This is a repair of Bose L1 tower PA system. These skinny tower speakers are somewhat of a marvel because they are designed to stand behind the musician (aimed RIGHT AT the microphone) and don’t generate feedback (in most situations). Ah, the miracles of DSP.
It turns out these PA’s have a frequent failure in their power supply. Read on for details.
UPDATE 10/22/2021: Since I first published this, I’ve gotten a lot of inquiries from people regarding repairing their L1s. For people local to northern CA, I’ve swapped working units for their bad ones plus cash. Then I try to fix the dead one I took in trade. All together now I’ve worked on five units and succeeded fixing all of them. I’ve also successfully repaired an aux supply board that someone extracted from their L1 and sent to me (Unit 3 below). For those trying repairs at home, here are my results (I’ll try to update as I see more units):
6 of 7 failures have been in the “Auxiliary Power Supply Board” (read below for more details). I recommend starting trouble shooting there if you L1 does not power up but please note the dangers of troubleshooting circuits that are hot to the AC line (noted elsewhere in this post).
Unit 1: My first attempted repair which started this blog post – D608 open – REPAIRED
Unit 2: D601 and D608 bad – REPAIRED
Unit 3: D601 bad – REPAIRED
Unit 4: D609 bad but also Transformer T601 open between pins 6 and 8. I was lucky in finding a broken wire that was accessible outside the windings. I reconnected the wire with some very carefully soldering. – REPAIRED
Unit 5: Had a working Aux Supply, but no sound from one half of the tower. The +/-27 power supply (there are three) for that channel power amp had smoked MOSFETs and other components. I harvested a +/-27V module from Unit 4 to get this one out the door, then repaired the supply by replacing two MOSFETs, two zeners, several resistors, and one rectifier.
Unit 6: A bit of a train wreck with many visibly charred components. I ultimately had to replace D606, D602, D614, Q601, R614, R620, R621, SCR601, TH603 and Z601. Whew! But now back from the dead. – REPAIRED
Unit 7: replaced D602 and D607 – REPAIRED
Tally so far on L1 repair is 7 for 7!
Back to the original blog post:
The tower speakers slide into the base where they are both supported and electrically connected to the amplifier. All the circuitry is in the base.
The bad news is if you have a problem with this first generation L1 (ours is about 10 years old), Bose is zero help. They will no longer repair them and only implore you to get a new system. To make matters worse they are completely uncooperative about supplying schematics (though I was able to find documentation on-line).
Our L1 suffered a power supply failure, which is apparently fairly common as these units age. The culprit is the “Aux Power Supply” shown in the block diagram at this link.
The below linked doc has schematics and instructions for disassembly. Even though it provides guidance, I still recommend taking LOTs of pictures. Especially of where connectors plug in.
Dissembling the base is no small task. There are screws everywhere, and you have to pull almost everything apart to get at anything useful. Below shows the top cover and control panel PCB after unplugging from the power chassis.
Pull the U-shaped top cover and you are in.
I pulled out the Aux Power Supply board (upper right corner) and the EMI filter board (lower right corner) in order to troubleshoot the power supply on the bench.
HUGE WARNING! The EMI filter board and large parts of the Aux power supply are NOT isolated from the AC line! I powered this from an isolation transformer before hooking an oscilloscope. Make sure you fully understand what this means before proceeding with any testing. Unisolated circuits are VERY dangerous to work on.
Warning #2! The output of the EMI board is about 330V and the big capacitors on that board can store that voltage for up to 30 minutes time after the base is unplugged. Check that the voltage has fallen to a safe level (<20V) before touching anything or removing any boards.
Aux power supply board
I found some unrectified AC waveforms in places that should have been DC, and pinpointed the problem to a large rectifier diode that was open-circuit. After replacing this diode, all the Aux rails powered up, but many voltages were off (by 5V or more) from the spec’d values. I spent several days chasing this with no progress, until I guessed that the amplifier and the rest of the L1 circuitry might need to be connected to the Aux rails before they would read the correct voltages. This turned out to be correct. When I assembled and temporarily wired everything together, the Aux voltages read correctly. These rails are not tightly regulated and needed the load of the circuitry to pull them to their proper voltages.
Other miscellaneous inside images below, mostly to remind me where all the connectors go:
152 thoughts on “Bose L1 Tower PA Repair”
Thanks so much for posting these details and circuits. Enough for me to repair a faulty aux power supply in my LS1 over in the UK. D606 and D601 both failed short, replaced and we’re back up and running. Would have been much more difficult without your guidance and inspiration.
