It would be nice if there was an AD811 among these op amps. I think that would clear up all doubts about op ampsOh no, no, no.
Plenty of opamp rolling🤣
This is the motherboard.OPA1656 was particularly liked by @mikorist in his PC sound card.
It's not normal how well this plays music. Almost like a DAC.
ALC codec.....GA-H81.AMP-UP......Seasonic PRIME 600 Titanium Fanless PSU
I ordered the OPA1641AID for the DAC.
Last edited:
I would like to try that one also.It would be nice if there was an AD811 among these op amps. I think that would clear up all doubts about op amps
Trailer :
I posted earlier 27R passive IV + 100x gain stage right after, in 3 favours.
https://www.diyaudio.com/community/...s-input-nos-r-2r.354078/page-271#post-7164070
https://www.diyaudio.com/community/...s-input-nos-r-2r.354078/page-290#post-7194093
https://www.diyaudio.com/community/...st-tht-i2s-input-nos-r-2r.354078/post-7160371
https://www.diyaudio.com/community/...st-tht-i2s-input-nos-r-2r.354078/post-7192531
Fran has finally done enough listening and will post his impressions later when he gets some time.
The IC versions can take power from your standard opamp IV sockets.
The JFET version needs a VERY good 32V regulated supply, like the Didden super reg.
Fun to build. Stay tuned.
Patrick
I posted earlier 27R passive IV + 100x gain stage right after, in 3 favours.
https://www.diyaudio.com/community/...s-input-nos-r-2r.354078/page-271#post-7164070
https://www.diyaudio.com/community/...s-input-nos-r-2r.354078/page-290#post-7194093
https://www.diyaudio.com/community/...st-tht-i2s-input-nos-r-2r.354078/post-7160371
https://www.diyaudio.com/community/...st-tht-i2s-input-nos-r-2r.354078/post-7192531
Fran has finally done enough listening and will post his impressions later when he gets some time.
The IC versions can take power from your standard opamp IV sockets.
The JFET version needs a VERY good 32V regulated supply, like the Didden super reg.
Fun to build. Stay tuned.
Patrick
Front end: “yellow” SD card player from Aliexpress with Mori clocks, Miro1360 AD1862 DAC with IV stages as listed below. 01a preamp, SS power amp and stacked quad ESL speakers.
Caveats:
IV stages are not as close to the DAC as they should be – because I am connecting/disconnecting I need to run longer lengths than are ideal
Quads go low enough to deliver satisfying bass, but don’t deliver anything below ~40hz
Different stages require different power supplies, and these may well have an effect
These are subjective opinions, and all subject to the biases I have… so treat accordingly
My long term reference IV stage has been one based on the 3a5 DHT valve, with Rod Coleman filament supplies and Ale Moglia’s HV supplies. This delivers a glorious mid rang with lots of detail, full soundstage with width and depth that draws you in. The quads are excellent at soundstage, so the combination of the two really matches well. This is not an easy build, and requires close attention to all aspects of the build. The basic circuit is pretty much exactly as outlined on Ale’s site:
https://www.bartola.co.uk/valves/2016/05/11/filament-bias-a-practical-example-with-3a5-dht/
Over the course of a week, I listened carefully to 5 different IV stages. These were played in different orders over that week, but with the same tracks on each occasion to try and avoid patterns.
First up is a passive IV stage using the AD8429 opamp - an instrument amp with 3 opamps inside, has high input impedance and negative feedback . The IV resistor is set at 27r, and the opamp configured for 100x gain, and powered by a +/-15V low noise supply from diyinhk.com. This delivers a really smooth gentle refined sound. In my set up the soundstage from side to side is excellent, but not as deep as some others. Excellent sense of “blackness” between the notes very evident in more sparse music. I would say overall this is a more refined laid back sound – if you had a partnering system that tended to be a little harder then this could be a perfect match. I built this on some adapter boards, and given the low parts count this turned out as a very compact size.
