Results 1 to 3 of 3

Thread: Speaker Amplifer - Repair and Capacitor Replacement

  1. #1
    Join Date
    Jan 2018
    Location
    Near Stuttgart, Germany
    Posts
    246
    Blog Entries
    25

    Lightbulb Speaker Amplifer - Repair and Capacitor Replacement

    A french NSX owner, active on NSX prime sent me his Bose speaker amplifiers (and the radio- but we'll come to that in a separate thread) for repair. The car has been left in a garage for a considerable amount of time and the audio system was non-functional as not even the head unit would turn on. The PCBs were from 1992 and in a bad state:

    Click image for larger version. 

Name:	IMAG4889s.jpg 
Views:	302 
Size:	56.4 KB 
ID:	14202
    right door - cover
    Click image for larger version. 

Name:	IMAG4886s.jpg 
Views:	652 
Size:	98.6 KB 
ID:	14201
    right door - PCB top
    Click image for larger version. 

Name:	IMAG4887s.jpg 
Views:	299 
Size:	94.8 KB 
ID:	14203
    right door - PCB bottom

    The right door amplifier hat a blown MOS-FET transistor which charred the PCB. The Bose PCBs are rather robust and withstand a lot of abuse but there are limits, of course. The reason for this failure is (of course) capacitor leakage but also due to the design of the amplifiers. Their main ouptut stage is a MOS-FET transistor bridge that connects each speaker pin to either GND or VBAT - that's why there are four transistors.

    Click image for larger version. 

Name:	IMAG4885s.jpg 
Views:	321 
Size:	97.8 KB 
ID:	14213


    The downside of this design is, that it fails catastrophically if a pin is connected to GND and VBAT at the same time - which happens if both transistor pairs are conducting at the same point in time. Such situations are usually avoided by switching the transistors correctly but in this particular failure case the acid from the capacitors carries voltage to one of the transistors, making it conduct permanently - if the second transistor is now activated it becomes a direct connection between GND and VBAT. Since the acid-affected MOS-FET typically conducts only a little at first, it keeps a remaining resistance which causes it to literally heat up and burn.

    Click image for larger version. 

Name:	IMAG4893s.jpg 
Views:	305 
Size:	58.0 KB 
ID:	14204
    left door - cover
    Click image for larger version. 

Name:	IMAG4890s.jpg 
Views:	555 
Size:	103.4 KB 
ID:	14208
    left door - PCB top

    Click image for larger version. 

Name:	IMAG4891s.jpg 
Views:	304 
Size:	96.6 KB 
ID:	14209
    left door - PCB bottom


    Click image for larger version. 

Name:	IMAG4897s.jpg 
Views:	308 
Size:	56.1 KB 
ID:	14210
    centre - cover
    Click image for larger version. 

Name:	IMAG4894s.jpg 
Views:	591 
Size:	98.8 KB 
ID:	14211
    centre - PCB top
    Click image for larger version. 

Name:	IMAG4895s.jpg 
Views:	308 
Size:	101.0 KB 
ID:	14212
    centre - PCB bottom

    Both left side and centre speaker amplifiers appeared normal from a first look. Due to the contamination with capacitor acid, all of the hot glue, all capacitors and the coils had to be removed. The PCB of the right door amplifier was washed with water and a brush, then rinsed with natural alcohol. Contaminated (dark coloured) traces were scraped of and polished with a fibre pen, then covered with solder. New capacitors installed and powered-up for a test.

    .. to be continued
    Last edited by Heineken; 08-03-2021 at 08:01 PM.
    1997 JDM Custom Order AT VIN 1400005 - Stock
    Heineken's Garage

  2. #2
    Join Date
    Jan 2018
    Location
    Near Stuttgart, Germany
    Posts
    246
    Blog Entries
    25

    Default

    Unfortunately, nothing happened. Checking the MOS-FET transistors reviled two potentially broken ones - their measured resistance remained too low to be correct. Event after replacing them with new ones (IRFIZ24N being a popular exchange type) no sound came out. Further analysis turned out that one of the Bose special ICs was broken - can't be repaired without a replacement from another amplifier.

    One down, two more to go. The centre speaker amplifier worked fine after the same treatment so it got a drop of hot glue to stabilize the coils and new conformal coating to protect the copper traces. Same for the right door amplifier with the burned transistor. Repairing such damage is not exactly optimal but despite a real alternative it seemed like the only sensible path to take.

    Click image for larger version. 

Name:	IMAG4933s.jpg 
Views:	570 
Size:	141.6 KB 
ID:	14215
    a repaired center amplifier (glue missing)
    Click image for larger version. 

Name:	IMAG4935s.jpg 
Views:	534 
Size:	136.5 KB 
ID:	14214
    a repaired right door amplifier

    It's kinda sad that only two of the amplifiers could be brought back to life but much better than none. Usage of IRFIZ24N MOST-FET transistors as a replacement can be confirmed, which helps to avoid having to source "official" Bose parts which are sold for a premium.

    PS: A blog entry on the repair of these amplifiers has been created.
    Last edited by Heineken; 25-03-2022 at 09:16 PM. Reason: wrong labeling
    1997 JDM Custom Order AT VIN 1400005 - Stock
    Heineken's Garage

  3. #3
    Join Date
    Jan 2018
    Location
    Near Stuttgart, Germany
    Posts
    246
    Blog Entries
    25

    Default

    The amplifier on the right picture just came back to my desk - with a blown capacitor. The reason can already be seen on the picture above: Wrong polarization of the lower power stabilizing capacitor. Both red markings need to face towards the outside of the PCB ..

    Click image for larger version. 

Name:	IMAG6089s.jpg 
Views:	193 
Size:	77.1 KB 
ID:	14568
    Boom!
    Click image for larger version. 

Name:	IMAG6090s.jpg 
Views:	179 
Size:	152.3 KB 
ID:	14569
    Repaired (again)
    Click image for larger version. 

Name:	IMAG6091s.jpg 
Views:	185 
Size:	107.8 KB 
ID:	14570
    Final result

    It's now playing fine again after cleaning the PCB from the black dust and replacing the exploded capacitor. Unfortunately, the Bose chip U2 died as well so an IC chip socket and a new Bose chip was required, too.
    An annoying mistake of mine but with these major take-aways:


    • Double and triple check the polarity of the capacitors
    • Aluminium Polymer capacitors can work with the wrong polarity for several hours before failing (catastrophically)
    • Ensure a prolonged test run
    • A shorted capacitor can kill the Bose U2 IC


    Did I forget anything ? .. hopefully not ..
    1997 JDM Custom Order AT VIN 1400005 - Stock
    Heineken's Garage

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •