful (I'm too under the weather for this today ;(

I will report back again when I start feeling better!

Oh btw it has not worsened or became any more frequent. I wonder if its safe to use as is. I mean just until I get it sorted out.
 
Last edited:
I wonder if its safe to use as is. I mean just until I get it sorted out.
I suspect there isn't a definitive answer to that without knowing what is actually causing the fault.

Edit: on re-reading your description, I think you can safely continue to use it. The type of intermittence you describe can be hard to trace until it gets worse, so using it may actually help, because you cycle the unit thermally each time.
 
Last edited:
A basic diagnostic step is to eliminate external effects. I suggest removing all connections except one pair of speakers. If you have some shorting plugs use them to short the inputs. Leave the unit running like this when you are not using it and see is the buzz occurs. If it doesn't the fault may be upstream in the preamp.

Alternatively, swap the left and right inputs (do not swap the outputs). Does the buzz swap sides? If yes the fault is upstream, if not it's in the power amp.
 

wasabii

Member
2021-11-30 10:48 pm
I suspect there isn't a definitive answer to that without knowing what is actually causing the fault.

Edit: on re-reading your description, I think you can safely continue to use it. The type of intermittence your describe can be hard to trace until it gets worse, so using it may actually help, because you cycle the unit thermally each time.

Good to know, and thanks. Because it almost makes my Bryston and Nad Bee integrated boring, I missed it while un-hooked.

I'm not a tech but it doesn't look like theres a whole lot there to weigh 48 lbs. I'll pop the top off mine and have a look around in a couple days, when feeling better.

The output caps in his aren't original as far as I know, they like ML have big blue Philips caps.
 

Attachments

  • 1750084-d7416b9d-proceed-amp-2.jpg
    1750084-d7416b9d-proceed-amp-2.jpg
    154.1 KB · Views: 50

wasabii

Member
2021-11-30 10:48 pm
A basic diagnostic step is to eliminate external effects. I suggest removing all connections except one pair of speakers. If you have some shorting plugs use them to short the inputs. Leave the unit running like this when you are not using it and see is the buzz occurs. If it doesn't the fault may be upstream in the preamp.

Alternatively, swap the left and right inputs (do not swap the outputs). Does the buzz swap sides? If yes the fault is upstream, if not it's in the power amp.
By process of elimination I narrowed it down. I changed every connection, the power wire then finally the speakers, the noise always came back. The last time it occurred I managed to switch off the preamp essentially isolating everything upstream. I'm 100% sure the trouble exists in the right channel of the amp.

I was relieved to hear potential cracked solder joint/s. By reading around these particular amps are pretty reliable work horses. Thanks again. Since I got it it I haven't lifted it off the amp stand. But I had to have it shipped half way across Canada to buy it, perhaps it was bumped along the way. I'll upload some pictures with the top off when I get the time to have a look inside it.
 

wasabii

Member
2021-11-30 10:48 pm
I think you mean 'power supply caps'. Specialist manufacturers do tend to match models in small batches and are victim to component availability. Lead times on bespoke capacitors can be 6 months or more so it is not unusual to see different types in different production runs.
That hadn't dawned on me, true. I seen dozens of ML amp internals I never seen the black ones before.
 

Mooly

Administrator
Paid Member
2007-09-15 8:14 am
This intermittent buzzing comes and goes at random. I may catch it for 5 or 10 seconds once in a month, or as often as every hour or two for the entire length of my listening session.

I had this very issue in a build of the 50 watt blameless class B amp built on the original official boards. Long story short... when the buzzing occurred I could also see some disturbance on the rails on a scope. The cause was noise on the mains and I suspect the source was our electricity supplier experimenting with mains based transmission of data. This was 25 years ago and they were looking at internet via mains amongst other things. I could see the noise by looking at the secondary voltage of a small external transformer where the noise was obvious. Large reservoir caps do not remove this type of noise.

So do you have any mains signalling devices?

Why did it effect one channel?

I my case it was because of compromises in using one PSU for a stereo pair. The noise disappeared when the inputs (left and right) were not effectively joined together at the preamp. In other words when not connected to a preamp the power amp was silent.

I would say don't assume that just because it is a Madrigal that it has grounding that is immune to this.

So in the first instance and to prove this you need to see the noise on a scope and see where it enters the chain.
 

wasabii

Member
2021-11-30 10:48 pm
I had this very issue in a build of the 50 watt blameless class B amp built on the original official boards. Long story short... when the buzzing occurred I could also see some disturbance on the rails on a scope. The cause was noise on the mains and I suspect the source was our electricity supplier experimenting with mains based transmission of data. This was 25 years ago and they were looking at internet via mains amongst other things. I could see the noise by looking at the secondary voltage of a small external transformer where the noise was obvious. Large reservoir caps do not remove this type of noise.

So do you have any mains signalling devices?

Why did it effect one channel?

I my case it was because of compromises in using one PSU for a stereo pair. The noise disappeared when the inputs (left and right) were not effectively joined together at the preamp. In other words when not connected to a preamp the power amp was silent.

I would say don't assume that just because it is a Madrigal that it has grounding that is immune to this.

So in the first instance and to prove this you need to see the noise on a scope and see where it enters the chain.
I have next to no experience fixing electronics, I wish I could be more helpful answering your questions.

But I was under no assumptions regarding ground immunity.


Perhaps it only effects one channel because of its true dual mono design, perhaps the other channel is isolated from being effected?

I powered up the amp this morning, its dead silent, in fact the quietest amplifier I've ever owned. So far its being good to me, there are no noises in the way of humms or buzzes, it sounds incredible actually.. the late NAD 326 Bee I had playing previously doesn't come anywhere close to it in terms of sound quality. I'll take a big hit it if it dies. I'll get in there with some chop sticks this afternoon. To look for cracked solder joints.
 
You don't want to hear from me, but isolating the buzz to one junction will prove productive whether the cause is noise on the P.S. rail, a bad solder joint, or a bad weld inside a component. While scopes are nice, the $20-60 meter listed in post 6 is as effective for blatant problems. You find the earliest place that the sound occurs, then you work in that area only! Newbies are likely to do damage soldering every joint willy nilly. You can overheat things & cook them, or bridge two traces with excess solder. There is a Mark Levinson 431 thread within the last month, download that schematic for a guide. I'd start by shorting the inputs, then looking at the P.S. rails with the meter or scope for excess AC voltage when the buzz occurs. That would eliminate or confirm the problem on the AC supply listed in post 16.
If not, then I'd look at the junction of R126 and the collector of Q103, the base drive to the Q117 the VAS. Pick the input to the VAS of the channel that is buzzing, of course. If the buzz is at input of VAS, the bad junction or component is between the input and that point. If not, afterwards. You keep dividing the circuit in half until you find exactly which stage is causing the buzz. Then you press jointss in that area with a chopstick to see if it makes the buzz come and go. If it does, the place you are pushing is the bad solder joint. That is what your would resolder. From the picture, you'll have to take the board out anyway to do any soldering on the bottom.
 

Mooly

Administrator
Paid Member
2007-09-15 8:14 am
I have next to no experience fixing electronics, I wish I could be more helpful answering your questions.

But I was under no assumptions regarding ground immunity.

From your original description of the fault (post #1) I still suspect that something external to the amp is likely to be the cause but it is very difficult to easily prove that.

It would be worth seeing if the problem occurs at certain times of day or when you have certain household items running.

@johmmath suggested using shorting plugs on the inputs and running the amp in isolation to see if it buzzes. That would be the first thing I would do tbh, you have to try and prove whether the problem with the amp or external to it. It has to be just the amp and speakers and with nothing else external connected to it.