klipsch kda-500 amplifier suitable power supply capacitor replacement ?

[r2digital]

I have a klipsch kda-500 that would turn on and allow me to get into the web management but it was picking up a lot of noise from audio sources so i decided to replace the smd caps on the power supply along with radial one. Some one was in here before me and they basically burned the bottom side of the traces off for the radial cap but i was able to repair it without any real problem. I thought i did a good job sourcing out replacement caps but apparently not and now i can't get the power supply to turn on it just constantly try's to keep turning on. Attached is a picture of the power supply before and after the cap replacement

Photo 1 shows the caps that are currently in the power supply and also you can see where i had to repair the trace for the radial cap on the top side. I tried to get replacements as close as possible to the originals

Photo 2 are all the caps before replacement. The caps that were identifiable with a series were mostly panasonic FC. However, there's the 2x 330u /25v with no series name, 2x 100u /35V and 1x 47u /16v

[r2digital]

I measured the impedance of the no series/brand caps after removal so i could find something close.

47u /35v - .400 mOhm was replaced with .200 mOhm
47u /16v - 2 ohm and replaced it with 1.7 ohm
330u /25v - 0.270 mOhm and was replaced with 0.08
100u /35v - 0.500 mOhm and was relpaced with 0.16

Does anyone think the impedance is too far off from original caps and thats why its not working ?

[CapLeaker]

Naah… I don't think the cap swap has anything to do with it, However I do question that top trace repair where the radial large cap goes. To me in the picture #1 that doesn't look right to me. Why would there be an isolated pad then joined by a little trace to GND? Even from a manufacturing standpoint it's just dumb, What does the backside look like and where does the trace go?
Does It power up at all or is it dead? If it's dead first thing I'd look at is that 5 legged TO220 thing mounted on a heat sink on the bottom of the pic. These 5 legged devices like to blow up a lot or even play dead. If you came too close to it and some dirt, a piece of lint, flux etc came in between its pins and you didn't clean it absolutely squeaky clean, I mean absolutely squeaky clean, it very well could be dead.

[r2digital]

It looks like it's isolated because i used some solder mask after the trace was repaired so i could definitely see how it looks wrong lol and i wish i had a more clear pic to show for that. The backside of the board had nothing left to solder too but luckily the trace in the center on top next to where the radial positive lead goes was also on the bottom side of the board so i bent the radial cap lead over to that and soldered it. But now im questioning if that will even work? I was over by that 5 legged device with flux and i even bent that little board out of the way a tiny bit so i could solder on the capacitor nearby with out issue. The red light comes on and i hear the power supply making some noise but after a few seconds it shuts off and keeps attempting to turn on repeatedly. I attached some more pictures of the board

Edit : Removed radial cap for clear pic of trace on top of board

[CapLeaker]

O.k, sounds like it has a problem with one of the output rails of the secondary. If there would be something wrong with the primary, it either would be dead, or it would went out with a nice BANG! Somehow there has to be a STBY voltage and hopefully that isn't clunking out.
Measure all the voltages on the secondary side of the PSU and make sure they are there, How does this PSU turn on? Is there a switch somewhere or does it turn automatically on and off like a powered speaker?

[r2digital]

there is a rocker switch for main power that turns the amp on directly into standby(solid red light) when in standby the amp is able to be managed through a web gui. The amp is currently configured to turn on by audio sensing 5mv but could be changed to always on if i was able to get the power supply going again. I didn't hear any loud noises or bangs so i was pretty surprised when it wasn't working after the cap replacement. I will double the voltages when i get home later today but does the trace repair look suitable or should i try to solder the cap to the top side or the board ?

[r2digital]

right to left




4 pin

brown -8.4v

black 7.6

white -.006

gray -.006




5 pin

green 32

blue 3.3

green 2.6

red/white 3.0

red 1.1




3 pin

yellow .770

light blue -.016

yellow 13.5




3 pin

black 165

brown -166

red -165

[r2digital]

I tried resoldering the positive lead of the cap to pin 1 on the transformer which is where the trace goes but still no luck

[CapLeaker]

There should be a rectifier after the transformer then the cap.

