Thanks Miro,
I jumped the train of pcb layout with the tda1541A long time ago but never has the time to try myself this hard task. It is indeed very hard to sorf out the theory and what is effective in relation to the dac chip, what is hearable or not, etc. And yes, that is my understanding than with pcm chips 4 layers are not always mandatory, and a trade off VS two layers. 4 layers can have issues too as crosstalks between layers, etc.
Anyway the ad1862 and your good work and also the one of Painkiller (again North European Viking team, I send him one of youf pcb as he sent me one of his discrete i/v one long time ago) member on that chip gives me the envy to try Kicad to try to do something with my fingers.
I have finished the double test today about I/V resistors with oaps. Just a little contribution to your work as many members did already. On 12 resistors one performs really above in my subjective setup. Unfornatully the Rhopoint I have has not the legs long enough to be tested. The surprise is they all have differences, sometimes subtles sometimes not. I sorted them out in three class:
1. I can not live with it, it wastes the DAC.
2. I can live with but non optimal according my expectations
3. Smell of High-end, it put the DAC in another class I can live with and that shows your work gives margins to passive parts, here people should try imho if they have not already a proven good sounding resistor as a Rhopoint or hand wirering Ayrton Perry a la John from ECdesigns.
I beginn this evening after the movie with family.
I jumped the train of pcb layout with the tda1541A long time ago but never has the time to try myself this hard task. It is indeed very hard to sorf out the theory and what is effective in relation to the dac chip, what is hearable or not, etc. And yes, that is my understanding than with pcm chips 4 layers are not always mandatory, and a trade off VS two layers. 4 layers can have issues too as crosstalks between layers, etc.
Anyway the ad1862 and your good work and also the one of Painkiller (again North European Viking team, I send him one of youf pcb as he sent me one of his discrete i/v one long time ago) member on that chip gives me the envy to try Kicad to try to do something with my fingers.
I have finished the double test today about I/V resistors with oaps. Just a little contribution to your work as many members did already. On 12 resistors one performs really above in my subjective setup. Unfornatully the Rhopoint I have has not the legs long enough to be tested. The surprise is they all have differences, sometimes subtles sometimes not. I sorted them out in three class:
1. I can not live with it, it wastes the DAC.
2. I can live with but non optimal according my expectations
3. Smell of High-end, it put the DAC in another class I can live with and that shows your work gives margins to passive parts, here people should try imho if they have not already a proven good sounding resistor as a Rhopoint or hand wirering Ayrton Perry a la John from ECdesigns.
I beginn this evening after the movie with family.
Last edited:
So I tested several 1k5 I/V resistors with both op1656 Jfet input on @Vunce adaptator w/o the thick DIL8 adaptator and the op1611 which is said to be one of the best Bjt input oap today. This is of course subjective. the few ohms difference from a resistor to another didn't give particular results. the volume pot level of the pre was unchanged. resistor were plugged but not soldered as the oaps pcb adaptators on my test box : Miro pcb with uf-l + Wave I/O on non isolated output. Powered with a TSA7 reg on a separated secondary R-Core traffo. Main PSU is the PSU2. Caps and resistors, active parts were sourced at Mouser or official distributors. Please do not input about the box DAC layout, it is just a Testing box for experiment with parts, front-ends, regs, etc... I do not care here of short wires with the front end or the power supplies.
Resistors are from the left to the rigth:
-Row1: 1/4W and 1/2W carbon film, generic
-Row2: Allen Bradley 1/4W (Hifi Collective) - Vishay MILLS05 5W - Ayrton-Perry wirewound
-Row3: Bulk Vishay LTO-030F (notice: 1K ohms) 30W.
-Row4: Amtrans Carbon Composition (Hifi Collective) - Vishay Wirewound ACO 1W (green cement)
-Row5: Vishay SFR16 1/8W - Vishay wirewound RS01A 1W (black cement) - Vishay metal film PTF56 1/8W (dark blue)
-Row6: CCF07 1/4W - vishay CPF2 wirewound (brown clear cement) 2W - Rhopoint 8G16D (my ref on my main DAC)
Resistors are from the left to the rigth:
-Row1: 1/4W and 1/2W carbon film, generic
-Row2: Allen Bradley 1/4W (Hifi Collective) - Vishay MILLS05 5W - Ayrton-Perry wirewound
-Row3: Bulk Vishay LTO-030F (notice: 1K ohms) 30W.
