Paradise Builders

[beetle]

Found it. Wrong cap. Will fix it tomorrow.

 

[Joachim Gerhard]

I can read brain waves C:

I fact it is always the same sh.....

 

[PierreQuiRoule]

I got pupu load of 327 (FEW UNDREDS) with HFE above 540 and few undreds 337 with with HFE around the 500 mark

PM on its way.

 

[FdW]

Found it. Wrong cap. Will fix it tomorrow.

That's easy done!

 

[analog_sa]

For some reason the SK170 in my output buffer fails to properly bias - gate stays positive and output offset fluctuates randomly within a hundred or so mV. Have not yet hooked a scope. Apart from that all other dc levels read fine.

Scratching my head...

 

[Joachim Gerhard]

What DC voltage do you measure between drain and source of the 2SK170 ?

 

[analog_sa]

1v in one channel and 2.3v in the other. Vgs is slightly positive in the first channel

 

[Joachim Gerhard]

Then the J310 cascode Fet does not provide enough voltage over the 2SK170.
When the Vds voltgae over the 2SK170 falls under a certain threshold there is not enough Vgs left to make the 2SK170 conduct negative.
Search for the highest voltage J310 you can find and use them as the cascode transistors over the 2SK170. This is simply bad luck.

 

[analog_sa]

Highest voltage?

 

[analog_sa]

I also experimented a bit without the servo or the output buffer. Quite serious random DC fluctuations at the transconductance riaa node. The servo manages to calm these down but because of the integration there is still up to a 100mV dc at output at times even with the servo on. Could this possibly be normal?

 

[analog_sa]

I am inclined to think the reason for the positive bias is a disbalance from the transconductance stage which the servo fails to fully correct. With the servo switched off there is a positive 1 - 3v appearing at the output of the transconductance stage. No idea why as my bipolars are carefully selected. Maybe i have made a mistake somewhere...

 

[pdul]

It sounds exactly like the problem i have too. the wandering offset on output, and riaa, but input and psu is ok. Finally someone else have the same problem

 

[pdul]

Could it be the fets playing tricks?

 

[pdul]

My problem is the same in both channels

 

[analog_sa]

Fixed the issue with the insufficient voltage across the output 2SK170. Indeed Joachim was correct - top FET needed replacement.

As for the randomly moving dc levels i still have no idea. And it is very consistent between channels.

 

[analog_sa]

Could it be the fets playing tricks?

Which fets would those be? The problem exists ahead of the output buffer... and is not related to the servo.

 

[pdul]

What else? The bipolars are well matched

 

[Joachim Gerhard]

I had initially an offset of around 25 to 40mV. That went down after a while. I assume that is forming of the electrolytics in the input stage ( 2 x 6800uF ). They have some leakage current that goes down with time in use.
If after several days in use the DC offset at the output does not go down there are instructions in the building manual how to raise the "authority" of the servo.
Yes, i ment searching for J310 that leave enough Vds ( Uds ) over the 2SK170.

 

[Joachim Gerhard]

There is also a little discussion on the other Paradise thread if the OP134 in the group buy are fakes. No conclusion yet but if that is the case then the servo opamps have to be substituted with a genuine one.

 

[pdul]

It is not the stable offset from the buffer, that is the problem, it is the wandering offset at the riaa note +/-100mv. Jumps around constantly as if it were sampling a bigger curve.