| homemade |
| Just picked up an older Denon DCD-2560 cd player , which uses 4(2 per ch) current output AD1862 dac's. Since there is not very much info and/or experience on this forum with passive I/V components(resistor values and such) with regard to these dacs, I would like to start with the op-amps. I have had extremely good experience with the dual OPA2107 as a buffer , but I'm unsure if it is up to the challange for I/V , as it has a relatively low slew rate, if that is any indication of usefullness here. Any suggestions are , as always , much appreciated. |
|
|
| Leolabs |
| Any FET input audio opamp will do. |
|
|
| jameshin |
OPA2107 is a bad I/V converter, Gain Bandwidth is too low. OPA2604 is much better.
I/V conversion OPA2604 > OPA627 x 2 > OPA2107
Analog stage OPA627 x 2 >OPA2107> OPA2604 |
|
|
| philpoole |
Hi,
I'm using OPA2107 as an IV converter for my TDA1541A NOS DAC, and I think it sounds great (although I admit I haven't tried much else).
Why is slew rate more important for IV conversion than for buffering? I would have thought the criteria would be similar for both situations.
Have you heard the difference?
Cheers,
Phil |
|
|
| clem_o |
I-V conversion would require a high slew rate for accuracy - the DAC puts out step current change for each new word it receives; a low slew op amp would take some time to convert this, causing HF innaccuracies. Aside from slew rate, one should (I think) also be looking at settling time, noise, and current drive capability of the op-amp, given that the FB resistor used to set the I to V relationship tends to be pretty low value.
After LPF, the signals are into the audio band, thus a buffer need not be exceptionally fast. Low distortion is needed here though...
Cheers |
|
|
| georgehifi |
| quote: | Originally posted by clem_o
I-V conversion would require a high slew rate for accuracy - the DAC puts out step current change for each new word it receives; a low slew op amp would take some time to convert this, causing HF innaccuracies. Aside from slew rate, one should (I think) also be looking at settling time, noise, and current drive capability of the op-amp, given that the FB resistor used to set the I to V relationship tends to be pretty low value.
After LPF, the signals are into the audio band, thus a buffer need not be exceptionally fast. Low distortion is needed here though...
Cheers |
clem_o, I use an AD825 as the I/V, this is the best I've heard so far in my rig, what do you think of it for the parameters you've recommeded above does it seem a good one or are my ears decieving me?
Cheers George |
|
|
| clem_o |
Hi georgehifi,
I'm not looking at the parameters of the AD825 at the moment, but I recall it was a pretty fast and good chip - well respected by many in the audio community. I have a very good reason not to bother looking at almost any AD product - it's not available here at decent prices... :-(
One thing I have tried is the THS4032 (texas instruments) - it's actually designed for video frequency I-V. We put this into a Pioneer CD/SACD/DVD-A player and the sound change simply blew our minds...
Cheers! |
|
|
| jameshin |
| quote: | Originally posted by clem_o
I-V conversion would require a high slew rate for accuracy - the DAC puts out step current change for each new word it receives; a low slew op amp would take some time to convert this, causing HF innaccuracies. Aside from slew rate, one should (I think) also be looking at settling time, noise, and current drive capability of the op-amp, given that the FB resistor used to set the I to V relationship tends to be pretty low value.
After LPF, the signals are into the audio band, thus a buffer need not be exceptionally fast. Low distortion is needed here though...
Cheers |
I remember that an article written by Walt Jung said that Gain Bandwidth is the most important sepcification in choosing an I/V opamp. Slew rate is a related element but not as crucial as GBW. |
|
|
| philpoole |
Interesting.
I will have to investigate this.
I was of the assumption that 4.5MHz GBP would suffice for IV. The OPA2107 has a slew rate of 18V/uS. The samle period (for NOS) is 1/44100 = 22uS. So initially it would seem it could easily handle a pk-pk voltage change within one sample period for NOS at least (unless, which is very likely, I have the wrong end of the stick).
I guess for an oversampling setup, requirements could be different, and generally, I can see how a faster settling time could well be of benefit. |
|
|
| Chris Daly |
Dual Opamps are a compromise , Singles are much better
but I can't fit them I hear you say - well you can Brown Dog
make an adaptor. My recommendation is the opa627
:angel: |
|
|
| Chris Daly |
I think the parameter Walter Jung referred to was Vth given
by the Slew Rate divided by the Gain Bandwidth
philpoole if this use of a TDA 1541 is using a inverting opamp
followed by a non inverting ie- same device if a dual , check your
schematic and apply a 22k resistor globally around both ie:
from inverting input (also the Analog Dac output pin ), then of opamp 1
to output of non inverting opamp2. Note feedback arising from
output of opamp 2 then applies to correct two devices and not
just one device. Simply amazing when applied to a Marantz
CD50 using 2x opa627 per channel , close to if not the best
sound I have heard. Simple and effective
:) |
|
|
| philpoole |
Hi Chris,
Currently (if you forgive the pun), its only a single opamp per channel (its my first revision of this DAC, I'm still in the early stages of playing), but I can see how the global feedback for two opamps can be beneficial.
It will definitely go on my todo list (there's lots to do, and its great fun).
At the moment, I'm busy making a balanced I2S driver receiver, so I can connect the DAC to my CDP more sensibly.
Thanks for the tip!
Cheers,
Phil |
|
|
|
