My QSC speaker fix, 5 years late! (LONG-you've been warned!)

[Reezekeys]

Here comes a novel, sorry. More than five years ago (!) I posted a plea for help with a problem in one of my two QSC K8 speakers. One of the two channels stopped passing audio and had a nasty buzz that made it unusable. Thankfully it originated before the level control, so by turning that control fully ccw I could eliminate the buzz. The other channel was functional so I switched to using that one.

 

I"ve always been a DIY type when it came to electronics, even though I know nothing about circuit theory beyond ohm"s law. I also knew that QSC service centers don"t do component level repair so they would want to replace the entire board â which costs around $300 plus labor. I decided to try to fix this myself.

 

I got an incredible amount of help from the users here. Some of the folks that helped me were MoodyBluesKeys, JeffLearman, WesG, BlueAndGreen, icarusi, and Karl Schmeer. One person in particular, MIDI2XS, was super-patient and helpful. After I shared the schematic, he had me measure resistances at various points and finally diagnosed a bad opamp. I"ve been an electronics hobbyist since my early teens but had never touched a board with SMDs before. Added to that, this opamp was surrounded by two large combo XLR jacks soldered to the board that made maneuvering a soldering pencil or heat gun to this opamp impossible â they would have to be removed first.

 

Here"s that original thread, if you"re a masochist and have time to kill: https://forums.musicplayer.com/ubbthreads.php/topics/2691393/

 

Anyway - I wound up putting this project on hold. The channel that went out has a mic/line switch (it just switches a pad in & out), while the good channel is line-level only. That meant I lost the ability to use the 'mic' (high gain) setting, but in actuality I only needed this setting when doing super-loud gigs like weddings â and I"ve stopped doing those gigs. So, this broken channel didn"t impact me much. I figured that someday I would get to the fix.

 

Cut to a few weeks ago when I went for a deal on a Chinese rework/soldering station. I got it because I"m about to embark on another project I started years ago â rehabbing by OBXa (I believe there"s another thread here about that. Stay tuned for more, lol!). I figured since I now had better tools (a statement I know the experienced EEs here will laugh at!) I thought I would give my QSC repair job a shot. Enter another super-helpful friend, Ethan Winer. I"m betting many of you know who he is, but in case you don"t he"s the guy that started the acoustic treatment company RealTraps. He"s also a long-time electronics projects designer and contributor to many magazine articles about audio. He had a recording studio in the 1980s with a console he designed and built himself. He also built a synthesizer that was featured in a Keyboard Magazine 'Keyboard Of the Month' article. I"ve known Ethan since my teens and he"s always been very generous and helpful with his knowledge. I recently spent a few hours with him where I got to practice soldering a SOIC opamp to a spare board he had. In an amazing stroke of luck for me, he also happened to have a few opamps that were not only the same configuration (dual), with the same exact package as my bad one (SOIC), but were also 'premium' grade, with much better specs than what I had â the Texas Instruments OPA1602 which as I understand is used in some high-grade audio equipment (higher-grade than my QSC at least!).

 

I know this is already way too long so I"ll just cut to the chase. Here"s my fancy-pants rework station (hey whaddya want for $50?) and a few other tools I had to do this:

 

 

I didn"t do too well removing the XLR connectors. I managed to lift two traces right off the board. Luckily they only served to short some pins on the connector itself so I reconnected them with wire-wrap wire.

 

 

I also messed up removing the bad opamp - two of the pads lifted off. Those were definitely attached to traces going to the rest of the circuitry so I thought I was probably f-ed! I carefully pressed them back, then spent quite a while maneuvering the new opamp into place and soldering it. The pics tell the story, it was quite the hack job! I truthfully thought I"d be extremely lucky if this worked.

 

 

Drumroll⦠my lucky streak continues! The channel is fixed. I almost can"t believe it. I had the speaker on last night and played through it for a few hours. All seems good, but there"s one issue and that"s shown in this next picture - I think my wire-wrapping fix on the XLR jack"s terminals kept me from seating it fully on the board, so it stands a little higher now. When I tightened the screws to the standoffs, the board got slightly bowed. I"m not gonna try to re-solder the connector, as I think that"s risking me lifting more pads or traces. My plan is to hit the hardware store for some washers to lift the board off the standoffs a little more. I"m really hoping that does the trick. Something tells me that a bent board, in a piece of equipment that"s gonna get banged around a bit, is not a good idea.

 

 

If you"re still reading this - congratulations! Thanks for hanging in and thanks again to all the folks that helped me out five years ago. I see that MIDI2XS hasn"t posted in a while, I hope he"s doing OK. Get ready, my OBXa rehab project is gonna start up soon!

