Bose Acoustimass 3 humming

We have an Bose Lifestyle 5 & Acoustimass 3 system. I have to say it sounds pretty good in the room, especially for classical music. Great room acoustics and just one cube per channel probably help. Lower bass and some midrange are a bit light.

The Bose has low hours from new. Over the past few years, it rarely has been used because it has a few issues

PROBLEMS
1- Distortion sometimes which overwhelmed any music, particularly on start-up. Some distortion sometimes subsided after listening for a while.

2. No sound from bass module sometimes.

3- Scratchy bass and treble pots.

4- Very dim blue display (although got a bit brighter after a few minutes of use).

Tonight, following some videos and advice of @kevinkr I looked at the Acoustimass 3 module. This seemed easier to open and work on than newer models. There is no service manual and very little information in forums so diagnosis options are quite limited.

BASIC DISASSEMBLY ACOUSTIMASS 3
- remove 3 control knobs
- remove small plastic cover plate and one screw
- pop off big plastic end cover
- remove 3 screws from top board
- remove 6 screws from bottom board / heatsink
- remove 2 wire connectors
- gently tilt top board, heatsink, bottom board; don't rip out the soldiered cables. Don't break the stand-off pins. Caution with heatsinked parts

--> I did not remove the 2 boards from the unit completely as that would likely require desoldiering connectors or removing the transformer underneath.

INSPECTION
- Overall, parts quality and board assembly was excellent. Nichicon caps littered everywhere. The pots were labeled Japan and looked like Alps blue but I didn't look carefully. Board layout and soldier work was very tidy.

- There were some spider webs and dust inside which I quickly cleaned with vaccuum, dust brush, canned air, alcohol.

- Caps looked perfect. Inspected soldier joints with a magnifying glass and frankly everything looked new.

SUSPECTED CAUSES
- There was a lot of spider web stuff bridging a few contacts on the larger wire connector.

- I suppose the power supply caps are always suspect.

- With no factory manual and very little advice on forums, diagnosis work is very limited.

--> I decided to tackle this one step at a time and see the results of the dust/spider web removal. Well, I couldn't resist cleaning the pots also, lol.

REASSEMBLY
- Cleaned and replaced thermal grease
- Followed advice below and elsewhere from @kevinkr
- Don't overtighten the 6 black screws. They can easily strip in the MDF. Overly tight screws will prevent the 1 screw for the white cover from aligning properly.

RESULTS
- 100% distortion for a few minutes, no music heard.

- I power cycled a few times. Distortion declined slowly.

- No distortion after 10 minutes. Sounds perfect.

NEXT STEPS
- I will keep unit unplugged overnight and try this for a few days.

- I will fix the display this week. I think that is just 3 caps to replace in the Lifestyle unit.
To get to the amplifier module there are 8 screws (10? I forget) that have to be removed to remove the bracket the amplifier sits under. Do not remove the plastic plate from the bass module as sealing is crucial and the original gasket should not be reused. When reinstalling tighten the screws by hand only and do not over tighten as it is easy to strip out the flange area as it is only mdf.. If you changed the power transformer you have probably already figured this out.

Watch for twisting in the woofer cable where it comes out of the box , avoid pinching. (Shorts possible - will destroy bass amplifier)

You can pop apart the amplifier module with a large flat blade screw driver if you need to - stick it in the slot and pry gently, so that you are pushing against the plate - do not twist. When you reassemble make sure the heat sink spring clamp is aligned properly with the holes in the board before you put the cover back on. Put end of cover without the aforementioned notch into the notch in the heat sink and press until the notched side locks into the notch on the other side of the heat sink. Be very careful when handling the heatsink not to scratch the area where the devices sit as this is a non-conductive anodyze. You should apply new grease to the devices before reinstalling the board.
 
Well, the fun didn't last too long.


This morning, we plugged in the system, a relay clicked and we heard distorted sound for 2 seconds then nothing.



Leaving the Lifestyle head unit plugged in, then...


- unplugging the Acoustimass to mains resulted either in a thump or relay click.



- replugging the Acoustimass resulted in a relay click and sometimes 0.5 seconds of distorted sound.


Also tried keeping the unit plugged in for 20 minutes with no improvement.
 
I have a Bose Acoustimass Multimedia Speaker System (bass module with 2 satellites), which seems to be a very close cousin of the Acoustimass 3. I've had issues similar to yours at various times, which I usually could at least partially fix, but never permanently fix.

If you have not seen a schematic, there is one over at hifiengine. Look under Bose, then under "Misc Schematics", there is a file for AM3P.

As you know, the different parts of the electronics in the bass module are distributed over two boards stacked on top of each other, and since you can’t access anything when they are assembled and you have to disconnect them to get in there, it is difficult to test and trouble shoot this unit. Seems it is "optimized" for a workbench only at a Bose facility.

Over the years, usually the problem I've had been loud static/distortion, sometimes followed by complete cutting out of the sound. One time it was clearly related to a dirty volume pot. Other times I have found broken solder joints on the connectors for the short cables between the boards (undoubtedly my fault from removing and reinstalling them many times), and also on several of the transistors that press against the heat sink.

Even after these fixes, there was still static on startup, but this became less, seemingly supplanted by a tamer buzz, usually more or solely on one channel. This would appear on startup, but it would fade over a period of time, usually 10 minutes or so, but the length of time got shorter with more consistent use, but never less than a couple of minutes. I've never figured out what it was; perhaps a cold solder joint that did better once warmed up, or some other failing component that liked to be warmed up.

I used this unit back in the mid-90s until about 2010 as a poor-man's home theater (back when TVs had analog out sound), and it served well for that period. I keep it around now mostly as a curiosity. Its sound once the buzz fades is only decent; I have so much better stuff laying around.

Good luck with your attempts to get it working. I find that consistency in behavior is hard to come by now.
 
@northpaw -thank you very much for your tips and advice. Very helpful. I will give this another shot. It is very cold to be running fresh air for soldier rework!



I appreciate Bose is not #1 here. But I am helping older family members get their holiday music running again as a favor. Thanks.
 
Even if the caps are branded Nichicon or some other respected brand, it doesn't mean they are of the same high manufacturing standard and grade as their hi-fi cousins. Their longevity probably depends more on the plant they were made in than the brand, grade and colour of the sleeve.

The faults I find with cheap audio systems generally, relate to components that are minimum spec. for the job, meaning that they fail more often and sooner than you would expect from older, more expensive, high quality products. Electrolytic caps that last only a few years instead of 20 or more are symptomatic of companies that just seek a bigger market share today rather than maintain yesterday's engineering standards and reliability since that does nothing to improve this year's bottom line.

You could follow that approach and just trash the system, buy a new replacement for the little it might cost now. If you want to complete the repair and servicing process though, I'd systematically re-cap it, at least the electrolytics and also increase the capacitance value and voltage rating where it's (likely) marginal, particularly for the main power supply smoothing cap(s). These are are the most common culprits for buzzes and hums that die away as the amp warms up.

I have a few of my own cheap amps in PC audio system clutter around the desk-top. Two of them take up to 10 mins or so from switch-on to stop buzzing or humming as the caps slowly recover and the audio begins to emerge from the crud. This is just typical of old equipment and caps of any original quality, since none of them will remain good for a lifetime and I'm just too lazy to repair stuff that I don't really need right now.

I'm not familiar enough with Bose electronics products to make specific comments but keep an eye out for small electrolytics used as coupling caps in preamps, tone controls etc. There can be similar signal cutout problems and humming etc. if they dry out, as all electrolytics eventually do.
 
Thanks, Ian, for your comments here and for sharing your knowledge and experience concerning the case of a start up buzz fading over a period of minutes.

I had checked the main PS smoothing caps (10,000µF/35V on +/-20V rails) on my unit with an inexpensive LCR meter, after lifting one leg, and they checked out OK, although I understand that this is hardly a definitive test. But there are indeed several 10µF electrolytic coupling caps in each channel in the preamp section of the AM3. If I open it up again, I'll change those out. When I recapped a late 70's Advent 300 receiver last year, I was struck how virtually every one of the small electrolytic coupling caps in it were way out of spec (although the Advent had not yet developed an issue of a start up buzz that faded with warm up).

Thanks again.
 
Thank you so much @Ian Finch and @sgrossklass for your tips on potential areas for further investigation.


Thank you again @northpaw for detailing your test of the main power supply caps.



I'll take some photos today and compare to the schematics. There is a big electronics supply store down the street so I can just pick up a bunch of caps and replace.
 
I'm looking at the schematic and the amplifiers look like they're straight out of a tutorial. They have the topology that the tutorial would say "DO NOT USE FOR HI FI." They're hard class B.

