ojg said:
Excellent articles that you can never link to one time too many.
Obviously those still recommending OPA627 and the like for I/V converters must know something that Jung and Gilbert does not
you might check employers, I guess you don't get to be an adi fellow by recommending ti/bb product
For fast settling time / high resolution / high sampling rates DAC's there's only one I to V OP: AD8066. Excellent! Careful - it works up to + - 12V only, and may need a heat sink. It is extremely fast with very low input bias current / high impedance - allows all those fine and fast current step-changes (out of the DAC) to arrive properly to the OP input (unchanged), and be "registered" and amplified, and smoothed - out. The resulting analog out will still have all those fine step-changes details present - wonderful!
I also learned to dislike BB: their DAC's sound cold, their OP's sound warm.
AD is good in both - very good!
Extreme_Boky
I also learned to dislike BB: their DAC's sound cold, their OP's sound warm.
AD is good in both - very good!
Extreme_Boky
Extreme_Boky,
I totally agree with you, but have a look on datasheet specs for OPA657 http://focus.ti.com/lit/ds/symlink/opa657.pdf
It looks very well suited for such job.
tibi
Ps: AD8065/6 is also my favorite OP amp.
I totally agree with you, but have a look on datasheet specs for OPA657 http://focus.ti.com/lit/ds/symlink/opa657.pdf
It looks very well suited for such job.
tibi
Ps: AD8065/6 is also my favorite OP amp.
I'm currently modding my MF XDACv3. It was using NE5532 for I/V and Diff and I replaced with OP2134 but that was before I did any real homework. There was an audible improvement but I concede that these OPA devices are not really well suited to I/V.
I've now ordered a stockpile of AD8066 and I'll putting them to use asap. Unfortunately the XDAC uses DSD1792 and I'm yet to hear these DACs deliver the resolution and 'liquidity', (for want of a better adjective) in the treble. Hopefully the op-amp upgrade will deliver but I'll post my subjective findings either way.
BTW, I've done quite a bit of RF design so I'll use best RF bypass practise (and my CRO) to make sure I don't construct an EMI generator in the process.
I've now ordered a stockpile of AD8066 and I'll putting them to use asap. Unfortunately the XDAC uses DSD1792 and I'm yet to hear these DACs deliver the resolution and 'liquidity', (for want of a better adjective) in the treble. Hopefully the op-amp upgrade will deliver but I'll post my subjective findings either way.
BTW, I've done quite a bit of RF design so I'll use best RF bypass practise (and my CRO) to make sure I don't construct an EMI generator in the process.
MF makes very good designs, which require attention when it comes to components replacement.
AD8066 is very revealing OP and emphasises the (basic) character of the DAC / CD player. In other words, if you use AD8066 in your MF DAC, you have to “balance” the sound-out by choosing "warm" sounding capacitors around analog stage / in power supplies. Try Black Gates and paper in oil (warm!) coupling caps. With this approach, you could end-up with very nice sound!!!
I have modified 2 MF CD players and a DAC, and all had excellent resolution, low noise - but sounded a bit sterile. The basics where there!
Extreme_Boky
Thanks for the advice. I'll definitely search around for the Black Gates and give them a go. I have already shorted out the coupling caps because my preamp input is cap coupled with polypropylenes, but if I'll look around for paper in oil and compare. BTW, the XDAC is now very close to being as good as my Marantz SA-11s1 on RB CD which is promising because it was behind before any mods. What brought about a significant leap forward was the installation of 78L12 and 79L12 voltage regulators on the PCB very close to the op-amps. That seemed to remove 80% of the sterility. I've actually removed the XDAC PC board from the original XDACv3 box and mounted it inside a single unit pro rack case which looks far better, and also gave me enough space to install a nice +/- 15V regulated supply and another board to allow switching of several SPDIF inputs. MF do sell an external PSU but from my experience you don't get great results unless you can regulate reasonably close to the analogue stages. It will be interesting to see (and hear) how far I can go with it.
Cheers,
D.L.