[Markyboard]

Way to go Rob! Soldering gets easier and cleaner the more you do it. But as long as you make solid connections that don"t short to adjacent pins or components it really doesn"t matter. Having the right tools, good eyes and of course skills makes all the difference. The techs and especially professional assembly people that do this all the time put me to shame.

 

I missed the original topic in 2015 but enjoyed the read. I have also received excellent help from MIDI2XS. I too hope he"s ok.

[Reezekeys]

Thanks Mark! I did my best and as I said, I think I got lucky. One thing is that I never checked or asked anyone what the correct temperature for the iron should be. Before, with my Rat Shack pencil, I just plugged it in. Now I have to select a temperature? It's probable I had it set too high and that may have been what took off those pads. Later on I backed the temp down while checking to make sure I could still melt the solder, and things seemed to go better â though I had to hold it to the joints longer before I could see the solder flow. I managed to nail the side of the opamp with the pencil too, as you can see from the mark showing on the picture. I guess these SMDs are built to take some heat!

 

BTW if you've ever experienced the "Japlish" manuals of old Roland gear as I have, I can tell you they don't hold a candle to the "Chinglish" manual for this rework station - it's virtually incomprehensible! I'm not complaining though. I got a pretty nice assortment of tips and three different width nozzles for the air gun. My next project is recapping all eight voice cards for my OBXa. I got the kit with all the components from Synthchaser earlier this year. Gotta do this before those electrolytics dry out! There's trimpots and a switch array in that kit too. I did finish redoing the bushings a while back.

[Markyboard]

I keep my iron at about 640 but I"m due for new tips. I"ll back it off if it starts causing problems which sometimes it does. I"ve had more problems with my desoldering gun. Short duration is key to not lifting vias /traces, especially boards that are more vulnerable to this type of thing. I always add solder on the points before desoldering. Also the rosin in solder acts as a cooling agent so best to continually feed solder anytime you"re applying the iron to the board.

 

Synthchaser is great- love his videos and his kits are well worth avoiding the pain of hunting down individual parts. Looking forward to your OB exploits.

 

Btw today I turned down someone who wanted me to fix their Jup 8. It"s easy to get burnt out on this electronics stuff and my own needed repairs seem to come in 3s or 4s. Time to play.

[OB Dave]

I love posts like this! Good work, and congrats on your debug & rework. This is pretty ambitious for a first go at SMT.

 

I was really hoping, though, that you were gonna say you found a drop-in replacement for the crappy TRS/XLR connectors that QRS used on the early K-series. A problem they never came clean about until after my K10s were out of warranty.

[Reezekeys]

Mark, you might not remember but you helped me out in another K8 repair thread I had here â this was about four years ago and involved an inductor where one lead broke off due to movement & flexing. The later rev QSC boards had a wire tie to stabilize this piece but mine didn't. After I installed the replacement (which you found for me!) I drilled a hole in the circuit board and put the tie around it. Guess what - I never did the other speaker (the one I've been working on now) - I decided to deal with that today since it was already apart:

 

 

I keep my iron at about 640 but I"m due for new tips. I"ll back it off if it starts causing problems which sometimes it does.

 

I was gonna post about how my cheap rework station only went up to 480 degrees before I realized it was in celcius! No wonder I had it set to 275 and it was still melting solder. Your 640 is fahrenheit right?

 

I"ve had more problems with my desoldering gun.

 

Do you mean the plunger guy like I have? (It's in my first pic). I find it hard to work it while I'm applying heat with the pencil â pressing the button to release the plunger shifts its position away from where i need to do the sucking. I usually have to repeat the process a few times at the same spot - which means applying the pencil to the same spot over & over. I think that's what's ruined a few pads for me.

 

Short duration is key to not lifting vias /traces, especially boards that are more vulnerable to this type of thing. I always add solder on the points before desoldering. Also the rosin in solder acts as a cooling agent so best to continually feed solder anytime you"re applying the iron to the board.

 

Great tips, thanks. I was wondering if I needed to get a separate tube of flux. I watch Louis Rossman's videos on youtube, that seems to be his method â he puts a gob of flux down first. There's a guy that makes all this look very easy!

 

Synthchaser is great- love his videos and his kits are well worth avoiding the pain of hunting down individual parts. Looking forward to your OB exploits.

 

I paid $76 for the full rebuild kit for all eight voice boards. A bargain, considering the time it would take me to source everything â or even figure out what needed to be sourced! Time to get this ball rolling.

 

Btw today I turned down someone who wanted me to fix their Jup 8. It"s easy to get burnt out on this electronics stuff and my own needed repairs seem to come in 3s or 4s. Time to play.

 

I'm not burned out on this work myself, but I have burned out a few electronics components in my time! I totally get it. I'm actually in the middle of repairing a dishwasher right now. Can't wait to get that over with!

 

I was really hoping, though, that you were gonna say you found a drop-in replacement for the crappy TRS/XLR connectors that QRS used on the early K-series. A problem they never came clean about until after my K10s were out of warranty.

 

I remember reading a few posts about these connectors causing problems but I never experienced any that I know of. Strange because I got my Ks a few months after they first came out â mine had the loose inductor issue mentioned above, which accompanied rev A power amp boards. Maybe they changed suppliers for the connectors somewhere along the way?

[Reezekeys]

All done, all put together, and it still turns on and passes audio!

 

My idea of using washers to level the circuit board worked. The panel controls are not perfectly aligned now, but the difference is minor and all switches & pots work fine. I can't say I'm proud of this part of my repair, but I wasn't about to start taking apart and desoldering stuff again to make it look good. I've owned these puppies 11 years and I'm confident I'll get quite a few more out of them. I have the fine EEs on this forum to thank for this. I couldn't have done it without you guys!

 

[Markyboard]

Let"s see if I can respond without more quotes.

 

Yes, 640 Fahrenheit. We"re still days away from going all metrIc here in the US and I wanted to take advantage

 

I have a Hakko powered desoldering gun. There"s no going back to those plungers as this is so much easier and faster. But like I said you have to be real careful using very short duration contact. Also if you followed my P5 refurb thread there was the solder melting on the opposite (component) side of the board that caused solder blobs. I"m still not sure what was different about those boards as I haven"t had the problem since.

 

Definitely good to have flux on hand. Most of the time I don"t need it but it depends on the board, the solder used (solder often has flux in the core) etc. If the solder isn"t taking to the board then yeah, it"s a good idea.

 

Don"t get me started on home repairs. I"ve pushed my limits this year and still live to tell about it. Natural gas lines, electrical box rewiring, plumbing. Check.

[Reezekeys]

I did some googling on that desoldering gun. They aren't cheap - best I saw was a few used ones for around $150 like this Weller. I figure by the time I'm done with my first OBXa voice card I'll know whether or not to bite that bullet. For now it's plunger and solder wick.

 

Yes let's not get started with appliances... I did bother some good folks at the appliantology.org forum â but as Buddy Rich once said, "not over here"!

[stoken6]

Man I'm in awe of this thread. I can just-about-adequately solder cables, but anything on a circuit board scares me.

 

Cheers, Mike.

[Markyboard]

This is probably the best manual solder sucker available.

 

 

 

I found it to be much better than the blue plastic type. But for big jobs like refurbs where you're changing out a lot of components a powered desoldering gun is really the only practical option.

 

Btw Rob I did end up with 2 of the above so if you want one PM me. You'll just need replacement tubing which you'll see on the link above is just a couple of bucks.

[elif]

Bravo!

 

There is definitely a reward other than merely financial that one gets by repairing rather than replacing.

 

Concering desoldering opamps and other leaded components from PCBs when using only a soldering iron, cut the leads and remove the body of the part. Then you can desolder each lead individually. You may need a better diagonal cutter than you currently have to isolate a lead to snip it.

 

Plastic thru-hole components are a problem to remove without damage to the PCB (XLR and other connectors). I have used a dremel tool, a jeweler's saw, diagonal cutters, whatever would work to cut away and remove the plastic body and any mounting tabs of the part from the PCB to to isolate the leads for removal.

[Reezekeys]

Thanks for those tips CH, I'm learning a lot in this thread and I know it's gonna come in handy when I start my OBXa voice card project.

 

Regarding the cutters â Ethan lent me these, they look similar but not as good as the one you linked to:

 

 

He said the same thing you did â that they'd come in handy removing the bad op-amp. You're both on the money there â I used them, but still managed to lift two pads off the board! iirc I couldn't reach one side of the opamp with these cutters because it was too close to the xlr jacks I didn't remove.

 

Mike - believe me I was quite scared putting my soldering pencil to this circuit board! I still think I was extremely lucky. I keep wondering when my luck is going to run out!

[Theo Verelst]

So it was the opamp anyway ? Good, it's like the first chip acts as some sort of protection for the rest of the circuitry, plugging cables output first input next aside, it is possible to overload the chips by cable plugging (only not when input and output devkce being plugged are both grounded, or mains disconnected.

