yeah I thought about that one but find I found a better balanced one. Want I send you a pouilated pcb for personnal bechmark ? (have you two ad1862 ic?). Of course the passive parts will have a choice job as well.
OPA1633 (successor to OPA1632) has about the same settling time as AD811 (if that is important). Works fine as I/V.
and also is super for modern dac with differential output for the that task it was made : the okto dac uses it 🙂 and is well liked as it is class A according their Stereophile ears. But there are bettersssssss....
@bohrok2610 , I have to apologize, I wrote to you find bad the op861 Pedja R. is using with its off the shelf flag ship. It was @Zoran finally if I am not wrong in another thread that told me so. It is one of the guys around there with you I trust, even so with mods I find the TDA151A dac from Rogic with a tda1541A taiwann 1988 IC and few mods wirh a sota digital front end is a very high end dac on the sounding side (one of my musician friend is opera tenor and very do not care about hifi but I pay him with wine to benchmark (blindly and with some musicians friends) my confs.
I have found in one of my protos the ad828 was very ok but needs some works on lythic decoupling caps to add it some thickness for a rigther tonal balance, at least in my setup.
As all of you are good designers, I wonder if it was not a setuping of the whole hifi with the room that matters at the end (i.e. the last sound going out from the loudspeaker after all the devices and loudspeaker interactions, which is hard to modelize ! Just for illustration a flat spl loudspeaker curve with a NOS dac over a sligthy diving curve from 1k to 2k hz till the "20 k hz". But that is off topic.
😊
@bohrok2610 , I have to apologize, I wrote to you find bad the op861 Pedja R. is using with its off the shelf flag ship. It was @Zoran finally if I am not wrong in another thread that told me so. It is one of the guys around there with you I trust, even so with mods I find the TDA151A dac from Rogic with a tda1541A taiwann 1988 IC and few mods wirh a sota digital front end is a very high end dac on the sounding side (one of my musician friend is opera tenor and very do not care about hifi but I pay him with wine to benchmark (blindly and with some musicians friends) my confs.
I have found in one of my protos the ad828 was very ok but needs some works on lythic decoupling caps to add it some thickness for a rigther tonal balance, at least in my setup.
As all of you are good designers, I wonder if it was not a setuping of the whole hifi with the room that matters at the end (i.e. the last sound going out from the loudspeaker after all the devices and loudspeaker interactions, which is hard to modelize ! Just for illustration a flat spl loudspeaker curve with a NOS dac over a sligthy diving curve from 1k to 2k hz till the "20 k hz". But that is off topic.
😊
Last edited:
Its noise below 1kHz (current noise primarily) would rule it out for me, it has more current noise than a CFB in the lower frequencies.
AD828 has the caveat that its not claimed to be unity gain stable - being designed as a +6dB video buffer. So its stability depends on the DAC output impedance relative to the I/V resistor. Input LTP degeneration makes it a tad noisy.
Perhaps low noise enough but greater linarity ? Important thing for i/v isn'it?
Btw, would you choose systematixcly a unity gain stable OPA for that transimpedance job?
Btw, would you choose systematixcly a unity gain stable OPA for that transimpedance job?
Linearity is rather moot if its oscillating or peaking grossly due to its noise gain being below minimum allowed.
These days I don't go for OPAs in digital audio applications. Theoretically a unity gain stable OPA isn't required unless the DAC's output impedance is higher than the I/V (feedback) resistor.
These days I don't go for OPAs in digital audio applications. Theoretically a unity gain stable OPA isn't required unless the DAC's output impedance is higher than the I/V (feedback) resistor.
The LME4924 also works great, although it is bipolar, it sounds no worse than 1632, but consumes less current (self-heating), about the same as 1633.
Alex.
As we are at it, somz liked also for the task the ad8065... but OP828 shows incridibly better low end bzlow 2k in the input jfet based family. Haven t tried this one.
Voltage noise plot for AD8065 shows 'designed for RF' characteristics with a 1kHz noise corner.
Hi thanks.
What are FDA letters are ?
Since I found the op861, which is noiser than a virgin at a megastor, sounding goid with care, I am less focussed on the nouse floor only. Op828 and op211 (big brother of op1611?) Are hard to beat below 2k hz each in tyeir own topology input, while many despite the very low noise floor found it sounds not so good at transimpedance job.
I should be a partisant to blind test them and measure after to try measurement correlations on many departments more than the harmonics spectrum only.
Off topic maybe : you talked of the output capacitance of gurrent output DAC. How measures it please if not on the datasheet?
OPA certainly terrible for that task...but the intetresting compacness of the layout. And easier for enthusiast like me not skilked asa EE. I liked Bohrok260 test of 4 I/V where he tested also Sergliss discrete 🙂
Last edited:
For ES9039Q2M composite opamp as I/V works wonders. Not so with ES9038Q2M probably due to dac chip limitations.