help philips cd 584 resets

[martinn]

got an old working philips cd 584 and did some modifications. After adding decoupling caps to the TDA1541, changing the output caps, fiddling with the opamp it all played fine, but when I replaced 33 uf and 47 uf caps with equivalent oscons the player resets itself at certain tracks and re-reads the disk like it was just inserted, otherwise it plays fine, but when it reaches a loud part of a track, well it resets.

I've checked the voltages on the new caps, reheated all the solder contacts (except smds), swapped power supply rectifiers for shottkys (1N4002->11DQ10)... nothing helped. I am thinking the player somehow runs out of juice, but the parts I've fiddled with seem fine. I am wondering what could cause the player to reset? Any help is much appreciated.

[PlasticIsGood]

Why loud parts I wonder?

My first inclination would be to reverse the last modification. Have you thrown the old caps away? In which positions did you put the oscons?

In some situations, an ESR much lower than in the original design might cause instability. How the control of the mech could be related to loud bits isn't obvious to me. Audio section power supply ringing infecting the digital ground maybe.

Do you have an oscilloscope and a circuit diagram?

[martinn]

Originally Posted by PlasticIsGood

Why loud parts I wonder?

I also wonder why. My reasoning is that the digital section should work regardless of the signal or not work at all. Since the player resets on tracks like hero's return form pink floyd which has a loud beginning I thought maybe there is not enough juice reaching the dac.

hero's return:
Pink Floyd - The Hero's Return (Parts I & II) (45 rpm single) - YouTube

People owning a Rega planet had similar problems with the player restarting. Replacing the rectifier diodes solved the problem there. This player has two rectifier bridges (schematic attached)

My first inclination would be to reverse the last modification. Have you thrown the old caps away? In which positions did you put the oscons?

I've reversed part of the caps reaching the transport. In the power section I've beefed things up - larger caps shouldn't hurt here? Also added caps around the dac.

The old caps are safely stored I'll revert first the caps around the dac and the filter chip.

I should mention when swapping the opamp I've placed it the wrong way in and burned the 15V voltage regulator (yellow sqare) on the power section. The player worked fine after replacing the voltage regulator with a new one. Then the oscons arrived and with them new problems.

There was a 33 uf oscon which had huge voltage drops when playing and I've replaced it with a 47 oscon. Thought this was it, the voltage was stable, but it made no difference.

In some situations, an ESR much lower than in the original design might cause instability. How the control of the mech could be related to loud bits isn't obvious to me. Audio section power supply ringing infecting the digital ground maybe.

Do you have an oscilloscope and a circuit diagram?

Unfortunately a multimeter is all I have. Also my knowledge about electrical devices is very limited. That's why I chose a simple, cheap cd-player with a tda1541 dac to learn on it.

Attached is a picture of the board:
- yellow square right 100 uf output caps replaced with fresh 100 uf nichicons.
- yellow left is the 15V voltage regulator smoked (and replaced) when I accidentally put the opamp wrong way in
- violet squares are oscons. I've exchanged 33 uf 16V for 33 uf 20V oscons and 47 uf 25V for 47 uf 16V oscons. (tda1541 from schematics needs 15V supply voltage).
This is where the reset error occured, added more oscons, measured all voltages. Where I could not measure I've replaced the white underlined caps with the originals and the 33 uf caps in the power section with the 100 uf bipolars from the audio section (which work). Replaced one 33 oscon with 47, because it was discharging rapidly while playing / under load.
- green marks the rectifier bridges and 3 additional diodes of the same kind (1N4002) exchanged for schottky's 11DQ10

On the bottom side of the board:
- added larger decoupling caps to the dac
- messed up the pcb on the audio output caps while trying to attach a too thick wire to the pcb to connect some large mundorf film caps, reverted to new nichicons
- found a crack on the pcb - where I've put some solder to fix the contact

[poynton]

Did you replace any of the 33uF/47uF caps in the area under the CD mechanism ??

If so, they must be the same type and value i.e. not oscons.


Andy

[Mooly]

Just zooming in on the third picture and the soldering looks a bit suspect tbh. Two components, one at left and one at the right look really dry.

