• State Resolved
  • Locked Locked
  • Replies 7 replies
  • Answers 1 answer
  • Subscribers 30 subscribers
  • Views 993 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

Original question:

PCM1792A: Resistor values for OPA134 family for minimal distortion.

PCM1792A: Distortion

Mark Willerton
Prodigy 140 points
Part Number: PCM1792A
Other Parts Discussed in Thread: PCM1792, , TINA-TI, OPA1678, OPA1611, OPA134

Hello again!

I was looking at the datasheet for the PCM1792 again (fig.37) and had another idea that might reduce distortion at higher audio frequencies (i.e. >10kHz).

As before, I am thinking that the op-amps should use around 2k2 as feedback resistors instead of below 1k. This would give an optimal balance between noise and distortion in the OPAx134. The problem is the first op-amp that converts the output current of the PCM1792 to voltage (U1 and U2). It would be difficult to use a resistor with much more than the proposed 820R due to the centre current (-6,2mA).

One way to solve the matter would be to connect a resistor between the 5V (quiet) VCC supply line and the output of the PCM1792. Setting the resistor close to 806R would effectively reduce the output centre current from -6,2mA to zero. Then I can use whatever feedback resistor I want in (U1 and U2).

Is it ok to do this or would it introduce other problems?

Thanks
Mark

  • 0
    TI__Guru 54983 points
    Hi Mark,

    I'll try this in a simulation tomorrow and see how this looks! Though I think the resistor would need to go to a negative potential to offset the current, and the PCM is a current source.

    I'll get back to you tomorrow.

    Thanks,
    Paul
  • 0 in reply to Paul_Frost
    Prodigy 140 points
    OK, thanks! I was initially thinking the same regarding the negative potential but double checked and it says in the PCM1792A datasheet that the centre current is -6,2mA (I was assuming this means 6,2mA goes into the DAC).

    Best regards
    Mark
  • 0 in reply to Mark Willerton
    TI__Guru 54983 points
    The negative/positive sign can be confusing, but know that this device is definitely sourcing the current. The IOUT pin is usually at a virtual ground potential created by the amplifier, and the current flows to the lower potential.
  • 0 in reply to Paul_Frost
    TI__Guru 54983 points

    Hi Mark,

    I have been playing with this, and it is interesting as it lets you use a bipolar supply, instead of negatively biased  supply.  In order to achieve the very low noise needed to realize the 130ish SNR, some fun tricks can be used with the gain of circuit.  For example, having high gain in the first stage and attenuation in the second stage.  The problem with having a large gain in the first stage is that the offset becomes problematic, as the output of the first stage is negative.

    By having the offset removed by having a resistive to a lower potential you can use a higher gain on the first page without increasing the supply.  This does present some problems, as the new resistive path creates an the input resistor of a inverting amplifier! So the noise is gained by the new -Ri/Rf gain.  My instinct is to make that resistor small and use a small negative bias so that there is less thermal noise from the resistor, but that makes the noise gain higher!

    I have attached a simulation where I use a -4V bias with a 645ohm resistor to center the output of the first stage around 0.  This combination has about the same noise as the version without it.  This does assume the -4V reference voltage has no noise, which will not be the case either.

    7635.PCM1794 with neg offset.TSC

    Thanks,

    Paul

  • 0 in reply to Paul_Frost
    Prodigy 140 points
    Hi Paul

    Thanks (a lot) for the feedback. Really interesting - can't wait to try this out!

    I am a bit shy to admit having trouble opening the .TSC file. Is it possible to send the file in a format that can be opened by say Altium, PS-Spice or similar?

    To the original point: Addition of the offset resistor would allow higher resistance (e.g. 2k2) to be used in the subsequent op-amp stages. Then there are the possible benefits regarding noise from having higher gain earlier in the signal chain as you say. Do you thing that careful selection of the 'offset' resistor and a well selected (very low noise) voltage reference would support this strategy?

    Best regards
    Mark
  • 0 in reply to Mark Willerton
    TI__Guru 54983 points
    Hi Mark,

    Try to download TINA-TI from ti.com: http://www.ti.com/tool/TINA-TI

    Whether or not this is a good strategy is a hard question to answer. I would recommend that first you figure out the SNR/THD goals of the project. The bipolar/JFET input of the OPA134 means that higher resistor values would result in higher noise. For example, I simulated that transimpedance amplifier generates less noise with the 2.2kohm resistor if it is a CMOS type, like the OPA1678. Meanwhile the second stage can still have relatively small resistors, so should still be a JFET/bipolar type, like the OPA1611 or OPA134.

    If noise is not the primary concern and you care more about distortion, then this might make more sense, but that is much harder to simulate. That may require real testing! Are you thinking about building a prototype of this design?

    Thanks,
    Paul
  • 0 in reply to Paul_Frost
    Prodigy 140 points
    OK, message received! Thanks for the help. Yes, it is a prototype but at its early stages. I will try things out and let you know how it goes!

    //Mark