Next is another passive IV-opamp stage but this time using AD797, well known, single opamp with negative feedback. This stage is very similar to the AD8429, but vocals are a little more forward, and there is more sense of depth to the soundstage. I remember thinking as I listened that it was impressive, and easy to build, low parts count. Either of these two stages would deliver long term enjoyment for very little outlay. Both opamp based stages were powered by the same supply.
The next IV stage is a discrete jfet IV stage, and powered by a Salas ubib shunt supply. This has no negative feedback, has output coupling caps, a single ended circuit with H2 dominant and not the lowest distortion ever. This stage is a little unusual, in that sound changes a little over the first 15mins or so – I waited at least 20mins before listening. You can imagine that this alone creates bias – I suppose so does everything. The scale of the soundstage is bigger here than with either of the opamp stages. Side to side width is about the same, but there is more front to back depth. Vocals, although bigger (due to the scale) I think are a tad softer than with the passive IV opamp stages, and piano is a little more dynamic, key strikes carrying more impact. There is more recording ambiance here too – say the slight reverberation you hear in a recording.
Next up is the FC CEN IV discussed elsewhere here, a JFET push pull current conveyor with no negative feedback and DC coupled at input and output – which in short I consider sonically to be the equal of the 3a5 reference stage. In other ways it is far superior – cheaper to build, quieter, less prone to noise pickup, tunable Sallen-Key filter to name a few. If I hadn’t built the 3a5 stage before FC CEN was available, I would never have made it. Bass is tight and with lots of texture. I believe there is more presence here than with other stages (ie the illusion that the performance is in the room) but I would think this is extremely subjective. This is a standard to aim for, it is at the top of the list – and over the last few years I have made and listened to a lot of IV stages, all with the same DAC and associated equipment.
I want to add that I am listening to these and searching out differences – the things I am pointing out above are generally small differences, and honestly are not that easy to determine. The differences between both opamp stages for example are really tiny. Having said all that….. 3a5/FC CEN IV remain at the top of the pile for me, then jfet behind by a small but definite margin, and the both opamp stages together. I would think that different power supplies might make a difference to that running order. I think the passive IV with AD8429/AD797 would make a great alternative for someone who wants to keep the small footprint of the Miro DAC onboard opamps (and supplies) where the JFET and FC CEN are more involved builds and need their own supplies
And no, I am not showing a picture of the 3a5 stage because its a rats nest and not fit to show in public!!! Sounds great though
Caveats:
IV stages are not as close to the DAC as they should be – because I am connecting/disconnecting I need to run longer lengths than are ideal
Quads go low enough to deliver satisfying bass, but don’t deliver anything below ~40hz
Different stages require different power supplies, and these may well have an effect
These are subjective opinions, and all subject to the biases I have… so treat accordingly
My long term reference IV stage has been one based on the 3a5 DHT valve, with Rod Coleman filament supplies and Ale Moglia’s HV supplies. This delivers a glorious mid rang with lots of detail, full soundstage with width and depth that draws you in. The quads are excellent at soundstage, so the combination of the two really matches well. This is not an easy build, and requires close attention to all aspects of the build. The basic circuit is pretty much exactly as outlined on Ale’s site:
https://www.bartola.co.uk/valves/2016/05/11/filament-bias-a-practical-example-with-3a5-dht/
Over the course of a week, I listened carefully to 5 different IV stages. These were played in different orders over that week, but with the same tracks on each occasion to try and avoid patterns.
First up is a passive IV stage using the AD8429 opamp - an instrument amp with 3 opamps inside, has high input impedance and negative feedback . The IV resistor is set at 27r, and the opamp configured for 100x gain, and powered by a +/-15V low noise supply from diyinhk.com. This delivers a really smooth gentle refined sound. In my set up the soundstage from side to side is excellent, but not as deep as some others. Excellent sense of “blackness” between the notes very evident in more sparse music. I would say overall this is a more refined laid back sound – if you had a partnering system that tended to be a little harder then this could be a perfect match. I built this on some adapter boards, and given the low parts count this turned out as a very compact size.
Next is another passive IV-opamp stage but this time using AD797, well known, single opamp with negative feedback. This stage is very similar to the AD8429, but vocals are a little more forward, and there is more sense of depth to the soundstage. I remember thinking as I listened that it was impressive, and easy to build, low parts count. Either of these two stages would deliver long term enjoyment for very little outlay. Both opamp based stages were powered by the same supply.
The next IV stage is a discrete jfet IV stage, and powered by a Salas ubib shunt supply. This has no negative feedback, has output coupling caps, a single ended circuit with H2 dominant and not the lowest distortion ever. This stage is a little unusual, in that sound changes a little over the first 15mins or so – I waited at least 20mins before listening. You can imagine that this alone creates bias – I suppose so does everything. The scale of the soundstage is bigger here than with either of the opamp stages. Side to side width is about the same, but there is more front to back depth. Vocals, although bigger (due to the scale) I think are a tad softer than with the passive IV opamp stages, and piano is a little more dynamic, key strikes carrying more impact. There is more recording ambiance here too – say the slight reverberation you hear in a recording.
Next up is the FC CEN IV discussed elsewhere here, a JFET push pull current conveyor with no negative feedback and DC coupled at input and output – which in short I consider sonically to be the equal of the 3a5 reference stage. In other ways it is far superior – cheaper to build, quieter, less prone to noise pickup, tunable Sallen-Key filter to name a few. If I hadn’t built the 3a5 stage before FC CEN was available, I would never have made it. Bass is tight and with lots of texture. I believe there is more presence here than with other stages (ie the illusion that the performance is in the room) but I would think this is extremely subjective. This is a standard to aim for, it is at the top of the list – and over the last few years I have made and listened to a lot of IV stages, all with the same DAC and associated equipment.
I want to add that I am listening to these and searching out differences – the things I am pointing out above are generally small differences, and honestly are not that easy to determine. The differences between both opamp stages for example are really tiny. Having said all that….. 3a5/FC CEN IV remain at the top of the pile for me, then jfet behind by a small but definite margin, and the both opamp stages together. I would think that different power supplies might make a difference to that running order. I think the passive IV with AD8429/AD797 would make a great alternative for someone who wants to keep the small footprint of the Miro DAC onboard opamps (and supplies) where the JFET and FC CEN are more involved builds and need their own supplies
And no, I am not showing a picture of the 3a5 stage because its a rats nest and not fit to show in public!!! Sounds great though
Now I know why the DAC chips are so hard to findOh no, no, no.
Plenty of opamp rolling🤣
Nicely done Fran .
Best sounding (To ME!!)... EUVL's FC CEN IV / Pedja Rogics OPA861 Transconductance Output Stage
.Budget Friendly Fav.... OPA1655
Best sounding (To ME!!)... EUVL's FC CEN IV / Pedja Rogics OPA861 Transconductance Output Stage
Hello,
I got the ball rolling on my Ad1862 build, im hoping someone could take a look at my intended transformer and psu setup.
The length of digital wire to the dac board is probably an issue, transformer and rectification looks a little cramped but my scale is off. The copper lines dividing the transformers and digital section will be some sort of isolation barrier. Im a little confused how to provide 2 x +5v for the i2soverusb, for the USB and Clock with the Super Reg. Any advice would be great or a post link that would enlighten me.
Thanks
I got the ball rolling on my Ad1862 build, im hoping someone could take a look at my intended transformer and psu setup.
The length of digital wire to the dac board is probably an issue, transformer and rectification looks a little cramped but my scale is off. The copper lines dividing the transformers and digital section will be some sort of isolation barrier. Im a little confused how to provide 2 x +5v for the i2soverusb, for the USB and Clock with the Super Reg. Any advice would be great or a post link that would enlighten me.
Thanks
North4,
To save some room in your chassis, use Miro’s board with integrated JLS I2SoverUSB, then no worries about long wires. Also, use R-core trafo’s with 4 secondaries. Probably only need 2 instead of 4 trafo’s. 👍🏻
You don’t have to worry about inrush, way less than 200va total.
To save some room in your chassis, use Miro’s board with integrated JLS I2SoverUSB, then no worries about long wires. Also, use R-core trafo’s with 4 secondaries. Probably only need 2 instead of 4 trafo’s. 👍🏻
You don’t have to worry about inrush, way less than 200va total.
Attachments
Thank you, i will give that more thought. I have habit of overcomplicating projects that i dont quite have the experience to deliver on. I like your build, had that one bookmarked. Vunce 1862 build.
I remember reading something about R-core not having inrush like a toroid does, low va is probably more influential. That would certainly simplify things. As far as Miros integrated JLS I2SoverUSB board, great design, am i wrong in thinking having the JLS on its own power like in my picture is worth the trouble? If not, i might as well send out for a set of the integrated boards.
I remember reading something about R-core not having inrush like a toroid does, low va is probably more influential. That would certainly simplify things. As far as Miros integrated JLS I2SoverUSB board, great design, am i wrong in thinking having the JLS on its own power like in my picture is worth the trouble? If not, i might as well send out for a set of the integrated boards.
Many thanks to Fran for the write-up, and of course the many hours spent on auditioning.
A few notes perhaps on passive IV.
People will tell you that current-out R2R DACs need to see a virtual Gnd. Hence the usual opamp IVs. But many of them actually work quite well with a small load impedance. For example in this tutorial by Analog Device, a 25R IV resistor is used :
https://www.analog.com/media/en/training-seminars/tutorials/MT-019.pdf
If we use a 27R resistor directly at the output of an AD1862, we shall get +/-27mV output. Amplify this by 100x then gives you about 2Vrms. The question is then how to do 100x amplification.
The easiest is to use a low-distortion, low-noise, low-DC-offset opamp, such as AD797 or AD8429, as they are used e.g. in phono preamps. The AD8429 is especially easy as the gain can be set by a single resistor. The AD797 has slightly better performance, so choice is more subjective. But both have good PSRR and can work with say +/-12V. So this makes them very easy to use in this particular application. And now you can do resistor rolling as well. Anything from Beyschlag MELF, Nikkohm RP-44, to Vishay bulk foil or fancy wire-wound resistors, the list is endless.
The discrete JFET gain stage is essentially also a 100x voltage amplifier with a tube-amp like topology. Which means it also has all the drawbacks such as no PSRR or capacitive output coupling. But it is H2 dominant, which gives you this tube-like sound that some prefers. As it is using all active components, it is low cost and fun to build. Only requirement is a low noise power supply, like the Didden super-reg. The otherwise excellent LT3045 will not do the required 32V output. What a pity.
Anyone want to do Beta-test, Fran has some boards left, so you can contact him direct.
Have fun,
Patrick
A few notes perhaps on passive IV.
People will tell you that current-out R2R DACs need to see a virtual Gnd. Hence the usual opamp IVs. But many of them actually work quite well with a small load impedance. For example in this tutorial by Analog Device, a 25R IV resistor is used :
https://www.analog.com/media/en/training-seminars/tutorials/MT-019.pdf
If we use a 27R resistor directly at the output of an AD1862, we shall get +/-27mV output. Amplify this by 100x then gives you about 2Vrms. The question is then how to do 100x amplification.
The easiest is to use a low-distortion, low-noise, low-DC-offset opamp, such as AD797 or AD8429, as they are used e.g. in phono preamps. The AD8429 is especially easy as the gain can be set by a single resistor. The AD797 has slightly better performance, so choice is more subjective. But both have good PSRR and can work with say +/-12V. So this makes them very easy to use in this particular application. And now you can do resistor rolling as well. Anything from Beyschlag MELF, Nikkohm RP-44, to Vishay bulk foil or fancy wire-wound resistors, the list is endless.
The discrete JFET gain stage is essentially also a 100x voltage amplifier with a tube-amp like topology. Which means it also has all the drawbacks such as no PSRR or capacitive output coupling. But it is H2 dominant, which gives you this tube-like sound that some prefers. As it is using all active components, it is low cost and fun to build. Only requirement is a low noise power supply, like the Didden super-reg. The otherwise excellent LT3045 will not do the required 32V output. What a pity.
Anyone want to do Beta-test, Fran has some boards left, so you can contact him direct.
Have fun,
Patrick