Quad 303 - High End PCB Upgrade Calibration issue

  • No one logged in.

30-Dec-2017 05:57 AM

BPD

BPD

Posts: 16

Hello all

I have just upgraded my 303 with a set of the Dada High End Upgrade PCBs.

It recently had a problem with the Right channel blowing with the original PCB's ( previously upgraded with standard Dada kit)

I replaced all the power transistors on the heat sink ( with 15003 x 4 and x 1 MJ802G) - New mica pads and conductive paste. ? Does it matter if the conductive paste touches another mica sleeve on external heatsink.

I have been able to set the 67 v and the 33.5 v no problems.

When I now connect to pins 4 and 6 to to measure and set the 6-10mV and attempt to adjust RV101 the voltage reading between pins 4 and 6 jumps to 400-600mV . The Heat sink where the power transistors sits gets very hot - I am assuming lots of current flowing !!

This behaviour happened with the old original Quad PCB's.

The commonality being the power transistors now with the high end PCB's. I have checked these transistors individually recently  ( with leads off and all test ok). Is there a trick to mounting these ??

Do I have a dud set of transistors for the L and R channel ??

When the voltage on pins 4 and 6 is within specs external heatsink is warm BUT NOT hot.

Appreciate any advice


  Reply

30-Dec-2017 08:24 AM

BPD

BPD

Posts: 16

OK further Dada website reading !

It seems if you adjust RV101you are setting the base to emitter voltage on TR107.

I find ( in my 303 ) if I adjust this voltage much above 6.5mV then "I get runaway" and I read about 600mV - alot of current here and a very HOT heat sink.

When you reach the critical point , even touching the pot will give yiou runaway.

I have fount the best way to set it for "stability of a number" is to set the value and turn on and off the AC power to ensure that the voltage across pins 4 and 6 is repeatable  ( within your set value).

It would be great if someone more knowledgable to me could expand on this bias setting and why  ( and  what) is this value.

The other question as I have the High End PCBs now is there a different voltage or setting across pins 4 and 6 - Stefaan / Joost comments please .

Ahhh the learning curve.

Bruce 

Sydney Oz.




  Reply

30-Dec-2017 09:55 AM

Joost Plugge

Joost Plugge

Posts: 962

The HE boards have the same setting as the normal boards, about 10 to 12mV DC across pin 4 and 6, no load, no input signal.

The heat conductive paste is an electric isolator, so no problem if you use to much paste.

Before you wire up new transistors, always check if there is no contact with the heatsink.

You could swap the boards, to see if the fault is in the board or the output side.

The part about the thermal runaway is correct, but should not occur with the setting of 10 a 12 mV.

Joost

  Reply

31-Dec-2017 03:58 AM

BPD

BPD

Posts: 16

Thanks Joost

I would like to actually measure the quiscent collector current whilst adjusting RV 101.

In the Quad Service Data on page 2, under Regulator Board section - Point 4.

"Adjust RV101 for 5 to 10mA quiescent  collector current measured by breaking the eternal lead to tag 2 of the left hand driver board and inserting a meter in series at this point, with no signal drive."


Can yo actually point out this position on the schematic so I can check this current.


Cheers

Bruce

  Reply

31-Dec-2017 04:42 AM

EJP

EJP

Posts: 1506

It's a lot easier to measure the quiescent voltage, by measuring across pins 4 and 6 (or, equivalently, the farther ends of the current-limiting diodes MR105/6) and dividing by 0.6 according to Ohm's Law (0.6 = sum of the emitter resistors you are measuring across). I'm not keen on measuring current directly at any time by any means (except a current probe, which I don't possess), and particularly not by breaking into connections that then have to be repaired and re-tested.

If you can't get a stable bias around 9-12mV try it at 6mV and then increase it to just below the point where it runs away. Sometimes, for reasons I do not understand, a lower bias, say 7-8mV, is required to prevent thermal runaway. This seems to be more the case in the earlier model with the four diodes around this spot, but I've seen it also in the later model with Tr107.
  Reply