Klipsch ProMedia 2.1 repair pt. 2, thermal boogaloo

[nsd_c]

Hi all,

So I previously posted that I had a set of Klipsch ProMedia 2.1 computer speakers (powered sub + a pair of powered satellite speakers) that I repaired. In that post, I didn't go into detail about the (board level) repair, but asked for some advice regarding one of the satellite speakers which worked but seemed off. (The original failure was that a bunch of transistors popped in the satellite speaker amplifier section on one channel, which probably put DC through the speaker, messing up the voice coil, but only enough that the speaker sounded "off" but still worked.)

Anyway, the original repair consisted of making an entirely new PCB to replace one of the amplifier boards; this PCB was just a copy of the old but with new components.

I tested the functionality of this with all of the amplifier guts laid out on my bench, and it worked!

So today, I worked on screwing the amplifier boards back into the enclosure. I then powered the thing up and--nothing. Not a thing. Not even for a second. No light, no "bumpf" turn-on sound, no green power light. Nothing.

Horrified, I yanked the power and opened the enclosure back up. No smoke. No smell. I checked component temps with my thermal camera. Nothing appreciably above ambient, except one bridge rectifier which was a scant 2 deg. C warmer than ambient.

I then realized my mistake: I forgot to put mica insulators back between the output stage transistors and the chassis to which they were mounted. These transistors are TIP41's and TIP42's and are driven straight by some bridge-rectified-then-capacitively-filtered (but not regulated) rails. I checked the secondaries (with power off ofc): these show low-ohms. Primary? Open.

The transformer has a thermal fuse.

So how screwed am I? Where are thermal fuses buried in transformers? Are these able to be replaced? Or should start looking for a donor unit to pull the transformer from?

Thanks all for any advice anyone might be able to give.

[Chris56000]

Hi!

The thermal fuses are underneath the secondaries adjacent to the primary winding, and they're very difficult to replace unless you've had a lot of experience rewinding power transformers!

Obviously if you can find a "parts mule", that'll get you going with the least aggro, but do a search on eBay for "R Core Transformer" , about 100VA you should be able to find one with two 40–0–40V secondaries, or one smaller one, in which case you might have to use two separate transformers, but the one feeding the H.F. (tweeter) units can be a bit lower power rating as the H.F. tweeters don't take as large a power supply current!

If you do use replacement R core transformers, I would add one 20mm x 5mm fuse of 500mAT (anti–surge type) in series with each primary winding if you use two transformers, or a 1.25AT fuse if you can find a single large one!

If you get a dual primary transformer, don't forget to wire the '0' and '115' ends of each primary together and then the two '0s' to mains N and the two '115s' to mains L.

Hope this helps!

Chris Williams

[nsd_c]

Hi Chris,

Thanks so much for the reply! That is some very useful information! I did not know about R-core transformers. These look interesting--I'll have to read more about them.

However, it turns out that I made a rookie dumb mistake--I neglected to notice that there was, in addition to the thermal fuse in the transformer, a 5x20mm 2A glass fuse in series with the primary side. When I tried measuring to see if the primary had blown open, I did so by flipping the panel-mounted mains switch 'on' and measured DC resistance across the mains plug--including the fuse in the circuit. D'oh! Checking the primary directly showed it to be low-ohms (~3 ohms), which seemed possibly concerning.

So, I replaced the fuse and hooked the transformer up to a variac with a current clamp around a lead on the primary side and the secondary disconnected. I then slowly increased the voltage supplied by the variac. Low AC voltage current draw was promising (tens of milliamps), so I dialed the variac up all the way and the draw @ 120VAC was on the order of 80-100 mA or thereabouts (I've forgotten exactly what already), and the secondary voltages read as they should (40 and 80VAC). So, ~10W no load. Seems acceptable?

I have not yet poked around where stepped-down AC enters the circuit, but fingers crossed there's no shorts / residual damage!

[nsd_c]

Update:

I checked resistance between the a variety of points (left channel to right channel, positive to negative rail, etc.) and everything looked good.

I already had one dummy load constructed (consisting of a few 2 ohm high power resistors in series), so I made two more (one each for the left and right channel satellite speakers, and one for the subwoofer), then I powered the whole thing on (boards wired up on the bench this time, and not installed in the unit).

I fed in a test signal and checked AC & DC volts across the loads. No DC (great!). I wasn't sure what to expect in terms of AC volts magnitudes, so I played with dialing the volume up and down, trying (emphasis on "trying") not to leave the volume knob at "max" for more than a second, and trying (again, "trying") to not have the whole unit powered up for very long--with the boards spread across the base, the output driver transistors were just flapping in the breeze as opposed to screwed to the chassis, so they had no heat-sinking.

Welp, considering that I've already made a ton of mistakes, what followed should be obvious: I quickly cranked volume to max, and PSSSHHTTTTTT, out comes the magic smoke in tremendous fashion. Output transistors Q9 and Q10 destroyed (for the Nth time now); death confirmed by shorts between all pins (in addition to what I've dubbed "transistor blood" oozing out of one--see attached image). I quickly yanked power, but the fuse got there first (thank you fuse!). Note that the output transistors only blew on one channel, and on the channel with new transistors... The old transistors (TIP41 and TIP42) were high-quality ST parts in nice looking TO-220 packages: thick tabs, some sophisticated epoxy package shapes which implied a great deal of care. The new ones claimed to be Fairchild by their logo, but seemed very cheap by comparison: thin tabs & very generic, crappy looking packaging.

After putting a new fuse in, replacing the power transistors (again with cheap-o's--I couldn't readily get nice ST ones), and poking around for shorts (none found--seems good) I again hooked up dummy loads & supplied a test signal. No DC, AC response seemed good; AC voltage changed with volume knob. Scope waveforms looked fine (other than some horrendous 100 kHz noise coupled in from the square-wave generator in the class-D subwoofer amp). I nervously hooked up the satellite speakers and powered the system on. It works! Again!

Next I need to re-attempt re-attaching all of the boards to the chassis (this time trying not to forget the mica insulators...)