Stax Question

[lpd]

I have a friend rebuilding a channel on a Stax Da-80. One channel was blow so replacement transistors, on output and driver board were replaced. On start up, it is makes noise ect in repaired channel then stabilizes and operates as it should. Any ideas? I can indicate what transistors were sub into the amp later tonight.

Here are some notes so far.


If I swing the DC balance control all the way clockwise and turn it on,
it's OK, but then dives negative rather quick. But not so quick I can't
follow it to zero with a screwdriver on the DC balance control to the
normal op point.

This only happens on an ice-cold start. Once she's warm, there is no
problem. Even preheating it with a heatgun is enough to keep her from
going wild on power on.

I didn't scope it because the rise/fall time is enough to see it
manually on the analog FETVOM.

[stratus46]

If you have a scope you _should_ scope it as that will tell you far more than a meter. So you have a temperature sensitive component but it could be anything as resistors, capacitors and semiconductors have all been known to do this. It could even be a flaky solder connection.

Good luck with this. You have one of the 'tough dogs' but it will be very satisfying when you solve it.

G²

[h_a]

noise ect in repaired channel then stabilizes

Could be a cold solder joint, you can try to reheat the solder joints of the power relevant parts.

This only happens on an ice-cold start.

Sounds to be related to temperature compensation being not sufficiently well in contact to the power transistors; I would recommend to have a look at that

Hannes

[Geek]

Hi,

Originally Posted by h_a

Sounds to be related to temperature compensation being not sufficiently well in contact to the power transistors; I would recommend to have a look at that

I'm the tech

The tempcomp however is working within the manual spec - 400mV bias across an output emitter degeneration resistor, settling to 350mV after 30 mins. warm-up. If it wasn't working, them bipolars would fly into thermal runaway right quick like.


Cheers!

[h_a]

settling to 350mV after 30 mins. warm-up

I meant that thermal comp might not be tightly connected Or cold solder joint.

Both might lead to functioning nicely when warmed up, but not immediately after turn on.

Hannes

[Geek]

/* snooty accent
By jove, I think I've got it!
snooty accent */

So tonight I rip 'er all apart again and for safety pull the output transistor leads (which just release the complimentary current boost, everything else stays put), hookup the meter to the output terminals and kick it on... it was fine.

Very confused, I decided to hit each transistor on the driver board with a little heat to see which makes it go flying to the rail - perfectly stable

I began tapping components looking for bad solder joints... meter didn't budge.

The driver board fits into a bakelite card-edge connector to the mother board. So I decided to screw in the mount screw and do more testing..

*CLICK*

The protector circuit kicked in and the meter went to V+

Hmmm.....

I wiggle the board and the meter zero'd and the protection clicked back out.

I gave the board a little press where the DC balance control was...

*CLICK*

Bottom line .... the connector it fits into is b0rk3d (I redid the solder joints on it days ago)

What I found was the piece of connector that was damaged, has the pin for the NFB loop, hence error correction. When I adjusted the DC balance, it wasn't really that, that caused the cure... it was the pressing on that edge of the card.

Why this is only a cold issue is perhaps the pins/sockets expand just enough to make contact? I dunno, but it looks like I'm going to have to find a connector that will fit.

Thanks all for your help, if anything else creeps up, LPD and I know where to come

Cheers!

[Duo]

If you can show me some clear pictures of the connector I may be able to dredge up a replacement from my junk pile.

[Geek]

Hey Duo,

Than ye muchly!!!!

I'll have them tonight (off to town now)

Cheers!

[jam]

I remember on a couple of these years ago and in both cases it turned out to be defective input jfets............well thats my bet anyway.

Jam

[Geek]

I hope not. Almost every part on this thing is unobtanium and needs a cross-reference. Hopefully this connector does it for good.

Duo - pics sent

Cheers!