I’ve made some progress on EUVL’s Current Mirror I/V converter boards, one pair is soldered. Just checked the output DC offset using my bench supply feeding +/-12vdc:
Board #1 = less than 1mV
Board #2 = 6.5mV
These measurements are without any trim resistors installed! I’d say Ripster did a Fantastic job matching the transistors!!! Thank You 🙂
Board #1 = less than 1mV
Board #2 = 6.5mV
These measurements are without any trim resistors installed! I’d say Ripster did a Fantastic job matching the transistors!!! Thank You 🙂
Attachments
AD1862, I2SoverUSB v.III - not flipped, 20-Bit uf.l connectors

@miro
Just received the new board from PCB house today, populate and fire up...it works first try!
So neat, no more jumpers, just piggyback.
Thank you so much!
I2SoverUSB not flipped (pin-headers installed on the bottom side on I2SoverUSB) ... the heatsink and USB connector are on the TOP 🙂

@miro
Just received the new board from PCB house today, populate and fire up...it works first try!
So neat, no more jumpers, just piggyback.
Thank you so much!
Last edited:
Meanie thanks for posting a confirmation. I was hoping one user could confirm the stacked boards before I built mine up. Time to order PCBs!
I've got DAC chips, i2s board, and psu board.
I just want to be like all you cool kids with your solder and your bits and your voltages
I've got DAC chips, i2s board, and psu board.
I just want to be like all you cool kids with your solder and your bits and your voltages
I'm looking through BOM's for power supplies. Some of these regulators are out of stock and Mouser + Digikey
Question: in the case of say an MC7805, what do the suffix letters mean?
In what ways are the MC7805ACTG different from say the MC7805ABGT ?
Question: in the case of say an MC7805, what do the suffix letters mean?
In what ways are the MC7805ACTG different from say the MC7805ABGT ?
Today I have discovered a way to make my miro's AD1862 sound significant better. I am pleased to share my joy here.
I have been comparing the sound quality between my modded DIYinHk ad1862 and miro's ad1862 and have been frustrated to find that I always like the DIYinHk version better.
I listen to music but mostly I use family videos I recorded myself as a benchmark to judge the sound quality of my dac, because I am much more familiar to how my family members and I myself are supposed to sound.
With the same i2s input, the same power supplies and the same discrete opamp (ss3601 for miro's and ss3602 for diyinhk's dac), the two ad1862 dacs sound very much different. The diyinhk dac has always sounded more correct and enjoyable. I was admittedly disappointed because I was expecting improvement over my diyinhk dac that I had put together before the miro's version.
I started to investigate what might be causing the sound difference. I bypassed the 100R output capacitor, but there was no change in perceived sound quality. The diyinhk version has stop clock mechanism and separate digital and analog grounds, miro's version has a nice ground plane.
Would it be possible that the stop clock mechanism is the main reason behind the difference in sound? I studied the data sheets of the various opamps I have until I found this:
View attachment 998699
I was excited to see that shunt capacitors can be used to absorb clock energy and current steps and decided to try it. I didn't have 2000pf caps on hand. The closest I had were 0.01uf C0G and 1000pf Wima film. I tried the 0.01 uf first, but I heard oscillations. Then I put a 1000pf right across pins 2 and 3 of AD797brz and was pleasantly surprised to hear a nice and big difference in sound!
The sound is now open, smooth and fluid. The resolution has been drastically increased. It is as if I were seeing music with a magnifying glass. The front to back sound stage has clearly expanded. Music is live and 3D. Every song sounds different. A layer of hoarsness has been removed.
Now, finally I am not sure which dac I will prefer. So far the miro's dac has sounded best, but I haven't implemented the same shunt caps for my diyinhk dac yet. I will order some 2000pf caps to try later.
Anyway, I am full of joy now. Kudos to the engineers at Analog Devices who had designed the AD797, who had invented the stop clock mechanism and who had invented the shunt caps mechanism. And thank you miro and diyinhk for designing these nice ad1862dacs!
Cheers😀😀😀
I have been comparing the sound quality between my modded DIYinHk ad1862 and miro's ad1862 and have been frustrated to find that I always like the DIYinHk version better.
I listen to music but mostly I use family videos I recorded myself as a benchmark to judge the sound quality of my dac, because I am much more familiar to how my family members and I myself are supposed to sound.
With the same i2s input, the same power supplies and the same discrete opamp (ss3601 for miro's and ss3602 for diyinhk's dac), the two ad1862 dacs sound very much different. The diyinhk dac has always sounded more correct and enjoyable. I was admittedly disappointed because I was expecting improvement over my diyinhk dac that I had put together before the miro's version.
I started to investigate what might be causing the sound difference. I bypassed the 100R output capacitor, but there was no change in perceived sound quality. The diyinhk version has stop clock mechanism and separate digital and analog grounds, miro's version has a nice ground plane.
Would it be possible that the stop clock mechanism is the main reason behind the difference in sound? I studied the data sheets of the various opamps I have until I found this:
View attachment 998699
I was excited to see that shunt capacitors can be used to absorb clock energy and current steps and decided to try it. I didn't have 2000pf caps on hand. The closest I had were 0.01uf C0G and 1000pf Wima film. I tried the 0.01 uf first, but I heard oscillations. Then I put a 1000pf right across pins 2 and 3 of AD797brz and was pleasantly surprised to hear a nice and big difference in sound!
The sound is now open, smooth and fluid. The resolution has been drastically increased. It is as if I were seeing music with a magnifying glass. The front to back sound stage has clearly expanded. Music is live and 3D. Every song sounds different. A layer of hoarsness has been removed.
Now, finally I am not sure which dac I will prefer. So far the miro's dac has sounded best, but I haven't implemented the same shunt caps for my diyinhk dac yet. I will order some 2000pf caps to try later.
Anyway, I am full of joy now. Kudos to the engineers at Analog Devices who had designed the AD797, who had invented the stop clock mechanism and who had invented the shunt caps mechanism. And thank you miro and diyinhk for designing these nice ad1862dacs!
Cheers😀😀😀
I was excited to see that shunt capacitors can be used to absorb clock energy and current steps and decided to try it. I didn't have 2000pf caps on hand. The closest I had were 0.01uf C0G and 1000pf Wima film. I tried the 0.01 uf first, but I heard oscillations. Then I put a 1000pf right across pins 2 and 3 of AD797brz and was pleasantly surprised to hear a nice and big difference in sound!
Hello Hidy,
Did you try 100R resistor as well? According to the datasheet:
"Optimum flatness and stability at noise gains >1 sometimes require
a small capacitor (CL) connected across the feedback resistor (R1 of
Figure 39). Table 6 includes recommended values of CL for several
gains. In general, when R2 is greater than 100 Ω and CL is greater
than 33 pF, a 100 Ω resistor should be placed in series with CL."
We are using 47pF in CL position, so maybe we should try 100R as well.
Regards
Laszlo
Would it be possible that the stop clock mechanism is the main reason behind the difference in sound? I studied the data sheets of the various opamps I have until I found this:
View attachment 998699
Your attachment is not reachable... . But may be this is only the AD797 datasheet.
Last edited:
Sorry, the attachment was somehow gone.
In case it goes missing again, let it be known that my attachment is a screenshot of the schematics on page 17 of the official data sheet of AD797 (rev.k). It is also on other versions of the AD797 data sheets, but on a different page.
There they are again:

View attachment AD797.pdf
I haven't tried the 100r yet, because I am using 20pf now, a value low enough to go without the 100R. But I will try the 82pf plus 100R set up once I have a 82pf available.
Enjoy!
In case it goes missing again, let it be known that my attachment is a screenshot of the schematics on page 17 of the official data sheet of AD797 (rev.k). It is also on other versions of the AD797 data sheets, but on a different page.
There they are again:

View attachment AD797.pdf
I haven't tried the 100r yet, because I am using 20pf now, a value low enough to go without the 100R. But I will try the 82pf plus 100R set up once I have a 82pf available.
Enjoy!
Actually, I may have to recalculate the value for 82pf, because I am using 1.5k resistor now, not the 3k resistor as specified in the schematics.
at asilker : you don't mind, it's often from what I understood about the package the factory is providing to the distributor : cuted band, sealed package, etc.
Better to focus on a good brand like JRC or Fairchild and buy the cheaper package on your distributor site. I will give the LM317/337 pcb a try, these chips are excellent, long known and always avaliable.
Better to focus on a good brand like JRC or Fairchild and buy the cheaper package on your distributor site. I will give the LM317/337 pcb a try, these chips are excellent, long known and always avaliable.
An equivalent value of 164pf is needed to bypass a 1.5k resistor. I will just try 120pf as that is the maximum allowed in the data sheet.
@Hidy,
A cap to ground at the inverting input of an opamp may cause frequency peaking or instability. Good to be careful with that if it is deemed necessary.
Also, if you find the dac doesn't sound as good as the diyinhk version, well, that would concern me. I am not a big fan of diyinhk designs so far. Haven't seen one that couldn't use some improvement itself.
A cap to ground at the inverting input of an opamp may cause frequency peaking or instability. Good to be careful with that if it is deemed necessary.
Also, if you find the dac doesn't sound as good as the diyinhk version, well, that would concern me. I am not a big fan of diyinhk designs so far. Haven't seen one that couldn't use some improvement itself.
Last edited:
Thanks.
Using a bigger value of 0.01uf did actually cause oscillation.
I wasn't comparing the sound quality to the default version of diyinhk board. I have added 10 x 100uf muse caps to the power supply pins, which has improved its sound quite a bit.
Using a bigger value of 0.01uf did actually cause oscillation.
I wasn't comparing the sound quality to the default version of diyinhk board. I have added 10 x 100uf muse caps to the power supply pins, which has improved its sound quite a bit.
@Hidy,
I am not a big fan of diyinhk designs so far. Haven't seen one that couldn't usce some improvement itself.
Could you be more specific as to where improvement is possible with the diyinhk AD1862 dac board? I'd like to try if possible.
By the way, have you personally heard how the diyinhk AD1862 sound?
Thank you.
- Home
- Source & Line
- Digital Line Level
- DAC AD1862: Almost THT, I2S input, NOS, R-2R