TPA6120 without input buffer

[mr.duck]

I've been thinking; does the TPA6120 really need an input buffer? Input bias currents are relatively high, so there will probably be major DC offset problems if used on it's own, but why not use a DC servo to correct for this?

Does using an opamp as an input buffer improve performance (on paper at least)? By loading the source less for example. The input bias current is 12uA according to Per-Anders site (I couldn't see it in the datasheet ). That still seems low. I did a calculation, if the source outputs 1v RMS, then with a 50k pot for volume control, would cause it to draw 20uA (1/50k). It seems the extra loading from the TPA6120 would be not much more.

Having an opamp in the signal path looks like a bottleneck when you compare some of the TPA6120 and some opamp specs (especially slew rate!).

Would TPA6120 without input buffers and with a DC servo be a sound idea?

[ren@darkertek]

are you going to use them in differential mode like in their datasheet?

if youre going to use them in non inverting or inverting mode, you can just use a compensation resistor on the +in.

choose a value thats equals the two feedback resistors in parallel, and you adjust the +in compared to GND to the same (theoretically) as the -in feedback circuit.

this assumes equal input bias currents on both inputs. i dunno of a way to fix input bias current offset issues.

[mr.duck]

Hi, planning on single ended input and output. Non inverting config.

There is a problem trying to balance the amp with resistor values.. the TPA6120 uses low value resistors in the feedback path (1k). So conpensation resistor wud be 500 Ohm. But when a pot is also placed before the amp for volume control, it screws this calculation up i think.

[peranders]

The TPA6120 is probably rebranded THS6012. The TPA6120 is _very_ much alike a THS6012!

The input bias current isn't mentioned in the TPA6120 datasheet and I find this very odd. I have never experienced that before!

The offset is 2-12 mV PLUS these "huge" currents. This can create a rather high offset if you don't know what you are doing.

Notice also that this is a current feedback opamp which means that you can't choose any feedback resistor value. I'll guess up to 1k or something is max.

[peranders]

Originally posted by mr.duck
I've been thinking; does the TPA6120 really need an input buffer? Input bias currents are relatively high, so there will probably be major DC offset problems if used on it's own, but why not use a DC servo to correct for this?
Would TPA6120 without input buffers and with a DC servo be a sound idea?

Sure you can use a DC-servo

[mr.duck]

Hi, Per-Anders. Yes that is odd, I thought I must have missed it. They say the feedback resistor should be 10k MAX in datasheet.

May I ask you to speculate on the performance without the opamp buffer? For example, the extra current draw should not be a problem for the source (very little extra distortion), and not having the opamp in the signal path will improve sound quality. That is my view on it (speculation as well), but I would value your opinion on the design.


I will make a schematic for the design soon.

[peranders]

Originally posted by mr.duck
They say the feedback resistor should be 10k MAX in datasheet.

Where did you find this info? Rf max 10k?? RL is 10k in a diagram.


I used a buffer simple because I wanted a problem free connection to just about any signal source including a pot.

[mr.duck]

Originally posted by peranders

Where did you find this info? Rf max 10k?? RL is 10k in a diagram.

Page 14 of datasheet near the top. Quote: "It is not recommended that the feedback resistor exceed a value of 10 k. The typical value for the feedback resistor for the TPA6120A2 is 1 k."

[peranders]

OK, thanks. Much to read sometimes

[mr.duck]

Here is a schematic for TPA6120 with DC servo. I have not designed DC servo before. Please comment on the component values. Also, I have no idea how much control range the servo will have. I have not used input DC blocking capacitor because I hope servo can take care of DC from the source. Is OPA134 a good choice for servo opamp?