Cheers,
D.L.
Extreme_boky,
as I understand your last post (the one where you refer to the modded MF CD-players) you find the AD8066 sound a 'bit sterile'. I heard the opamp as a buffer in a DAC and found it very revealing also unbalanced in the sense that the the higher frequencies were amplified more than the rest of thye sound. I also didn't like the more forward presentation of the sound, as I am used to a layed back presentation.
I wonder if you ever tried the LM6181 or LM6171 as I/V stage, as I tried those after reading positive posts here about them and I must agree they perform better then the opa627 I used in a lot of cap configurations. They are not so revealing as the AD8086 (and hence less analytical) however they sound more correct IMO using a decent PS, propper decoupling caps and decoupling methods. They don't require compensation by using fancy caps as with the AD8086's as you mention.
David: it is my experience as you start using the better opamps for IV or whatever audio-stage, the power supply and decouplnig becomes critical and standard regulators as 78 or 79 series as well as TL431, used as described in there datasheets, won't let you experince the potential of those opamps. Try descent shunt-regulator design (lost have been posted on this forum, also the search engine is still a night mare to find anything here) or the ALW-regulator which I use and find very good for low-power (low-noise, low-imp) applications.
Henk
as I understand your last post (the one where you refer to the modded MF CD-players) you find the AD8066 sound a 'bit sterile'. I heard the opamp as a buffer in a DAC and found it very revealing also unbalanced in the sense that the the higher frequencies were amplified more than the rest of thye sound. I also didn't like the more forward presentation of the sound, as I am used to a layed back presentation.
I wonder if you ever tried the LM6181 or LM6171 as I/V stage, as I tried those after reading positive posts here about them and I must agree they perform better then the opa627 I used in a lot of cap configurations. They are not so revealing as the AD8086 (and hence less analytical) however they sound more correct IMO using a decent PS, propper decoupling caps and decoupling methods. They don't require compensation by using fancy caps as with the AD8086's as you mention.
David: it is my experience as you start using the better opamps for IV or whatever audio-stage, the power supply and decouplnig becomes critical and standard regulators as 78 or 79 series as well as TL431, used as described in there datasheets, won't let you experince the potential of those opamps. Try descent shunt-regulator design (lost have been posted on this forum, also the search engine is still a night mare to find anything here) or the ALW-regulator which I use and find very good for low-power (low-noise, low-imp) applications.
Henk
I wouldn’t call it sterile. It is very reviling of everything else. If you pay attention to lowering the noise, clock oscillator, wiring....the whole design - not only bits here and there, the 8066 is fantastic.
However, if you decide to replace only analog OP's with this beauty, the result is going to be "to hard to handle".
I suggest 2134 on one (good for ‘masking’) side, 275 as good all rounder, 627 as fast but a bit suffocating, 826 as very good!, and finally - 8066. Haven't tried LM61** - can not get it down under - it's like living in isolation down here. I think it's close to 826.
Extreme_Boky
unother reason to use the lm6171 or 81,
is it max V+ and V- is 18V compared to th 13V of the AD8086. One of my rules is that higher voltage improves the linearity of an amplifying component and is therefor preferred.
This aspect is not the reason of the different sound of the National and Analog Device opamps. They are to different in design for that as the lm's are based on bipolar devices where the AD opamp uses jFet technology (at least for the input).
According to the specialists the lm require (decoupling) caps of V+ and V- to GND where decoupling of the 8086 performs the best as caps are placed between V+ and V-.
Henk
is it max V+ and V- is 18V compared to th 13V of the AD8086. One of my rules is that higher voltage improves the linearity of an amplifying component and is therefor preferred.
This aspect is not the reason of the different sound of the National and Analog Device opamps. They are to different in design for that as the lm's are based on bipolar devices where the AD opamp uses jFet technology (at least for the input).
According to the specialists the lm require (decoupling) caps of V+ and V- to GND where decoupling of the 8086 performs the best as caps are placed between V+ and V-.
Henk