Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Violent buzzing from computer speakers

Status
Not open for further replies.

longtallsally

New Member
Hi there, a total noob here with what should be a fairly simple question and answer, but based on my limited advanced knowledge of circuit design, I was hoping for some help from those more experienced than me.

So it is a set of Monsoon elecrostatic speakers with a powered subwoofer (typical setup really). I have done some general googling and research on the "ground loop" humming that many suffer from and resolve with various electronic bits. However, I'm pretty certain that this is not my problem as no matter what I've done and in multiple locations the problem persists- so I ruled that out as the source.

Being unemployed and needing a break from the job search, I tore into it to see if I could see any obvious issues, which of course I saw none. Here is the setup:

**broken link removed**

**broken link removed**

If I touch in the upper center of the board here, I get a horrible buzz. I figured I was just grounding out something, but thought I'd mention.

**broken link removed**

If I touch the spot mentioned above and the power supply it actually cancels out and the sub goes completely off, but the speakers stay on.

So my question is what component has gone bad and where might I be able to source it? I'm not a noob with a soldering iron or a meter, so if I can get a little guidance, I should be good.

Thanks a bunch in advance as I love these speakers and it is driving me nuts to have to use the stock ones in the iMac.
 
Don't try to run before you're walking...

Is the buzzing still there when you disconnect the amp from all audio inputs and have it running with only the speakers attached?

If not, are you certain that you are not connecting it into a noisy source?

Have you tried connecting a different amp to the same source to test this?

Is the source drive level compatible with your input? If not you may simply be overloading the input stage on the amp causing hard clipping in the output stages etc.

Sticking your finger here and there on the board will not achieve anything other than popping a transistor or chip if you're unlucky, so don't do that any more.

A common problem I have found on small-ish surround amps is a leaky bridge rectifier placing mains hum on the DC lines. This could possibly be your problem.

The Bridge Rectifier, in case you don't already know, is the 4 legged device you see mounted on the Aluminium bracket/ heatsink at the top of pic3 and marked + ~ ~ -.

The only other problems I have seen with them are blown outputs due to shorted cables. This is usually accompanied with a burning smell around the area the outputs are located, sometimes a bubble or hole is present on the devices that are blown.
The output device on your amp is a 20 odd pin chip mounted under the bracket at the back of the finned heatsink. I think this is unlikely in your case, but it could still be faulty and presenting DC to your speakers.

rgds
 
Well said. So then:

No, the buzzing is not there when disconnected from audio inputs. I have tried other amps and different power sources with the same results.

Roger on the no touching of stuff any longer. Didn't think it was good, but it was giving me some ideas on tracing of sources of stuff.

Good call on the levels of all parties involved. This leads to more info I forgot to mention: only one channel (one of the speakers) works when it acts up, and the sub behaves like it is not crossed over. That said, I have lowered levels dramatically for testing purposes so I don't think clipping the amp per se is happening.

I like the leaky Bridge Rectifier and I think that speaks to a degree to the added info I put in from above. Oh yeah, and it used to "fix" the issue if I ran my hand over the heat sync, so you really are on the right lines.

I'll look closely into the above and report back. As such, if it seems bad, how would I go about fixing it?

Thank you very much for the help.

Don't try to run before you're walking...

Is the buzzing still there when you disconnect the amp from all audio inputs and have it running with only the speakers attached?

If not, are you certain that you are not connecting it into a noisy source?

Have you tried connecting a different amp to the same source to test this?

Is the source drive level compatible with your input? If not you may simply be overloading the input stage on the amp causing hard clipping in the output stages etc.

Sticking your finger here and there on the board will not achieve anything other than popping a transistor or chip if you're unlucky, so don't do that any more.

A common problem I have found on small-ish surround amps is a leaky bridge rectifier placing mains hum on the DC lines. This could possibly be your problem.

The Bridge Rectifier, in case you don't already know, is the 4 legged device you see mounted on the Aluminium bracket/ heatsink at the top of pic3 and marked + ~ ~ -.

The only other problems I have seen with them are blown outputs due to shorted cables. This is usually accompanied with a burning smell around the area the outputs are located, sometimes a bubble or hole is present on the devices that are blown.
The output device on your amp is a 20 odd pin chip mounted under the bracket at the back of the finned heatsink. I think this is unlikely in your case, but it could still be faulty and presenting DC to your speakers.

rgds
 
Last edited:
Thanks to tunedwolf, I found the following statement which looks like it very much applies:

"The correct diagnostic instrument is an ohmmeter. You measure the collector-to-emitter resistance of the output transistors. If they're shorted, they will measure zero ohms. In most instances, you need to check the driver transistors, the bias transistor and/or diode(s), and the bias potentiometer. Drivers often get leaky -- or short themselves -- especially if the output transistors are shorted to the base as well as collector-emitter. Bias transistors and diodes like to open, and bias pots get touchy.

For a reliable repair, I will usually replace output transistors, driver transistors, bias elements, emitter resistors and perhaps a few small resistors in the driver circuits."

So I think we are on the right track b/c if I wiggle the volume control plug, all heck breaks loose. Now I just need to know what to measure the resistance across.
 
Investigate what's going on with the "all hell breaks loose" when you wiggle the volume control plug.
You are correct in that a multimeter would be used to test P-N junctions in transistors etc, however it would normally be set to diode test mode. Ohm's mode doesn't normally have sufficient test voltage to bring the junction into conduction.

As there are really only two active devices on the board that I can see, and assuming there is not a dry or broken solder joint somewhere, or a knackered socket etc, then I think I would be checking the bridge rectifier to see if it's leaky and then replacing the amp chip, or both if it is!

rgds
 
Again, excellent advice. However, seeing as you have very accurately narrowed down the options, I still have a couple noob questions:

- how do itest to see if the bridge rectifier is bad?
- is there a procedure for testing the amp chip?
- if one or both and the jack test as bad, where do I source said parts? Radio Shack? Hobby store?

Thanks again.
 
A Bridge rectifier is nothing more than 4 diodes inside one package connected to form a full wave bridge arrangement.

Each leg of the bridge is a connection point of two diodes, where marked a.c. (~) the connection is one anode and one cathode of a diode pair, where marked +, the connection is between both diode pairs' cathodes and where marked -, the connection is between both diode pairs' anodes.

Normally a single diode would be marked at the cathode by a coloured ring round it, however in a bridge, you can't see the individual diodes as they are all inside the one package. The symbol for a diode looks something like this, where A=Anode & C=Cathode:

A C
---|>|---

If you imagine for a moment, that is actually an arrow pointing in the direction of current flow, positive to negative. Diodes allow current to flow in one direction only.

This is how a bridge arrangement looks:

A C ~ A C
|---|>|---o---|>|---|
| |
o - o+
| A C ~ A C |
|---|>|---o---|>|---|


To test a single diode, you would set your multimeter to the diode test mode, then place the red lead on the anode and the black one on the cathode. You should see the voltage drop in mV on the meter scale. Swapping the test leads around should result in no flow or zero mV. If there is a reading when the leads are swapped, the junction is suspect.
Remember, there are 4 diodes in the bridge, so you simply work round each one in turn.
Also the surrounding circuitry will affect the readings you get, so remove it from the circuit and then test it. Lastly, most meters are fairly sensitive, so don't put your fingers on the test connections, hold only the plastic handles otherwise you may end up reading your fingers!

Diodes come in a multitude of flavours, each with their own characteristics and specifications, so there is no absolute guarantee that the junction is bad by performing the test, however, most silicon rectifier diodes will generally follow the rules.
If you are unsure about the specs for the bridge, google the data sheet for the type number and read up on them.

Testing the bridge rectifier should be pretty straight forward, the amp chip however would require a complex series of tests to confirm it faulty or not. For all the cost of the chip, I would just assume it's knackered and replace it anyway. If the bridge is leaky, the chip will have been stressed well beyond it's spec and may well fail later anyway.

I'm in the UK, so I'm not 100% sure where you will obtain replacements, but Digikey and Mouser are two possibilities that come to mind in the U.S. You could always try the local TV repair shop and ask them if they could order them in for you.
Other folks here might be able to shed some light on a source of parts if you post the type number of the devices.

Sorry about the rubbish ascii art, but you get the idea.
There is a half way decent explanation of stuff (with much better artwork!) here if you need further help:



rgds
 
Forgive me but I am just now getting to do tests on this thing. So I removed what I'm pretty sure is the bridge rectifier and performed the tests. Here are a couple pics of the results (can you tell I'm a visual learner?). BTW, the descriptions you gave were wonderful and made it very easy. I tried the test on the board and it failed miserably, but thanks to the link you provided, it was stated that when on the board, the test may not be accurate. So I removed it and retested. From what it looks like, the bridge rectifier is performing as it should- despite being backward in terms of anode and cathode.

**broken link removed**

**broken link removed**

**broken link removed**

So now I am kinda back to square one as it sounded like the amp chip could be the culprit. So which device is that? Forgive my ignorance again, but I'm trying to get this...

And again, a huge thanks for the guidance!
 
Last edited:
so the more research that I do on the topic and the more I think about it, could this be something as simple as the crossover(s) being bad?

Here is my logic: when I was in my undergrad, I bought one of those horrible and cheap in line subwoofers the you didn't need to have an amp for and you hooked your speakers into it and then hooked it into your receiver. I didn't have any money and when it went south (basically did pretty much what this setup is doing) I decided to play around a bit. I figured out the wiring and ended up getting a couple capacitors and putting them in and voila! It not only fixed the sub, but it sounded better than ever. Here is a link to the concept I'm describing:

Passive Crossovers

If you look at the third picture I initially posted, I think the capacitors are just underneath (as the picture lies) the bridge rectifier.

Am I on drugs?
 
From what it looks like, the bridge rectifier is performing as it should- despite being backward in terms of anode and cathode.
The bridge rectifier has 4 diodes in it. See the diagram on the wiki page. The + and - symbols are the same as marked on yours. You need to check all 4 diodes in the bridge, forwards and backwards.
longtallsally said:
No, the buzzing is not there when disconnected from audio inputs. I have tried other amps and different power sources with the same results.
Sounds like the source of the buzzing is the audio source. Try a different source of audio for the amp instead. A battery powered item would be best.
 
If the defective component/part is thermally sensitive, you can sometimes use a spray can ... with a small tube outlet ... to temporarily chill specific parts .... and thereby determine which is causing the problem.
Usually, the spray chiller is available at electronic supply stores.
Takes some care to use ... Doesn't always work, but it might.
 
So I think we are on the right track b/c if I wiggle the volume control plug, all heck breaks loose. Now I just need to know what to measure the resistance across.
Resistance measurements are pretty useless for this kind of troubleshooting.

have you checked all the grounds along the "bad" path especially at the input? You say it doesn't hum when there is no input line attached.

If you wiggle any "plug" and hear noise, you need to clean that plugs contacts and pins.

I am guessing: if it is a "volume control plug", does it connect the volume control potentiometer to the PC board? If so, a bad connection will open circuit the gain stage input and it could hum like crazy when that happens.
 
Last edited:
Of course the amplifier will produce a buzz sound if the input cables are not shielded audio cables because the wire picks up mains hum like an antenna.

The input cables are actually two wires each. A signal wire that is shielded by a shield wire. If you use only a single signal wire then it is an antenna that picks up mains hum.
 
The bridge rectifier has 4 diodes in it. See the diagram on the wiki page. The + and - symbols are the same as marked on yours. You need to check all 4 diodes in the bridge, forwards and backwards.

Yep I did that. Sorry the pics were just to show an example and if I was doing it wrong. I did it across all the terminals

Sounds like the source of the buzzing is the audio source. Try a different source of audio for the amp instead. A battery powered item would be best.

Yep tried this as well- hooked up an iPod not hooked to power or anything- completely standalone. Now I did this a while back and not recently...

So I decided to put the bridge rectifier back in and try this again and see what I could see and perhaps give more detail.

This was bad.

When I flipped the power switch I got a pop. Now I got nothing. So now I have bigger fish to fry. Here is the hinkyness, when I tested for continuity (just to see if the connections were still there) everything is still connected and the fuse is good and I can feel the coil from the amp getting warm, so I know I have juice, but now I have no sound at all. I obviously shut it down to keep from doing any further damage, but really am not sure of what I messed up.

Thanks a bunch for all the responses and guidance. I'm really trying to learn here and don't espouse to know too much (I can't stand people that come on to forums and think they know what they are talking about and don't) so I appreciate the patience.
 
A pop then nothing might suggest that you have accidentally fitted the bridge back in the wrong way round, so double check that the terminals line up with the board markings, or if no silk screen is present, follow the + terminal on the bridge to the smoothing caps and make sure that it goes to the positive side. Also a solder splash in the wrong place could cause this too, so double check your solder work.
The other possibility is that the amp chip has finally died, this is the multipin device under the bracket at the rear of the heatsink in your photo.
I'm a little surprised that the fuse hasn't gone, but it's not unknown.
As the bridge appeared to be ok, and assuming that you have not damaged it by removing and refitting it, then I would replace the amp chip.

rgds
 
OK, this sounds good. I'll check the bridge rectifier again.

On the amp chip, am I correct in saying that it is under the bracket holding the bridge rectifier, and then there are 2 capacitors under that and then it is right under there? I'd post another pic, but as usual, my smugmug account is down.

So where might I source one of these? And is it as much of a bear as I think with over 20 connections? How exactly do I remove something like that? My soldering skills aren't THAT good.
 
Yup, 20 pins sounds about right. It's the rectangular package, clamped to the heatsink (the bit with the fins at the rear) by that bracket.

Desoldering it shouldn't present any problems with either a desolder pump or solder wick.
If you're not so handy with an iron, use solder wick, sometimes called a solder mop, it's less likely to cause you trouble.

Basically you work on each pin in turn. Place a clean bit of wick on each pin/ board pad and touch the iron tip gently on top of the wick. The wick mops up the solder using the impregnated flux in it. Be gentle with the iron pressure, and do not hold the iron too long on the pad or you may distrurb the pad bond to the the board by overheating it.
If it's not going your way, add some fresh solder to the pin and start over.

Always cut off the used wick, the excess wick takes longer to heat with the iron, can scorch the board and can cause you other trouble too like shorting caps or accidentally desoldering an unintended surface mounted component without you realising that you had even done it, or indeed where on the board it might have come from. Inevitably when that happens you cut it off and toss it with the used wick without even knowing it happened, well not till it's too late anyway.

After desolding a couple of pins you will get a feel for how long the iron needs to be there to mop up all the solder. Remember to lift both iron and wick off the junction at the same time otherwise you will merely solder the wick to the junction.

If you are completely committed to replacing it, the easiest way to remove it is with a sharp pair of precision snips. Just cut each pin at the package, then when all pins are cut, just hold the pin with small pliers or a croc clip etc, and touch the solder side with the iron to let you gently pull it through the board and out.
Use the wick to clean up the pads after you remove all the pins from the board before fitting the replacement.
This method is especially good with double sided boards as there is less risk to the plated holes on the board due to a joint not being completely desoldered before you try to pull the component out.

Remember to double check that you have not accidentally splashed solder anywhere it's not supposed to be every time you complete a task, if you have, clean it up with the wick before moving on.

As for where you will get the chip, well I can't really say. I don't know what the part/ type number is. It might be available from mouser or digikey or somewhere like that. Also your local t.v. & radio repair shop will usually be happy to order one in for you if they don't have it in stock, just be friendly with them. I would sincerely doubt rat shack will have it, but it's worth a phone call I suppose. There will most likely be other places locally in your neck of the woods, perhaps someone in the US will be able to point you in the right direction after you can supply a part/ type number.

rgds
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top