 

It's possible changing the opamp type makes the digital processing a bit "off", even if it's got better specs, might be interesting to know. Like the imperfections can be tuned for one reason or another, even the buffer function needs to be considered when improving opamps.

 

T

[elif]

... Regarding the cutters â Ethan lent me these, they look similar but not as good as the one you linked to:

Those look like flush cut. Obviously, they worked well. I had one of those white handled Lindstrom cutters until my father-in-law reached into my toolbox and used it to cut a piece of 3/32 steel music wire, instantly turning it into a wire stripper.

[Reezekeys]

Those look like flush cut. Obviously, they worked well. I had one of those white handled Lindstrom cutters until my father-in-law reached into my toolbox and used it to cut a piece of 3/32 steel music wire, instantly turning it into a wire stripper.

 

Yes, Ethan called them flush-cut. They look similar to the ones you linked to but from your reply I assume they aren't quite the same? I see a Lindstrom flush-cut guy on Amazon â it's $80! That's pretty major coin for a cutter if you ask me, but the cutting head looks very precise and well made, so I assume that's par for the course for a pro-level tool like this. Sorry to hear about your Lindstrom.

[Reezekeys]

So it was the opamp anyway ? Good, it's like the first chip acts as some sort of protection for the rest of the circuitry, plugging cables output first input next aside, it is possible to overload the chips by cable plugging (only not when input and output devkce being plugged are both grounded, or mains disconnected.

 

It's possible changing the opamp type makes the digital processing a bit "off", even if it's got better specs, might be interesting to know. Like the imperfections can be tuned for one reason or another, even the buffer function needs to be considered when improving opamps.

 

T

 

I was thinking it would be interesting to see if this better opamp made a difference in the sound of the speaker. Since I have two K8s - the one I just fixed, with the OPA1602, and my other one with the stock MC33078, all I need do is feed the same signal to both & see if one sounds or measures better, right? Just a few issues though: first, I have no test equipment like a signal generator or scope. Second, the speaker is a closed system, so my feeling is that the only meaningful measurement would have to be done on the sound coming off the drivers. Another guess is that you'd need some kind of calibrated microphone to do that. I suppose you could tap into the power amp output stage and feed that to a signal analyzer, but what would any differences shown there matter if you couldn't be certain they'd be audible or not? I suppose some gearheads might find it interesting but I'm more interested in the practical.

 

Despite all this, for grins I hooked up my laptop's headphone output to both speakers, into their A channels, with the mic/line set to "mic" (high gain), then turned the levels all the way up. I did not play music through this â that would have blown the roof off my house (or just shut the speakers down, lol). I did stick my ear right next to the woofer and the horn of each speaker, one after the other. I thought I might hear that that my newly-fixed K8 would at minimum exhibit less quiescent noise than the other one â the 1602 is known for that, I think. Well, it certainly might have - but whatever advantage it might have was easily overridden by the digital hash coming from the headphone output of my laptop through an unbalanced cable! I know I should have shorted the inputs or hooked them up to a real piece of audio gear using balanced cables, but seeing as I use these speakers only with the laptop, I didn't feel the need to go any further with this little experiment! Not to mention that the K8's DSP algos might be involved in altering any low-level artifacts of its class-D power amps anyway.

[MIDI2XS]

[...]

Drumroll⦠my lucky streak continues! The channel is fixed. I almost can"t believe it. I had the speaker on last night and played through it for a few hours. All seems good, but there"s one issue and that"s shown in this next picture - I think my wire-wrapping fix on the XLR jack"s terminals kept me from seating it fully on the board, so it stands a little higher now. When I tightened the screws to the standoffs, the board got slightly bowed. I"m not gonna try to re-solder the connector, as I think that"s risking me lifting more pads or traces. My plan is to hit the hardware store for some washers to lift the board off the standoffs a little more. I"m really hoping that does the trick. Something tells me that a bent board, in a piece of equipment that"s gonna get banged around a bit, is not a good idea.

 

If you"re still reading this - congratulations! Thanks for hanging in and thanks again to all the folks that helped me out five years ago. I see that MIDI2XS hasn"t posted in a while, I hope he"s doing OK. Get ready, my OBXa rehab project is gonna start up soon!

Thanks for the update, and it's great to see that the repair was successful!

 

Thanks also for well-wishes - I'm OK, just otherwise busy and my dial-up internet connection limits my ability to participate in this forum.

[Reezekeys]

Nice to hear you're doing OK, and thanks again for helping me diagnose my speaker issues. Dial-up? Is that by choice? Maybe you should look into one of Elon's Starlink terminals!