Hi, if it’s of any help, here’s a UK-based guy who repairs Bose stuff that in my experience no-one else will touch. He’s been a great “last resort” help to me in the past. Cheers.
I have a dead unit. Was wondering if I could send to you for repair? I’m in Pennsylvania.
Hi Jason – My experience is that it is not practical to ship these things for repair. The last time I checked it was close to $100 each way, and I expect I will be more now. Also, there is a high chance of shipping damage. If you can find some magically cheap shipping, I’d consider it, but so far I have only worked on units that people have brought me.
Hello lens42 –
Can I power the Bose L1 (first gen) tower and bass cab with a different amp/processor/etc…much like a Bose rack mount amp setup? Basically bypassing the old heavy base. Mine still work great but I’m noticing a noise when I power down on ne of my units. Might be an indicator of bad things to come.
Superb advice – D606 and 601 restored the power – can be monitored at 8 pin multicoloured connector CN4098 when in situ (+15V, 0V, -15V, +3.3V, 0V, 5V, Fault, +24V). Still to diagnose deeper Fault on my L1 Model 1 – any hint on how the ‘fault’ voltage rolls or why all voltages are ok but no power LED green/red – there was once after the Aux PSU was first powered up but no more … thanks in advance – I’ll get the meter and scope out about the power amps and psu’s
Hi Peter – The power amps have their own separate +/-27V supplies, but I’m not sure if a failure in one of those will lock our the power LED.
First off, thank you for this great head start on getting this power supply to work. I have had mixed success with mine, in that, the voltages are there, but not quite to spec: (+11.46, -14.38, 3.68, 7.13, and 18.8V). I have changed out most of the troublesome diodes (D606 was originally a dead short), and otherwise pulled every suspicious component, checking them all with my LCR meter. My spidey-senses say it’s IC603, but I cant prove it.
Do you have any advice as to where I should focus my scrutiny?
Andrew – I mentioned in my blog post that it’s normal to get odd voltages if you run the aux power board with no loads connected. Are you are getting these voltages readings with loads connected or unconnected? If unconnected, you may have fixed it and the outputs will be their correct values when the loads are hooked up.
Hi, thanks for the reply. My Model 1 saga continues (here in UK) now that the aux psu is repaired and all voltages are thereabouts, note the aux psu also sets the power amp voltages (all 7 voltages rely on a single TL431 regulator and a few zeners). Referring to the Digital Signal Processor PCB Rev J, Mic analogue signals disappear into U302 TC9459F volume control and line analogue signals disappear into U385 A/D converter pins 16/17 and 20/21. Although clock signals are present, there is no analogue signal out from U388 2/3 or 5/6. So my problem lies in digital terrain. Has anyone else tried by passing the digital equalisation by connecting CN104B pins 1 & 3 through a couple of resistors to CN115B pin 3? It seemed to work (I don’t have or need a bass unit) and my external mixer can add some equalisation & lpf etc.! PS the non-operating power led is driven from pins 44 & 45 of the mcu via CN119A/B. Maybe I’ll rebox it and the night fairy will reboot it good!
Peter – You have drilled into the digital parts of these amps far further than I. Nice work. If you feel adventurous, the CS5361 (U385) chip can probably be replaced with some careful soldering, but your bypass solution seems reasonable also, and a lot less work.
I bought two of these Bose units new in 2003. I sold one about five years ago when it was working. I hadn’t fired up the other one for a few years, fearing the worst with all of the talk about these failures.
To my surprise, it came on and sounded great….for about 6 hours… then it died. Can’t wait to try the diode replacements to see how things go. It’s too nice of a system to not at least give it a go!!!
When you get it open, first check the voltages on the aux power board outputs.
If I can’t get my system working, do you have any idea where the crossover point is between sticks and subs? I may try to power them with other amplifiers which would require known crossover values.
I was wondering if you are excepting units to repair?
The shipping price is of no problem for me.
Can I send the aux supply?
Or if I turn this unit on can I report what I see to you and receive further debugging instructions?
I have an electronics background and completed the diodes and capacitors replacement to no avail.
I noticed this device is a mobius design in that controls go back and forth from board to board which makes it very difficult to troubleshoot for me.
Thank you for any advice.