Sabriena’s Speakers: In Surgery
I don’t appear to have written about it, but nearly ten years ago when I bought my 7700k desktop parts, Sabriena got a nice set of speakers as well. They’re AudioEngine 2+, what’s now known as the A2+, but there have been several iterations since she got hers, and they’re basically “budget-friendly audiophile” grade gear, though they were still what I considered shockingly expensive at the time. Still, my wife let me spend a stupid amount of money on a desktop PC build, I wasn’t about to quibble over a set of speakers.
Truthfully? They’re pretty fucken nice. I’m aspirationally audiophile - I like gear, I like things that sound good, but honestly my ears aren’t good enough to tell things apart in a lot of cases, unless something is truly dreadful (that perhaps makes me about the only audiophile on the planet honest about my capabilities though?), but they sound pretty good to my ears. The most common conclusion drawn by anyone talking about these speakers is “amazing vocals, not enough bass extension” but for the sort of stuff I listen to I find the vast majority of stuff has way too much bass anyway, so these sound pretty damn stellar to me, particularly for the money.
Not enough for me to have sprung for a set over the years mind, but yeah… nice enough speakers, have lasted fairly well, drivers are still in great shape. However…
Over the years, they started developing some nasty habits. When we moved into this house, Duncan and Sabriena’s desktops were in what can really be described as a sunroom - basically half windows and half open space, the only place to put their desktops was against the windows. This meant that lots of her stuff baked in the sun, the speakers included.
After being left on for too long, they’d begin to buzz. Not the general background hiss you’ll get from basically any amplifier, but an irritating roar audible from across the room. This progressed, until just before Christmas when she’d get about 10 minutes out of them, and so most of her gaming would be done with headphones, which she hates. She replaced them with the same cheap-shit ali-express tier speakers I’m running, and I can tell she’s not super happy with it. There’s also the issue of the volume control knob being scratchy as all get out.
But I’m a fiddlefingers with just enough electronics knowledge to make me dangerous to myself, we should be able to fix this right?
Crack them open, and from a cursory inspection everything looks nice enough. No caps leaking, everything looks in reasonable shape. A lot of it is globbed in either hot glue or epoxy, making service a bit more difficult, but a hypothesis began forming… the amplifier IC (a TDA7265) is bolted an alloy heatsink, which sits in the path between the woofer and the bass reflex port. The heatsink is then coupled to the metal back plate of the active speaker, though there’s no fins or anything on the exterior. But surrounding it is a bunch of globbed up thermal compound (I’m sad I did not take a photo before cleaning it), so I theorized that over time the thermal compound had marched out due to thermal cycling, and the IC was overheating, causing the buzz. This idea was reinforced by the fact I’d been beating the shit out of these things with 90s alt rock and shitty punk for two full days with the back cover off and it had not made any untoward noise once.
So after a long week at work on Thursday, I cleaned it all off with some rubbing alcohol. At this point I noticed plastic between the heatsink and the IC, and memories of people cooking their CPUs because they neglected to remove the plastic cover from the heatsink face came back. I peeled the plastic off, cleaned up the surfaces, smeared some of Jaycar’s finest transistor thermal compound it, screwed the IC back to the heatsink and let it have it again… cool as a cucumber this time!
Satisfied, I began to reassemble it, while still playing because I’m a neanderthal and whatever song was on was good, when it shot fire at me. Big sparks. Huh, that’s weird. I looked around for me having pinched a wire or something, when I simultaneously remembered two things: first, a line that my subconscious had clearly absorbed from the datasheet: “tab connected to pin 6”, and those fucking bits of plastic. Heading back to the datasheet for the amplifier, and yeah, pin 6 is -Vs, or -17.5VDC or whatever it is after any power conditioning in the circuit. Whoops, that’ll do it.
Sheepishly, I cleaned off the rest of the thermal compound from the plastic slices (there are actually two, so it’s heatsink->compound->plastic->compound->plastic->compound->IC), gooped the lot of it up, reassembled, and off we go.
At this point I’m starting to feel pretty good about myself as Mr Fixitman, when I went to adjust the volume and the noise came back. That’s weird, it’s never done that before. Sure enough, yes, the noise appears to be related to the fucking scratchy pot, I misdiagnosed the shit out of this thing, the thermals were probably fine. Well, not fine, it was definitely hotter when it’s making the noise, but I think the heat is a symptom rather than the cause.
Peering at the volume pot under a magnifier, the only markings I can make out appear to be B503, but by poking around on electronics sites and comparing specs, I’m fairly sure it is a Bourns PTR902-2015K-B503. It’s sealed up good, but in my desperation I gently drilled a tiny hole in the casing and fed contact cleaner into it, exercising the thing back and forth… not sure if I just can’t get it in the right place or what, but I think the pot is too far gone for my skills to salvage. It’s better, but still scratchy as fuck and still does the noisy thing.
I’m not 100% sure I have the make/model correct, but I couldn’t find anything that matched better. It’s weird that the B503 is a linear curve, I was under the impression that volume knobs normally use a logarithmic curve? But that’s definitely what’s stamped on it so it lines up with what’s on the datasheets.
Alas the PTR902s appear to be unobtanium now. I can find the exact same model number in a PTD, which is the same thing without the rotary switch. TT made a P092 which was similar but is also discontinued and out of stock everywhere. I just can’t find anything in stock that’s the same specs! So the following ideas have come to mind, and I’m not sure I’m good enough at electronics to work out which is the most sensible:
- Do I drill another hole and have another go at cleaning the pot up? I can glob them up with epoxy or something afterwards. I do risk punching the drill into the juicy bits and destroying it though.
- Can I use something else? Like a log curve part that’s otherwise the same? Not sure if they’re available.
- What about 100kohm instead of 50kohm? It’d just mean it’d go down quieter and the curve at the top end would be wrong (it would get louder much faster near the top), but Sabriena basically doesn’t touch the volume knob once it’s on anyway, tending to use software instead.
- Speaking of, can I cut the pins, leave the switch in place, and just solder some appropriate resistors in place to lock the volume? Can I just bridge it to leave the volume at 100% and just use software to limit it?
I’m not sure what to do at this point. I wrote AudioEngine’s support folks to pick their brains, and while they’re friendly they’re (justifiably, I suppose) not super forthcoming with information for someone who’s not an authorized tech, but what I can do is buy a replacement back panel, with an updated (USB-C!) amplifier board that’s a drop-in replacement for about $100AUD, so that’s an option too.
