Troublesome Switch mode PSU

This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.

I have got this car amplifier.... a JVC KS-AX504. A 4 channel poweramplifier...

When I try to turn it on, the "power on" LED just blinks on and off with ca. 2 seconds intervals...

As most car amplifiers, it has a switch-mode PSU thingy... Measurments shows that the B+ voltages is in the neightbourhood of +-18V. The negeative 18 volts is quite stable, but the + rail keeps dropping down to ~2V (from 18V) when the LED goes off. When the LED turns on, it swings back up to 18V for a sec. or two...

I am just wondering if anyone have got any clue, or maybe have encountered related problems in the past....

Some advice...

The first thing that I would do is to check the obvious. Start by carefully removing all of the devices from the output stage and test them. Be absolutely sure to mark each device you remove in such a manner that you can put them back into the same place! You don't want to upset and biasing without a service manual to put it back! I would also check the diodes and any filter chokes in the PSU stage. Usually the diodes are half bridge and look like a TO-220 style device. I have seen these fail, as well as the coil of a filter choke failing. I'm not sure about JVC amps, but +/- 18v seems really low. Most amps that I have worked on use between +/- 25 to +/-35v. It is also possible that a mosfet has failed in the PSU, but since you say the negative voltage is stable, this makes no sense. Let me know what you've already tried, maybe this will help eliminate some possible failures. I have worked on nearly every major brand of car amplifier, and most all of them work the same, however, everyone does something a little different from everyone else. I've never worked on a JVC.

thanks for replying!

This is what I've done:

- OK'ed all the output transistor.
- OK'ed all the PSU mosfets (IRFZ 44's).
- All diodes, including the TO-220 double diodes (two of them)
worked ok.
- Checked 4 Zeners, 1N4744 (15V, 1W types), they where not
the problem.
- Got out the "transformer" thingy, a large ferrite ring with some
few turns on it with thick copperwire. No signs of shortcuts
- Tried to change the PSU caps on the + rail with new ones...
- Disconnected two of the channels from the PSU, because I only
had to take some jumpers out. The other two channels had
PCB wiring from the PSU, and I didn't want to cut off the tracks.
The amplifier behaved like before...
- I even checked the Heat dependant resistor thing that is
connected to the heatsinks. I tried to power the amp up
without it, resulting in a completly dead amplifier.
- I did find some cracked solder-joints, so I went through most
solder-joints with fresh solder, but no results.

This amplifer has not been misused, that is, no components / PCB tracks / wires shows any signs of exessive heat. As far as i know it has not been used in a while.

- stig
check to see which mosfet driver is being used for the power supply. Is should be a UC..... or an SG......... It is the PWM driver for the power supply. It is possible that it is going through a soft-start cycle because of a shorted load. This would probably tell you that some of the output devices are shorted in the output stage. Look for the chip and cross reference its data sheet.

No big deal

It's not necessary to remove the output devices to check them. I repair on average 3 car amplifiers every day...of all makes and models. Use the Diode-Check function on your DMM. You'd be surprised how many techs don't do this.

99% of the time, protection circuits that activate after the supply has powered up indicate either a shorted output (B-E or C-E), open emitter resistor(s), damaged RCA input ground connections or shorted OpAmps. Yes, shorted to rail voltage sends +/- 12-16 volts DC straight into the amplifiers. In the rare event that a driver transistor has failed apart from an output shorting, you will get a high DC offset right before protection activates.

As the amplifier powers up, check the speaker output voltage (DC). If you get more than a few volts of "thump" DC, you've got troubles.

On the REAL dogs, I recommend powering up the audio sections separately using a dual voltage, current limited supply. I use a Power Designs unit I picked up on eBay for less than $75. Pull the rectifiers out and solder a wire to the remaining center pin connection. Observe polarity! Easiest to find tracing back from the filter caps... Ground at the nearest point - usually the center tap of the toroid or a speaker negative terminal.

Doing this saves literally hours of troubleshooting time and makes everything easier since you can limit total current draw as much as you want. You can extend your trace time from microseconds - or none at all - to many minutes...even without clamping to the heatsink.
Okey, I have done some more testing...

I had no separate PSU with + & - output, so I used two 12V batteries in series. Put two 220R resistors in series with the + and - rails to limit the current. Then I pulled out the TO220 double diodes, obviously the rectifiers.

The amp behaved like this;

- rail at -11,9, almost no current.
+ rail at 8,6 volts, pulling about 10,5 mA through the 220R resistor.

No parts got warm, but I did'nt have the batteries connected for a very long time...

The PSU Mosfet driver is a TL494. Can't see any shorts anywhere around it...

Found the error!!!

It was the driver transistor to one of the channels that had shorted between Collector / Emitter. Used the Diode-check on my DMM, and when I pulled the transistor out, the amp worked!!! Have +-19,7V on the rails now.

Now I only have to find a substitute. It is marked C1027. The complementary is C1023. The C1027 is a NPN, and I pulled out a C1023 and measured the HFE to be 98. It is probably best to change both of the driver transistors, yes?

Anyone knows of any substitutes for these transistors? I cant find them in any catalogs...


I was in the midst of posting a return comment when your post came up...was going to discuss the importance of driver transistors - but you already found it! Good Job!

To tell the truth, JVC and other car audio companies bias their amplifiers so low that just about anything "Tall TO-92" NPN within the power rating will work fine. It's sad, but true...but profitable. :)

Try C1627, A1266. NTE289A complimentary to NTE290A. I found these in a Kenwood amplifier which probably share many of the same components.
Thanks for all the help!!

think I have actually learned something here :cool:

I read your post, and measuerd on the outputs, one of the channels had a DC of ~3,5V, so then I knew which channel I had to work with. Then I just switched to Diode-check and it was the first transistor I tested, and now I am proud as hell! :D

I have a box full of various transistors, probably at least one of the types you mentioned, so I'll just throw one of them in and observe the result.

Thanks for all the help!!

Stig Larsen
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.