ADM 1540 EQ, adjusting DC offset into output transformer...

[tonedude]

I just fired up a pair of 1540 eq's I recently bought, and they seems to work just fine, great! I then recapped them with a definite improvement.

I then thought I should check the amount of DC offset running into the output transformer and I measured that one gets - 2,1 mV and the other gets 0,7mV. Under 1,0mV seems to be regarded as just fine to send into a output transformer, without using a coupling capacitor blocking dc. So what to do with the one getting - 2,1mV?

Questions;

- Is -2,1mV fine into this transformer (original ADM output trannies) or should I try to adjust the circuit to get below 1,0mV?

- Does anybody know what values to use if I want to implement the circuit in the 1540 schematic to adjust for high DC offset and get it closer to zero (components R81 to R84 in the schematic, those are not specified anywhere in the schem as far as I can see)?

 

[tonedude]

Part of the schematic...

 

[RuudNL]

Personally, I think this is a rather tricky construction, since there is never any DC feedback to the input of the OpAmp.
(The output is almost grounded for DC through the transformer winding.)
One or two millivolt won't hurt the transformer, but I think I would add a (large) capacitor, to avoid any problems.

 

[JohnRoberts]

Personally, I think this is a rather tricky construction, since there is never any DC feedback to the input of the OpAmp.
(The output is almost grounded for DC through the transformer winding.)
One or two millivolt won't hurt the transformer, but I think I would add a (large) capacitor, to avoid any problems.

Yes there is DC feedback from the very output (after the discrete transistor buffers) to the opamp input. That connection eliminates and DC error from the buffer transistors. 

JR

 

[RuudNL]

Yes there is DC feedback from the very output (after the discrete transistor buffers) to the opamp input. That connection eliminates and DC error from the buffer transistors. 

True. But the output is 'shorted' by the low DC resistance of the transformer winding.
If there would be 1 V. offset at the output, this offset would never reach the input of the OpAmp!

 

[JohnRoberts]

Yes there is DC feedback from the very output (after the discrete transistor buffers) to the opamp input. That connection eliminates and DC error from the buffer transistors. 

True. But the output is 'shorted' by the low DC resistance of the transformer winding.
If there would be 1 V. offset at the output, this offset would never reach the input of the OpAmp!

The transformer has a low, not zero impedance. The buffer/driver has a nominal source impedance of 4.7 ohms, so it can drive perhaps amps into that load. At DC the opamp has open loop gain of 100k or more, so it could easily drive the current required to stabilize with mV of DC offset.

====
To answer the OP's question there is a trim shown on the schematic with dotted lines, but it may just be easier to drop in a more modern bifet opamp with better DC specifications.

JR

 

[tonedude]

To answer the OP's question there is a trim shown on the schematic with dotted lines, but it may just be easier to drop in a more modern bifet opamp with better DC specifications.

JR

Do you have any clue what values to use for the trimpot and the three resistors? (the dotted lines indicates that that part is missing on the circuitboard, it's an optional fix ant there is no info on what values to use... )

 

[RuudNL]

I would use a high value for R81, otherwise the adjustment would become very critical.
Maybe even 1 M.ohm or so. For the pot anything from 10 K to 250 K (or so) would be usable IMHO.

 

[tonedude]

I would use a high value for R81, otherwise the adjustment would become very critical.
Maybe even 1 M.ohm or so. For the pot anything from 10 K to 250 K (or so) would be usable IMHO.

Ok, thanx! So R84 and R83 is't really needed in there?

 

[JohnRoberts]

Well it is pretty straightforward to predict effect of R value at the node. The feedback R is 7.5k, so a 7.5k trim R would vary the output +/- the full PS rail voltage. 75k= Vrail/10  , 750k= Vrail/100.  !Meg seems like it would deliver well more than the reported 2mV of offset. The shunt resistors across the pot will slow down the trim sensitivity near the middle, while allowing it full range of trim at end points.

JR

 

[tonedude]

Well it is pretty straightforward to predict effect of R value at the node. The feedback R is 7.5k, so a 7.5k trim R would vary the output +/- the full PS rail voltage. 75k= Vrail/10  , 750k= Vrail/100.  !Meg seems like it would deliver well more than the reported 2mV of offset. The shunt resistors across the pot will slow down the trim sensitivity near the middle, while allowing it full range of trim at end points.

JR

Thank you JR, much appreciated!