Am I missing something?
I don't know. What I have (Acoustimass Multimedia speaker system) was marketed as a computer speaker system in the mid 90's (before computer HiFi was a thing), where the bass module could be hidden under the desk and the two satellites, each about the size of a softball, could be placed out of the way. So it wasn't intended to be HiFi, despite perceptions of how the later Acoustimass systems were marketed (which was a powered speaker system providing room-filling music with almost nothing in view), which appealed to those users who valued compactness and hiddeness over fidelity. I used it for TV sound, where it did way better than in-set speakers. Things have evolved quite a bit since then.
 
I just feel like I'm missing something when I look at the schematics. If that amplifier actually works like I think it works, then it's crapola. It's a circuit I might use for motor servo drive... but not for audio.

If I'm wrong, then I want to know how those output transistors are actually biased. It would be a new trick for me.

I built the old op amp driving transistors from the power supply leads like 30 years ago. It worked better than I thought but I would never actually use it. Maybe it's OK for a super cheapo subwoofer amplifier.
 
Indeed. It would seem to be acceptable enough for the BASS OUT (thanks to still-high opamp loop gain and low required slew rate down there) but otherwise it's crossover distortion galore. They didn't even try to get close with e.g. one silicon and one Schottky diode or so between bases to reduce the necessary voltage jumps at least, no, the output stage is as Class B (or rather, Class C) as it gets. Awful awful awful. You never ever want any hard discontinuities in your output stage, unless you're getting into RF modulating your way around them (Class D, tape bias). The resulting distortion is really hard to get rid of and most severe when signal levels are around the magnitude of the discontinuity involved. This is why you see people going on about the importance of the "first watt".

BTW, in terms of culprits for the hum I'd be looking at 560 µF caps C20 and C24, as well as +/-8 V regulators U1 and U2 if need be.
 
Last edited:
I built Walt Jung's 12 volt amplifier, 3 watts into 4 ohms. Dead simple: opamp, two transistors, a handful of resistors was about all it was. It worked and did not self destruct. But distortion rose sharply above 8 kHz or so.

Notice the savings in parts that this topology offers. No bulky emitter resistors, etc. And no thermal runaway either.

If you look at distortion curves for op amps and power amplifiers, the curve is always high at lowest power (voltage), and then drops as power (voltage) increases. Then it rises as max power (voltage) is approached. The low power distortion is largely crossover distortion, I suspect.
 
Thanks to you both for this discussion on the design of the amp in this Bose unit. I’ve a very rudimentary knowledge of amp design, but I do understand how elevated bias in a Class A amp eliminates crossover distortion, how Class AB restricts it to when higher power is demanded, and how Class B does neither. What I don’t know, since I don’t measure distortion except with my ears, is what I would hear from an amp that corresponds to crossover distortion, beyond the obvious case of gross distortion when driven past its limits, or some malfunction like that described in the OP.

I do know that I prefer listening to systems that are powered by an amp with Pass’s Stasis design (which I feel fortunate enough to experience in a Nakamichi PA-5 amp and in Nak SR and TA receivers). I find that these produce a sound that I can only describe as being clearer and sweeter (especially while in the Class A region) than the sound I hear from other good quality equipment of similar cost. I am unable to articulate my experience any better than that. I’ve suspected it is related to the careful (non-global) use of feedback in the design and reduced IM distortion, but I don’t really know.

Anyway, for s&g’s, being retired and not having anything especially important to do this afternoon, I hauled the Bose unit up from the basement where it has been since the last time I played with it in the early Spring. I desoldered the 560µF/16V caps and tested them. They were slightly high, but less than 10% high, and they are ±20% caps. Vloss was ~1%, and 1-2Ω ESR. I was going to replace them with some 1000µF/50V caps I have on hand, just to see, but those all exceeded the 17mm max height clearance off the board for that position, so I reinstalled the originals. Searched around for a more bad solders; found a few possible partial crescents and reflowed them, and reassembled.

Had static/buzz at start up, on the bass module and left satellite only. Based on this, seems hard to pin it on the 560µF caps or the voltage regulators, as both the +8V and -8V are used on the left and right channel opamps (in fact, each opamp is a dual, with one half used for left and the other half for right).

As has been the case before, the buzz fades off by the 2 minute mark, leaving buzz-free music.

