advertisement


Op amp question

I

IDM

pfm Member
I am using an AD844 for I/v conversion with a TDA1541a using a circuit originally put forward by Pedja Rogic. When I adjust the dc offset on the input to 0v I can measure about 400mv offset on the output (from pin 5). Could this high offset come from slightly unsymmetrical voltages + and - on the supply to the AD844? If so should/could I trim the regulator voltages to get the output to near 0v? I do understand that the offset may drift, I just want to minimise it.
Thanks
Ian
 
I once had a highly modified Phillips CD80, one of their 1541 flagship models. The IV conversion there was done by a ladder of Vishay resistors. Even back then it was an expensive way, but sonically very good. These days I am sure there are more capable OP amps that may do an even betterjob, but back then 5532 ruled.
 
IDM said:
I am using an AD844 for I/v conversion with a TDA1541a using a circuit originally put forward by Pedja Rogic. When I adjust the dc offset on the input to 0v I can measure about 400mv offset on the output (from pin 5). Could this high offset come from slightly unsymmetrical voltages + and - on the supply to the AD844? If so should/could I trim the regulator voltages to get the output to near 0v? I do understand that the offset may drift, I just want to minimise it.
Thanks
Ian
Click to expand...

Is the design intended for and optimised for the AD844 (which is a current feedback op amp)? Output is pin 6 BTW. You can't just fit one of these where a normal op amp should be used!
 
Hi Arkless

I know pin 6 is the usual output. However, the Rogic design uses pin5 with a resistor to earth (as I understand it the resistor is to convert current to voltage. The output is then take to a separate buffer amplifier to output. Pin 5 is described as to from memory.

Also as I under the AD844 is highly unusual in that it can do this trick.

So can I minimise offset by rebalancing the supply rails?
 
IDM said:
Hi Arkless

I know pin 6 is the usual output. However, the Rogic design uses pin5 with a resistor to earth (as I understand it the resistor is to convert current to voltage. The output is then take to a separate buffer amplifier to output. Pin 5 is described as to from memory.

Also as I under the AD844 is highly unusual in that it can do this trick.

So can I minimise offset by rebalancing the supply rails?
Click to expand...

Unlikely. You could try the offset null facility. See datasheet. The pin 5 turns out to be a high impedance output, a transconductance output. Also check for oscillations as it is a very wideband device.
 
Oscillations, that’s a good thought, I had not considered that. If it is oscillation what would be the best way to control it? Ferrite beads?
 
The Pedja Rogic design deliberately uses the transconductance output from pin 5 of the AD844, to mirror the current into an output impedance for i/v

(It's easier for most than to build a common-base stage to do the same)

No , you can't swap opamps, and yes, if nullled to 0.00v at the input (as it should be - the TDA1541A requires no more than 25mv compliance voltage from 0v for least distortion - i.e. a very-low-impedance load) then there may well be a voltage offset at the output. Which means use a coupling capacitor at the output of the analogue stage.

And no, fiddling with voltage rails will not alter this output offset - you are seeing an artifact of the way the internal circuitry works.
 
Hi Martin,
Thanks for the really helpful post. When I adjust for zero offset on the output of the Ad844 (pin5) I end up with approx 20mv on the input. From your post this sounds within the acceptable limit, is that the correct interpretation?

I also found in one of the Rogic papers that you should adjust for zero offset on the output of the ad844, he doesn’t mention checking the input for being within 25mv.
Cheers
Ian
 
You must log in or register to reply here.


advertisement


Back
Top