Emmy’s Speakers Hum
08:40 PM: [Emmy] brought pair of Harman / Kardon powered computer speakers that hum excessively, with a pretty pure-sounding 120 hertz tone. She said that one afternoon a couple weeks ago, they made this loud crackling and snapping noise, and from that point on, they hummed loudly. Investigating… Details of the investigation supplied in the attached comments below.
November 22nd, 2009 at 4:00 pm
07:00 PM: My initial observations, before taking the speakers apart are as follows:
But they’re not. So it isn’t.
* a 120 Hz. tone indicates that the power diodes are, at least in part, rectifying the AC power as they should.
* The speakers do pass signal. That is, you can touch the input leads and hear buzzing and static in both channels. This suggests that the amplifiers are probably okay.
* The power adaptor has an AC output voltage, which means that I’ll not be able to repair these speakers simply by replacing the adaptor, which would have been nice. The rectifiers and filter capacitors are located inside the speakers. Had they been located in the adaptor itself, as they often are, this job would have been easy.
07:20 PM: The electronics are located in the right speaker. So I took this one apart after a bit of a struggle; two of the screws were hidden under the cloth speaker grill and served a dual purpose of not only holding the cabinet together, but also holding the speaker motor itself in place. But I got it apart without breaking anything anyway.
07:30 PM: I applied power to the speakers for a few minutes, and then felt around the circuit board for any unusual heating, and smelled for any unusual odors. Nothing found. So this humming is probably not the result of any excessive current drain in the amplifiers. So I’m suspecting that it’s the filter capacitor.
07:35 PM: I located the capacitor. It is a 3300 UF, 25 volt electrolytic, PCB-mounted unit. I then turned off the power, discharged this capacitor, and read across it with my Radio Shack volt-ohm meter. I saw no charging action with either polarity. Now, I’m thinking that this capacitor has opened up.
07:40 PM: I revved up my 35-watt soldering iron, and once hot, found my solder-sucking squeeze-ball, and unsoldered the capacitor. No problems here. It came out easily, and sure enough, out of the circuit, the capacitor read open. It’s definitely defective.
07:45 PM: Then, I checked my stock for a replacement. I found a few with axial leads (one lead on each end as opposed to both leads being on the same end. But these are quite a bit bigger than the original because they’re rated for 35 volts; not 25 volts. So I won’t be able to mount the replacement directly to the printed circuit board.
07:50 PM: So, I soldered two wires to the lands where the original capacitor had been connected. There appears to be enough space inside the cabinet that I can either tape or glue the new capacitor elsewhere, and then run these 4”-long wires from the PCB to it.
08:00 PM: I connected the new capacitor and then powered up the unit. The hum was gone, and the speaker still passed signal okay, so, thus far, I’ve not damaged anything. I let it run for a few minutes.
08:05 PM: However, I then heard a hissing noise coming from the new capacitor. I touched it, only to find that it had become very hot and that it had begun leaking electrolytic fluid onto my work bench. Immediately, I unplugged the speaker from the wall outlet. Hmmm. Might I have connected the capacitor backwards? With polarized electrolytics, you have to use the right polarity; else they’ll hiss and be destroyed, with possibly violent explosions.
08:10 PM: I double- and triple-checked the polarity. It was correct. So I began suspecting that the replacement capacitor I had connected was itself, bad. So I connected the second of the three of these I had.
08:20 PM: Same results. The second replacement also heated up, hissed, smelled, and leaked fluid. Hmmmm.
08:25 PM: Then, I remembered my electronics teacher at WPSBC talking about how, over time, electrolytic capacitors can become de-polarized when not used for several years or more. This could very easily be the case here, because I purchased these capacitors on eBay back in 2006, and even at that time, they weren’t new; not even being new, old stock (NOS). In fact, these ones were definitely used, as they had extra wires soldered to them. So it’s hard to say what condition these might be in.
08:30 PM: But I’m getting sleepy now. So I’ll get back to this project tomorrow. Stand by.
November 23rd, 2009 at 10:31 am
11:30 AM: Good day. Anyway, continuing: I only have one of these capacitors left. So I’m going to hook it up to my variable voltage power supply for a few hours in an attempt to re-polarize it.
12:30 PM: So far, I’ve applied 15 volts DC across its leads and it’s not getting hot. Next, I’ll increase the voltage to 25 volts. If it does okay with that, then it should work well in the speakers; assuming of course, that there’s no other problem with the electronics in there. I did check the power diodes and they should no reverse-bias leakage current. So I don’t think it’s the case that the first two replacement capacitors got hot because of the presence of AC.
07:05 PM: Well, I’ve left the capacitor on the power supply all afternoon and it’s neither got hot nor hissed at all. So now, I’m going to connect it to the speaker power supply and see if behaves any better than the other two I tried did yesterday. Stand by…
07:25 PM: Well, when I put the new capacitor in the speaker, it began heating up, just like the last two, even though it did not get warm all afternoon on the bench test supply. So I checked the polarity again and guess what? I had been hooking them up backwards.
No wonder I was getting the heating and hissing. 
So I reversed the polarity and all appears to be well; no hum, and the speaker passes the test signal cleanly. I’ll let it run for a few hours to “burn it in” so to speak, before putting the unit back together.
08:00 PM: I ran it for a half-hour or so and it worked great. So I strapped the new capacitor in place with a nylon ratchet band and soldered the leads permanently into the circuit. I then put the cabinet back together and gave it to [Emmy] to use to listen to the rest of the hockey game with; that will give it a potent workout, for sure.
08:05 PM: Barring any problems discovered during the burn-in process, I’d say we’ve solved this problem.
November 24th, 2009 at 6:14 pm
08:15 PM: The speakers worked flawlessly for several hours last night. So we’ve definitely solved this problem.