QRV08 output voltage offset and the servo

[Kingston]

Hi,

I recently built the QRV08 amp and it worked really well from the first power on. I built it without the output servo first. I was able to tune output DC offset to 0mV with just the input trim. Perfect performance!

I decided to test the servo as well. I am using LME49990, by the way. But the servo actually biases the outputs and little bit off. One channel is about 20mV, the other about 30mV! I know the servos are working fine since input trim no longer has any effect on output. But the outputs are biased steady at 20 and 30mV instead of zero.

How can I set the outputs to perfect 0mV with the servos? Is the 20-30mV bias at the output actually a problem?

Is there any point using the servos since I get zero offset without them?

Thanks,
Mikko

[peranders]

The input bias currents are much higher in the LME49990 than a JFET based one. Could be this.

[Kingston]

That makes sense. I will test with a FET input opamp and report back, thanks.

But further, is there actually any point using the servo if I'm able to trim the output voltages to perfection zero?

Mikko

[peranders]

It depends... but if you feel that you have a stable output you can omit the servo. In case you have the input DC connected to your signal source you may need the servo.

[Kingston]

It depends... but if you feel that you have a stable output you can omit the servo. In case you have the input DC connected to your signal source you may need the servo.

That's good advice. I won't be using the input cap so the servo stays just in case.

Also, I just soldered in OPA132 and the output is a perfect zero DC now. Input bias current difference between 5 pico amps and 20 nano amps made all the difference. Large difference between bipolar and FET! Time to close the lid and set this unit to active use. Thank you very much.

Mikko

PS. If there is one critique that must be said about this project is that the trimmer should have been a multiturn one. It's jumpy and difficult to set input to zero with a single turn dial.

[peranders]

The idea was to let the servo do the fine tuning.

I don't mind a picture later on.

[Kingston]

I still have to finish the case. This headphone amp is a part of a monitor controller 1U rack unit in a studio. There is balanced passive XLR input from mixer main outs (4-deck balanced rotary switch attenuator) and a separate unbalanced headphone input (2-deck rotary 100k attenuator). Three choices of input to headphone amp: XLR main input or RCA unbalanced input (Lundahl LL1579 input transformers) or plain unbalanced (no transformers). I will post a picture when its done.

[Kingston]

This is the end result.




Left side knob is for passive XLR pass-through. Right side is the 100k headphone input attenuator. Front panel switches select input to headphone amp (balanced with transformer, unbalanced with transformer, and plain unbalanced). Left & right have separate switches because I could not find a 2-deck 4-pole switch on short notice. There's some minor tweaking left to match all input levels to the headphone amp.

[peranders]

 :-D Nice!