Well, my ears and my brain are different than yours so subjective evaluation isn't of much value to me because what sounds good to you may not sound good to me. I believe you when you say that what you are hearing is better but obviously I can't evaluate that myself without going through the very significant trouble and expense of designing and making an appropriate circuit. I may not be capable of hearing the difference either, so a higher performance circuit may be lost on me anyway.
Ok, I modeled this circuit in Tina using Analog's spice model for the AD844. Distortion is low, but not screaming low. Noise is high. S/N is less than 110dB @ 1kHz. Bandwidth rolls off @ 20kHz. Gain is high @ 68dB or so. It's not suitable for the PCM1794a because the output of that DAC is too high. It's ok for the PCM1704 @ 1.2mA.
I don't see any evidence so far that shows this is better than the normal feedback IV converter. That's not to say it doesn't sound better to you.
I'm so impressed with the LM6181 that I'm going to buy 4 of them for my IV converter. Noise is about the same as the LME49710, but distortion is even lower in an IV converter application, even at high frequencies. It needs feedback to function so it can't be used in place of the AD844 circuit on this thread.
Hi Dirk you modeled it the way I used it in post 62?
Can you do the same for Peja's way as well in post 63?
I would particually like to know what the input impedance differences are of both ways is from 5hz to a few mhz. Not interested in disortion or noise if you want to keep it simple, as I cannot detect any of these over normal opamp based I/V's.
Thanks George
Can you do the same for Peja's way as well in post 63?
I would particually like to know what the input impedance differences are of both ways is from 5hz to a few mhz. Not interested in disortion or noise if you want to keep it simple, as I cannot detect any of these over normal opamp based I/V's.
Thanks George
Last edited:
Yeah, I modeled the one in post 62. I'll get to the other later.
I'm really in love with the LM6181 now. I partner it with the BUF634 and use a 200 ohm feedback resistor.
It would be interesting to know the reasons why Ayre stopped utilizing the AD844 in the output stage of one of their earlier CD players, I don't recall the model. I believe they used the chip in the manner Pedja Rogic does, open loop with the transimpedance node connected to ground via a resistor.
The LM6181 to me (which isn't much) looks good also on paper, but it doesn't have the magic TZ (pin 5) that the AD844 has that allows one to use it without feedback.
I feel this is one of the reasons the AD844 sounds so good. It's as though feedback in an I/V stage is a big no no, as it can get corrupted with all the HF glitches and stuff comming out of the dac chips. Please correct me if I'm wrong all you I/V aficionados that know better than I do? Because I'm going more on what's sounding better, rather than what's measuring better.
Cheers George
I feel this is one of the reasons the AD844 sounds so good. It's as though feedback in an I/V stage is a big no no, as it can get corrupted with all the HF glitches and stuff comming out of the dac chips. Please correct me if I'm wrong all you I/V aficionados that know better than I do? Because I'm going more on what's sounding better, rather than what's measuring better.
Cheers George
Last edited:
If feel Tina is not giving the correct info here as I, look on the scope at the TZ point and the output buffer point of a 1khz test square wave I have on cd going into the AD844 from the PCM1704.
And the square wave was pin point square (no rounding at all) on the leading and trailing edge, it's as though it's bandwidth was at least out to 100khz, it did have a very small Gibbs ringing on both corners but that is natural, the sine wave sweep that I have goes from 5hzh to 20khz, it's dead flat all the way. I will take pics next time it's on the bench and show them.
Cheers George
Last edited:
It's all a bit above me some of you may find it interesting.
A quote from John Curl on the designer of the AD844 Barrie Gilbert
"This is what makes Barrie Gilbert's analysis significant and useful. It shows a potenial distortion that is NOT removed by global feedback"
And here is Barrie Gilbert's paper called "Are Op Amps Really Linear"
http://www.linearaudio.nl/Documents/Are Op Amps Really Linear.pdf
Cheers George
A quote from John Curl on the designer of the AD844 Barrie Gilbert
"This is what makes Barrie Gilbert's analysis significant and useful. It shows a potenial distortion that is NOT removed by global feedback"
And here is Barrie Gilbert's paper called "Are Op Amps Really Linear"
http://www.linearaudio.nl/Documents/Are Op Amps Really Linear.pdf
Cheers George
Last edited:
The phase shift for the AD844 circuit is minimal over a huge bandwidth. That may have something to do with it.
Sorry, my mistake. I ran it on +/- 5 V instead of 15VDC.
Distortion is nothing to worry about.
From what I recall Charles Hansen saying (and my memory is rather hazy), ADI moved (shrunk I guess) the original process and as a result some of the newer chips turned out noisier than the originals. Perhaps the earlier parts were beating the DS specs comfortably and the newer ones had less margin? He didn't sound too impressed with ADI's customer support anyway - their response being (as you'd expect) 'it meets the datasheet spec'.
Here is a little from the now obsolete AD846 datsheet also can be used as current feedback opamp, seems it was the forerunner to the AD844, also had a TZ type pin 5. but was mainly used for bandwidth limiting compensation or as a low mA output point but they don't say what the output impedance is at that point either, looks to be similar architecture though.
Cheers George
Cheers George
Attachments
Quotes from Charles Hansen designer of the Ayre CDP's
"Those are not op-amps. An op-amp (by definition) operates by using feedback. Our 7-series products use Analog Devices AD844's (designed by the great, but very stubborn, Barrie Gilbert) in the analog signal path, but with no feedback. So while the part was designed to be used as a current-feedback op-amp, we don't use it that way. Just one reason that they are in the "Class A" rating in Stereophile."
Quote from Jan Didden
"Yes, I agree, the AD844 are one of the best kept secrets in audio. I used them in my error correction amp as precision gain-of-one-without-feedback.
You can set the gain with a single resistor and get an open loop buffer to boot.
You remember the AD846? Almost the same except that the max lin current for the transconductance output was 10mA in stead of 1mA for the '844. Discontinued, of course.
Barrie sort of confided that he had a much better part on the shelf but unless there was sufficient demand, they would probably not produce it.
If I only had a 100 grand!"
Cheers George
"Those are not op-amps. An op-amp (by definition) operates by using feedback. Our 7-series products use Analog Devices AD844's (designed by the great, but very stubborn, Barrie Gilbert) in the analog signal path, but with no feedback. So while the part was designed to be used as a current-feedback op-amp, we don't use it that way. Just one reason that they are in the "Class A" rating in Stereophile."
Quote from Jan Didden
"Yes, I agree, the AD844 are one of the best kept secrets in audio. I used them in my error correction amp as precision gain-of-one-without-feedback.
You can set the gain with a single resistor and get an open loop buffer to boot.
You remember the AD846? Almost the same except that the max lin current for the transconductance output was 10mA in stead of 1mA for the '844. Discontinued, of course.
Barrie sort of confided that he had a much better part on the shelf but unless there was sufficient demand, they would probably not produce it.
If I only had a 100 grand!"
Cheers George
The Ayre player/s were the 7 series, I believe he used 4 per channel as I/V, filter, and two for the balanced and single ended outputs all used as the zero global feedback way. I wished they had old school multibit dacs though, then I would look/search for them s/h.
Ayre CX-7 CD player | Stereophile.com
Cheers George