Just got around to changing the OPA2132 I/V for the PCM1796 dac to the LM6172 for I/V duties, and all I can say abraxalito it's far better in every respect, I'm still using a OPA2132 as the output buffer/filter though, it seem to be a great pairing for I/V and buffer for the PCM1796.
What a great I/V opamp the LM6172 is, I would have never thought it would be this good.
Cheers George
Last edited:
This was one of the most important evidence produced by Abraxalito a while ago here and i fully agree with it.
Welcome and join the gang
If you look at the simplified schematic of LM6172 you'll see its essentially the same as an AD844 but with a buffer feeding the -ve input pin which gives it the characteristics of a VFB opamp. I reckon the PSRR of this buffer is likely its weakest link, most likely its a diamond (as shown in various LT amps with the same architecture). Incidentally has anyone tried those Linear Technology offerings as I/V, I mean LT1818/9 and LT6274/5 ?
The psrr is good enough...we have fast regulators anyway, what really troubled me was the current noise as i would have wanted to use it with 7x lower iout dacs of the past.THS4631 would probably qualify for that purpose better than lt 1818/6274. It has about 800v/us and much lower current noise while the voltage noise is just about the same .Its serious drwback is the quiscent current though.At least its pad is connected to gnd not at v- as in ada4898.
Lt6274 might have that same buffer as 6172 as it shares the low psrr.In its datasheet is a diamond buffer indeed...maybe the missing part of the lm6172 simplified schematic.
I had a serious reason to get back to the audio op-amps of the past(30 years old) with my own circuits as analyzing all these new op-amps they never have it all in one package.
Lt6274 might have that same buffer as 6172 as it shares the low psrr.In its datasheet is a diamond buffer indeed...maybe the missing part of the lm6172 simplified schematic.
I had a serious reason to get back to the audio op-amps of the past(30 years old) with my own circuits as analyzing all these new op-amps they never have it all in one package.
Last edited:
Current noise doesn't look to me to be a problem but then I'm only using a 16bit DAC. For LT1818 its 1.2pA/rtHz which means you'd need to have a 5k FB resistor to get to the same magnitude as the voltage noise (6nV/rtHz).
With JFET opamps it pays to watch how hot they're getting - I agree the THS4631 is a tad on the toasty side. JFETs when warm have considerably higher current noise.
With JFET opamps it pays to watch how hot they're getting - I agree the THS4631 is a tad on the toasty side. JFETs when warm have considerably higher current noise.
Last edited:
LM6171 was really designed for driving stuff, it's unity gain bandwidth isn't that impressive, but the high slew rate lets it achieve that bandwidth with large output swing.
It's a bit noisy and distortion in the audio band could be better, as you pointed out.
THS4631 is a nice op-amp, at least you get a FET input for all that voltage noise. If you're going to go with a single and you want a JFET input, then might I suggest OPA828.
It's a bit noisy and distortion in the audio band could be better, as you pointed out.
THS4631 is a nice op-amp, at least you get a FET input for all that voltage noise. If you're going to go with a single and you want a JFET input, then might I suggest OPA828.
There were repackaged OPA627s right?
I'm mostly joking. I doubt anyone would go through the trouble of counterfeiting AD8065, but I've been wrong before.
LOL
They could well be recycled ones - de-soldered from boards and cleaned up. The current consumption was spot-on for the typical for an AD8065 but doubtless that could be faked with a suitable current source between the supply pins....
The datasheet shows NE5534A's throughout for I/V conversion and LPF, in the circuitry used in measuring the performance of the PCM1796. Something at least as good in noise and distortion is a wise choice. The OPA1612 has been mentioned and that ought to work fine (though NE5534/2's will win in price).
All the circuitry in the datasheet is very low impedance, so current noise of good BJT opamps is unlikely to be an issue.
__Mark T
Well, after living with now for around 10 or so 2hr sessions, it's not for me.
The LM6172 (properly decoupled and not oscillating) make the great recordings sound stunning but bit too clinical for me, but it makes the bad recordings unlistable.
The OPA2124 on the great recordings has less etched detail, just as much bass if a touch rounded, but has a more musical presentation, and makes the bad recordings quite listenable.
THE SEARCH GOES ON.
Cheers George
Last edited:
Looks like some good contenders from TI's new "SoundPlus" range, Essential, Premier, Ultimate.
http://www.ti.com/amplifier-circuit/op-amps/audio/soundplus.html
Trouble for me is adapter boards needed, to much mucking around to try all these.
Does anyone have a favourite for a dual for I/V among all these??
What's better for I/V fet or bi-polar input?
Cheers George
http://www.ti.com/amplifier-circuit/op-amps/audio/soundplus.html
Trouble for me is adapter boards needed, to much mucking around to try all these.
Does anyone have a favourite for a dual for I/V among all these??
What's better for I/V fet or bi-polar input?
Cheers George
Last edited: