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.

PGA2320: Minimum input impedance

Part Number: PGA2320

Hi, 

 

 Could you kindly clarify the MINIMUM input impedance (worse case attenuator setting) for the PGA23** series volume control chips?

I thought I remembered reading it was approx. 8K ohms minimum but cannot find that figure mentioned anymore...?

I need to know for calculating preceeding capacitor value, for minimal low bass drop.

Very best regards,

Simon

  • Hi Simon,

    Thanks for your post. The datasheet recommends a source impedance of 600 Ω or less. Source impedances up to 2 kΩ cause degradation.

    -Tamara

  • Hi Tamara, 

     I'm afraid that does not answer my question.

    I need to know the minimum input impedance of the PGA23xx.

    Not maximum output impedance of previous stage,

    Please advise further.

    Very best regards,

    Simon

  • Simon Quarry1 said:

     I'm afraid that does not answer my question.

    I need to know the minimum input impedance of the PGA23xx.

    RTFDS: for PGA2320, see Page 3, "Electrical Characteristics," under parameter "DC Characteristics," it's right there: input resistance is 12 kΩ, input capacitance is 18 pF.

    Definitely drive the PGA2320 input from an op-amp output.

  • Yes. That  is the static resistance

    Also this input impedance was said to change with differing level set on the attenuator.

    Hence my request for clarification.

    RTOFP!!

  • Hi Simon, 

    What does "RTOFP" mean?

    Looking into the input of the PGA2320, you will always see 12k-ohm (typical), regardless of the attenuation/gain chosen. Current will flow through the resistors and not into the MUX:

  • Hi Tamara,

     "RTOFP" was directed at the other chap's "RTFDS", not you.

    I was convinced I'd read somewhere that there was a slight but measureable change, to as low as 8K at certain settings.

    I can no longer find this source. I will allow for lowest 8K,  to be sure.

    Very best regards,

    Simon

  • Hi Simon,

    Understood. It is important to note, though, that this input resistance is a "typical" value. It may vary up to 15% depending on process variation.

    Good luck with your design!

  • Simon -- the CS3310 device (and its brothers) from Cirrus specs a minimum input resistance of 8 kΩ with a typical of 10 kΩ. That implies that the tolerance on the input resistance is +/- 25%, worse than a mechanical potentiometer.

    TI doesn't provide the minimum resistance, unfortunately. It's obviously related to process variances. I agree that it should be characterized, but the part is so old now that I doubt TI will do this.

    There is no easy way to compare the TI process with the Cirrus process and whether TI's trimming is better or worse. My experience with digital pots of various sorts is that the all have fairly lousy absolute-resistance tolerances of about +/- 20%.

    You say, "I need to know for calculating preceeding capacitor value, for minimal low bass drop." This tells me you want to add a DC blocking capacitor in front of the PGA2320, to basically "de-scratch" it and get rid of clicks when the volume level is changed. But I don't think the cap is necessary. The PGA2320 has logic that makes the volume changes occur only on zero crossings.

    I built a couple of things around these parts (a studio monitor controller was one). I was concerned that a user might make a request for "large" changes in volume, so I added some logic that would break up a "large" change into several smaller changes. That is, if the user wanted to go from -50 dB to -20 dB, the firmware would send commands to change the level as such: -47, -44, -41, -38, -35, -32, -29, -26, -23, -20, over a few tens of milliseconds. And then I compared that to just using the one large step. The large step didn't give me any noticeable clicks. I did listening tests, and I monitored the audio output on an oscilloscope triggered by the rising edge of the SPI chip select. But I thought that the ramping was less frightening for a listener, especially when monitoring at "loud" levels.

    The first cut of my first board design with this part had a footprint for a decently-largish aluminum electrolytic cap between the driver op-amp and the PGA2320 for both channels. You can check the zero-crossing detector behavior by stuffing a cap for one channel and put a jumper in place for the other and compare the two outputs with a null test.

    Good luck.

  • Hi Andy, 

    Thanks for that, this is helpful.

    I always used capacitors on the old mechanical pots to keep DC out of the pot. I realise this is no longer required, however as you said zero crossing artefacts were a worry.

    I have no qualms about leaving capacitor in, it was just I managed to put one that was a bit on the small side for the  ~12K ohms following it losing me bass in specification, if not audibly...

    Interestingly we also preferred "ramping" the value too up or down and this makes for a very good solution overall, also allowing for rotary encoder issues!

    Very best regards,

    Simon