I played some Hadyn piano sonatas from an Alfred Brendel CD that I greatly enjoy and am very used to listening to on, for me, a very good system (Adcom GDA-700 DAC to McCormack TLC-1 Deluxe preamp, to a Nakamichi PA-5 driving ADS L730s with renewed tweeters and midranges, supplemented with a NHT SA-2 sub amp driving a NHT Sub One i). The presentation on the Bose was overall pleasant, as I have always remembered it to be, but of course pales to the main system, if for no other reason than two tiny satellites and a bass module can’t reproduce sound in a room the way the ADS L730s and a real subwoofer can. And I can’t use the Bose amp to drive a different set of speakers. That fact makes it difficult for me to separate out what deficiencies in the Class B amp in the Bose might have been contributing to what I was hearing, other than it was not in that noticeably “clear and sweet” league that I mentioned above. Listening carefully, the most obvious feature I felt I could detect that was not speaker related was a slight warble or wavering in some of the longer piano notes. Perhaps there are other features that I simply am not discerning enough to articulate.

But my overall assessment remains the same as it was before: the Bose is pleasant enough for casual listening and background music, and great for an old TV.

To Michael, I hope you get that unit sorted and your family members can get back to using it.
 
Two problems now lol.
1. LOTS OF DISTORTION DURING FIRST FEW MINUTES
On Tuesday, loud distortion disappeared after a few minutes and system sounded perfect. I turned on the system a few times and it still worked fine. Since then...

2. NO SOUND ANY MORE
Can't get sound for more than 0.5 seconds. Sometimes music will sound for less than a second. Sometimes I can hear a relay click once.

I suppose I need to address problem #2 first. Some areas to sniff out:

- "short circuit protect" circuit

- "amplifier dc offset latch-off protect" circuit

- "on off control" circuit J7 (1/8" cable coming from the Lifestyle head unit) has wonky jacks that could be broken from pushing the Lifestyle & Acoustimass back into the wall. I tried changing the 1/8" cable and wiggling it but that did nothing.

- Maybe related to #1
I'm not familiar enough with Bose electronics products to make specific comments but keep an eye out for small electrolytics used as coupling caps in preamps, tone controls etc. There can be similar signal cutout problems and humming etc. if they dry out, as all electrolytics eventually do.
I picked up a cap tester this afternoon and will begin to test (and replace) the smaller caps in the volume, treble, and bass arenas. These three pots worked fine on Tuesday after cleaning but they were wonky for a decade.

Downstream I see compress, compressor detect, level detect, dynamic bass eq, audio detect.
BTW, in terms of culprits for the hum I'd be looking at 560 µF caps C20 and C24, as well as +/-8 V regulators U1 and U2 if need be.
I'll pull these for a quick test tomorrow - thank you very much for the tips.
I hauled the Bose unit up from the basement where it has been since the last time I played with it in the early Spring. I desoldered the 560µF/16V caps and tested them. They were slightly high, but less than 10% high, and they are ±20% caps. Vloss was ~1%, and 1-2Ω ESR. I was going to replace them with some 1000µF/50V caps I have on hand, just to see, but those all exceeded the 17mm max height clearance off the board for that position, so I reinstalled the originals. Searched around for a more bad solders; found a few possible partial crescents and reflowed them, and reassembled.

Had static/buzz at start up, on the bass module and left satellite only. Based on this, seems hard to pin it on the 560µF caps or the voltage regulators, as both the +8V and -8V are used on the left and right channel opamps (in fact, each opamp is a dual, with one half used for left and the other half for right).

As has been the case before, the buzz fades off by the 2 minute mark, leaving buzz-free music.
This is super helpful for diagnosis work.
To Michael, I hope you get that unit sorted and your family members can get back to using it.
Thank you - your narrative and information was very inspiring.
 
A correction to my post #18: The buzz is on both the left and right satellites, and not on the bass module. When I checked yesterday, I had the right satellite sitting too close to the bass module, and I mis-identified where the sound was coming from. This morning, I held the satellites against my ears. So yes, the 560µF caps and the ±8V regulators are not eliminated as the source of the buzz.

I also was able to verify that the buzz varies with volume setting, so the problem is in the preamp.

Sorry about not getting this right yesterday.

Michael: I have no experience with a Lifestyle module. I don't know if it is possible for you to test them separately, but obviously it would be useful to be able to do so.