[r2digital]

I checked several times, there is no rectifier there. It goes from that positive cap trace that was repaired directly into the transformer. I was able to check for continuity from the transformer pin on top to the repaired trace on the top of the board

[r2digital]

ok, i have a pretty good update. I used some freeze spray on the 5 leg TOP265EG and the power supply started and stayed in stand by a few times but also a few times it turned it self completely on and was detecting signal present with in all 4 channels.Im almost positive it is bad

[CapLeaker]

Just as I mentioned in my post #3! These 5 legged devices always have been a problem source.

[r2digital]

Yup! you were right all along so thank you again for your help. I'll post back in a few days with an update once i get a replacement in but i swear i never even touched that thing

[CapLeaker]

I've got quite a few good stories about these 5 legged switches in that package. This is the worst designed switcher ever, Just make sure to clean, clean and clean that area again. It's got to be hospital grade clean. No stuck lint from a paper tower or swabs, flux etc.
Take that switcher off and wash that whole area with 99 % IPA a few times.
Good luck!

[r2digital]

I wish i was posting with a complete fix of the power supply but i am making progress. After the replacement of the switcher it now turns on and stays on between 1 and 8 attempts. Unfortunately when it does comes on all the 4 channels in the front light up detecting there's audio even when nothing is attached . However, i found that Q2 is the culprit of the random starting of the power supply. Does this seem like a suitable replacement https://www.mouser.com/ProductDetail...A%2FBgJw%3D%3D ?

[r2digital]

I have a working KDA1000 which i opened up to compare voltages to so maybe some one can point me in the right direction.


kda500 power supply not working

right to left

4 pin

brown 8.06
black 7.93
white -0.11
gray -0.10

5 pin

green 30.16
blue 3.29
green 2.58
red/white 3.01
red 1.11

3 pin

yellow -.759 <—— this should be -13-15VDC
light blue -0.024
yellow 13.78


3 pin

black 165

brown -166

red -165


Kda1000 operating normally

4 pin

brown 8.52
black 8.92
white -0.003
gray - 0.003

5 pin

green 32
blue 3.33
green 2.6
red/white 3.05
red 1.1

3 pin

yellow -13.7
light blue 0
yellow 13.8

3 pin

black 157.5
brown -157.8
red -157.8

55bz<—- need help identifying what this is as it is not giving me proper voltages

these are the measurements from the working power supply

1 -18.2
2 -18.2
3 -5.83
4 -18.71
5 -18.71
6 -18.71


The measurements from the non working power supply 55BZ chip were very low

1 0.380
2 0.380
3 -20
4 0.420
5 0.400
6 -20

[r2digital]

purple line shows where the trace starts at the yellow wire and ends at the negative trace of capacitor 35v 47U capacitor and should be -13-14VDC

Removed cap for better view

[CapLeaker]

Looks like the negative is having a through hole to the other side side of the board. Where does it go? It's got to be connected to something else somewhere? Do some more tracing.

[r2digital]

https://datasheet.octopart.com/FDC65...et-8322017.pdf

It has all the proper voltages near by except gets messed up at the 55BZ chip and i think the above pdf is a replacement for the 55BZ. There is nothing on the bottom of the board that is connected to this chip. Attached is a trace of its path for the 55BZ. C24 negative path is not connected to the nearby 55BZ. Pins 1 & 2 are connected to the same trace and pins 5&6 are also connected together

[r2digital]

Finally figured it out after taking the entire heat sink assembly out. R35 resistor was reading in mega ohms while it should've been a 10k ohm but at the same time while the heatsink wall out i replaced capacitor C15 since the only time you can get to it is with the wall out. All capacitors on the power supply and the main input output board have been replaced