My studio monitor is sad. Is this a good place for amp diagnostic advice?

I have a pair of M-Audio M3-8 3-way powered speakers. Product Page Certainly not high end, but I really enjoy them. -One of those rare situations when a product outperforms it's price point. According to the mumbo-jumbo, it's class A/B amps in a tri-amp setup.

One of them has decided to start making a @1550Hz tone from it's mid driver at all times, regardless of what is or isn't connected to it. The tone is immediately present and steady when powered on, and sweeps down quickly when powered off. Volume pot on the speaker does not effect volume of the tone. Input signal does not effect tone. Transistor? Capacitor?

My first instinct was to find a used one with a good amp (or new old stock) and just swap out the amp plate, but no one seems to be selling, and a tech wants 2/3 the original cost of the speaker before they even look at it. (not to imply that they don't deserve it, but I'd prefer not to spend more than the purchase price to repair it.)

I'm an Audio Engineer (which doesn't mean much lol) I know how to solder, use a multimeter, and read a schematic. Is this something I can potentially figure out and fix on my own with some help, or should I just begin the grieving process? :cry:

Thanks for reading.
I would bet on the speakers having switched-mode supplies for the internal amplifiers, and one or more of the capacitors has cooked. If so it would probably be visible, either swollen or even burst. Replacing the electrolytic caps in the PSU with higher temperature ones (same value and voltage) should give you speakers that will last longer than they did before. Broken caps can make the PSU unstable, resulting in the squeal that you can hear.
if this is the cause, order parts for the other speaker too, as it will fail in the same kind of way soon enough, and you will have parts on-hand for it.
  • Like
Reactions: 1 user
SMPS are usually run above audio frequency, or everything would squeal like a stuck pig. And the required size of the magnetics is smaller above 20k. Running them at 1.55k would be highly unusual.

And a bad SMPS would affect woofer, mid, and tweeter. Maybe blow the tweeter.
  • Like
Reactions: 1 users
Find the schematic if you can. Then probe with what ever you have at different stages of the crossover and amplifier looking for where the oscillation originates. You can use your DMM on AC setting and probe the signal path at different stages starting at the crossover. With the input shorted you should see close to zero AC signal all the way through a working amplifier. Since you have two of these you can compare the working one with the broken one at each stage. For an oscillation to happen there is some phase shift that is new to the circuit, so a failed capacitor is the first thing to look for. An oscilloscope is really handy in this case. If you don't have one, there is free Oscope software for PC's that use a sound card for input. Since your problem is in the audio range, that could work. You need to make a set of probes with over voltage protection and maybe a resistor divider attenuator to protect the input of your sound card. There are also some pretty cheap USB scopes out there.
  • Like
Reactions: 1 user
SMPS are usually run above audio frequency, or everything would squeal like a stuck pig. And the required size of the magnetics is smaller above 20k. Running them at 1.55k would be highly unusual.

And a bad SMPS would affect woofer, mid, and tweeter. Maybe blow the tweeter.
I agree, but I’m talking about loop stability. The SMPS capacitors have a hard time, and are usually the first parts to fail in consumer electronics. The cross-over network would filter the 1.5KHz from the tweeter and base speakers.
As the OP doesn’t have a ‘scope I would look for physical damage, and SMPS caps often show that when they fail.
First of all - Wow! You guys are so generous to be so helpful and share your knowledge like this. Thank you!

No luck finding a schematic so far, but upon visual inspection:

This guy looks suspicious, right?


And it just happens to be right next to the Mid Out?



Could it actually be as 'simple' as replacing that one cap?
If so, do I need to be worried about anything still being energized, and if so, what do I need to do?
Also, other than a resistor to dim an led light that's too bright, I don't know anything about sourcing components like this. Any help in selecting the replacement (upgrade?) would be great. :) Thanks again!
Looks like that cap was bumped by a soldering iron during assembly. Not sure that's your problem, but it could certainly be easy to remove and test it. Test the value. A replacement is easy to find. Measure the diameter and height and look at Get the same value and voltage. Splurge and get a low ESR style if you like.
Test the value. A replacement is easy to find. .
Value at 2 v test voltage does not predict value at 50 v. Replace it. Mark - of caps on board with sharpie before removing.
Test main AC in caps at below 2 v to analog ground before touching any metal. The big ones rated 200 or 400 VDC. For complete safety test rail caps (60 to 120 vdc rating) at <2 vdc also. If not, discharge with insulated lead connected to 100 to 2000 ohm 5 watt resistor connected to analog ground.
Wear safety glasses soldering, especially desodering. Solder is a liquid, can splash in your eyes. Wear safety glasses testing while out of case too, mistake can cause parts to explode.
Low ESR caps are for the switcher supply oscillator areas, not for the analog out areas.
Post 9 looks appropriate, also.
Last edited:
The amplifier in the M3-6 and M3-8 monitors is made in class A/B. For heat dissipation, large radiator fins are provided. The midrange driver and tweeter are housed in an isolated chamber. The amplifier board is located in the same housing as the woofer.

For the low-frequency speaker, an amplifier assembled on the basis of two powerful B31N20D SMPS MOSFET transistors is used. Two TDA 2052 chips are used to amplify the signal to the tweeters and midrange speakers.
Last edited:

There is solder, but there's also some heat discoloration there that I'm not seeing anywhere else. Could that be something, or from manufacturing?
-resistor has glue on it, not a crack.

He really did fire the parts cannon, didn't he? Lol. I might be up for that.


C66 has a bulged top there, but all other caps 'appear' to be okay.
  • Like
Reactions: 1 user
That’s exactly what I was expecting. Next to that coil, it looks like it is part of the switched-mode power supply. I would replace all 4 of those capacitors, and possibly the 2 lower down in the picture. Use 105-degree parts and good quality low ESR (Nichicon, Panasonic…). While you are there there is no harm in replacing the physically damaged one too, and remember to get the parts for the other speaker at the same time. That might be enough to get free delivery from Mouser or Digi-Key.
Take pictures with your phone before you remove anything so that you can be sure to put the right parts in the right places with the right orientation.
  • Like
Reactions: 1 user
I'm having 'like' envy. I'm guessing that I'm on some sort of forum probation and am not allowed to 'like' posts?

Thanks again for all of the help and attention here. It's amazing to have a resource like this.
I am ordering caps today. I'm tempering my expectations, but it would be so cool if this does the trick.

Does anyone feel like there's a need to replace the big 4700uf caps while I'm at it?
The high-voltage caps are relatively unstressed and unlikely to fail. The output ones work hard, and those are the ones in the picture with obvious faults.
if this is the first repair the OP has tried, I would suggest initially changing as little as possible to return the amp to correct operation, and possibly follow-up with further work if needed and confidence is gained. Changing too much at once and breaking things accidentally would be a tough one to recover from. The original caps have come this far (most of them…) so no reason not to let them work a little longer, having gained knowledge on what to look at if further problems arise.