DRV134's single end to balanced noise analysis?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I simulated DRV134's noise analysis, on datasheet, DRV134 is pretty noisy (80nV/hz 1khz), but I'm confused with the result, should I measure the noise between the Vout+ Vout- or Vout+ to Gnd (VM1 on my chart).

Noise between Vout+ to Vout- is only 11uV (VM1),
but noise on Vout+ or Vout- is staggering 326uV.

on another hand, if I use opa1632 to convert signal,
Vout+ is only 4.4uV , VM1 is only 2.5uV...

So should I measure noise between Vout+ to Vout-,
or only Vout+ or Vout- if I want to use DRV134.

Thank you for your help.
 

Attachments

  • Drv134 and opa1632 .jpg
    Drv134 and opa1632 .jpg
    104.2 KB · Views: 244
  • DRV134.jpg
    DRV134.jpg
    102.8 KB · Views: 230
  • Opa1632 noise.jpg
    Opa1632 noise.jpg
    46.1 KB · Views: 224
These two circuits are fundamentally different. If you short one of the terminals of the OPA1632 to ground, nothing changes on the other terminal. It is a differential amplifier, but it's not the cross coupled ground compensating circuit used in the DRV134.

The ground compensating circuit is tremendously noisy, no matter whether you make it from scratch or use one of the pre-built chips. This is one of the reasons why I don't favor it's use - entire, complex devices can be 10-20dB quieter than just this output amplifier. 11µV is pretty darn noisy for a line level low gain driver amp.

To answer your question, I'd measure the noise between the outputs. But, for the DRV134, I'd also test it into an unbalanced load by grounding one side and then measuring the other side - that's the whole point of using the ground compensated amp design. You need to verify stability as well as noise etc.
 
The ground compensating circuit is tremendously noisy, no matter whether you make it from scratch or use one of the pre-built chips. This is one of the reasons why I don't favor it's use

Your answer is very detailed, I thought those 2 circuit is similar stuff, and didn't realize it's fundamentally different. I ran some simulation according to your suggestion, I think OPA1632 might be more suitable for my use. I'm looking for a way to convert single ended signal to balanced signal, then maybe bridge power amp chip or op amp + current stage.

OPA1632 has better THD, noise, phase performance, but DRV134 have lower DC offset. and the high noise which will degrade SNR 15db on 20khz 1V, any gain behind DRV134 will make the sound miserable.

The downside of OPA1632 for this task is it's DC offset, does this number shift with signal amplitude? Can I put a DC servo connect to Vocm, but I'm kind of afraid to inject noise into the Vocm pin.

Or maybe there're better circuit for single to balanced converter .
 
Account Closed
Joined 2010
5 years ago i built a drv134 headphones amp as in this link, except i used darlington transitors on the output and modified the superdiode acordingly:
Pro Audio Design Forum • View topic - Using the THAT1646 As A Transistor Pre-Driver/Headphone Amp

but i don't remember to hear any noise once the amp was finished, not even in a pair of 600 ohms AKG headphones.i used some voltage regulators for +-12v .I also used the smd version as i destroyed a Matrox proffessional video-audio sound card which had them on it.At the time when that Matrox was launched was about 4000 dollars...I doubt that they would have used annoing noisy devices next to the most expensive akm dacs and adc's of the time.
Apparently today i was looking for that headphones amp and i can't find it , but i think that was one of my best amp for headphones.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.