[martinn]

Originally Posted by poynton

Did you replace any of the 33uF/47uF caps in the area under the CD mechanism ??

If so, they must be the same type and value i.e. not oscons.


Andy

no, the three 33 uf caps in that area were restored with the original ones and that did not make a difference

[martinn]

Originally Posted by Mooly

Just zooming in on the third picture and the soldering looks a bit suspect tbh. Two components, one at left and one at the right look really dry.

Thanks for all the replies! Added some solder there and replaced the 47 uf 16 V oscon near the dac with the original 47 uf 25 V nichicon. Now the player plays the track that causes the reset for 13 s... which is an improvement The voltage measured is 14,89 V so I don't see why the 47 uf oscon would make a difference, but somehow it did (maybe the ESR issue mentioned earlier?).

Now I'm replacing the 47 uf oscons in the power supply area... fingers crossed

[martinn]

Now all the oscons are reverted back with the original caps, checked the schematics if I made a mistake and mixed up some caps, but all looks fine (was not so hard since there were only three 47 uf caps replaced, the rest was 33 uf).

Rechecked all the contacts, added some solder here and there and cleaned with contact spray (teslanol). The player makes it a whole 16 s into track 5 and then resets.
took out the opamp in the output stage.. no change

well the board looks much less colorful without the oscons

[PlasticIsGood]

Chin up! I'd be happy with a player that refused to play Pink Floyd.

Your headphone amp's the wrong way round. 4560D, top right.

I would remove that blob of solder over the crack in the track, and use a wire link. The way you've done it, it could be that the solder has only taken to the bare edge of the crack, and a little flexing could break the joint. Clean back a section of track to bare copper before you solder...a fibreglass pencil is good but fine abrasive paper will do. The board tends to flex due to the weight of the transformer. They can crack from the back of the audio socket to the front of the heatsink too. The tracks are flimsy, as you've found.

Carefully remove all the solder splash (from using a solder sucker?). I agree with Mooly that some joints look dry. Is your iron hot enough?

Are you sure that track played before? The single-board CDM4 machines can be a bit choosy about CDRs, I find.

[martinn]

Originally Posted by PlasticIsGood

Chin up! I'd be happy with a player that refused to play Pink Floyd.

Your headphone amp's the wrong way round. 4560D, top right.

I would remove that blob of solder over the crack in the track, and use a wire link. The way you've done it, it could be that the solder has only taken to the bare edge of the crack, and a little flexing could break the joint. Clean back a section of track to bare copper before you solder...a fibreglass pencil is good but fine abrasive paper will do. The board tends to flex due to the weight of the transformer. They can crack from the back of the audio socket to the front of the heatsink too. The tracks are flimsy, as you've found.

Carefully remove all the solder splash (from using a solder sucker?). I agree with Mooly that some joints look dry. Is your iron hot enough?

Are you sure that track played before? The single-board CDM4 machines can be a bit choosy about CDRs, I find.

Ok, no more pink floyd

I've played mike oldfield now, and it played the whole cd flawlessly. Then I've placed back the cover and it reset at track 4, which is kind of loud. If I let it cool down a few minutes and remove the cover it plays the normally again. I guess it's a heat issue.

Thanks for the advice, you are absolutely right about the headphone opamp. Things are starting to make sense now. Some time ago I tried replacing the headphone opamp with NJM5532 or opa2134 and the 15 V regulator (yellow square) started burning. Then I've replaced the burned 78L15 ACP 826 with L78L15ACZ TO-92 for 100mA.

-> and it gets quite hot. I'll try replacing it with a bigger 2A 15V regulator with a heatsink (have to order one). It would make sense that when it overheats the new voltage regulator has heat protection and switches off for a short period... and the cd player resets.

Hm, my soldering temperature is 350-400 C. New solder looks more smoth on the surface and the old solder is whiter or pale, sometimes I've added some new solder to old contacts, then it's a mix. I also get some dark stuff mixed in, probably carbon I rub the solder iron against to get rid of excess solder. Is this what you meant by dry joints?

Thanks for all the help!