-Row4: Amtrans Carbon Composition (Hifi Collective) - Vishay Wirewound ACO 1W (green cement)
-Row5: Vishay SFR16 1/8W - Vishay wirewound RS01A 1W (black cement) - Vishay metal film PTF56 1/8W (dark blue)
-Row6: CCF07 1/4W - vishay CPF2 wirewound (brown clear cement) 2W - Rhopoint 8G16D (my ref on my main DAC)
Attachments
There were big difference enough to sort them out in 3 class (look at the post above). Totally subjective, but I try to explain briefly what was heard for you to adapt in relation to your own setup. My own language, simplified, sorry for that. There is one clearly above and the worst are in the class that is imho better to avoid. Notice and it is important: resistors are almost all new, so no breaking in. It may changes the game. Allen Bradley certainly needs some (?).
Row 1 : carbon film generic 1/4W & 1/2W were a good surprise and among the best. perhaps the good inductive behavior of carbon comp for that task ? I noted: clear, better on voice that the Vishay black cement wirewound. Good on strings. Bass well sized. good texture while not meaty (clear). Not colored. Non fatiguing. Very advised, neutral. Subjective note 3.5/5 - Class 2
Row2 : Allen Bradley carbon composition, a liitle clearer than the generic carbon. Attacks (dynamic rise= transcient?)) are drilling a little mùy ears when in the upper medium and low treble. As the carbon, they are detailed in this I/V task. Bass are snappy (much more than basic metal film, see later). While they are nice i found them a littlr fatiguing because this sorta of H" on dynamic charges (transcients). Subjective note 3/5 - class2 (the simple generic carbon are better w/o doubt but maybe burning in is lacking there, go figure).
Row2 : Mills MRA05 - 5W, wirewound Ayrton Perry layout. Often seen in loudspeaker filters, I had some on hands for the test. My expectations were high with them but hard is the fall. Bass are subjectivly short and weaker. Bad decay. If highs are not particulary bad low mid lack of texture. A little flat sounding. Colored, notes are somewhat simplified. Subjective note 2/5- Class I (=better to avoid, I can not live with, it waste the design).
Row3 : Bulk Vishat LTO-030F : a not too much expensive bulk I had to try. It is a strange resistor here. texture on indian tablas and bass strings are among the best. Females voices are not nice and drilling on attacks while below all is a little muddy. It lacks bass (weak or shortened) that are too round as well. It is a little flat without meat/weigth while dark background, there is also a veil feeling. Artificial, a very strange resistor, somewhat colored. If they are better that the simple metal films, they lack of clearness. Subjective note 2.5/5 - Class 1 (better to avoid, I can not live with).
Another way to read it : if your hifi equilibrium is different from mine at home, defaults could be qualities for you... but I doubt ! My loudspeaker is a 3 ways with metal mid 125 hz to 2600 hz . aluminium treble - sealed bass 35 hz (f3) to 125 hz.
Here the simple carbon films at 2 cents are the clear winner of this first round.
Row 1 : carbon film generic 1/4W & 1/2W were a good surprise and among the best. perhaps the good inductive behavior of carbon comp for that task ? I noted: clear, better on voice that the Vishay black cement wirewound. Good on strings. Bass well sized. good texture while not meaty (clear). Not colored. Non fatiguing. Very advised, neutral. Subjective note 3.5/5 - Class 2
Row2 : Allen Bradley carbon composition, a liitle clearer than the generic carbon. Attacks (dynamic rise= transcient?)) are drilling a little mùy ears when in the upper medium and low treble. As the carbon, they are detailed in this I/V task. Bass are snappy (much more than basic metal film, see later). While they are nice i found them a littlr fatiguing because this sorta of H" on dynamic charges (transcients). Subjective note 3/5 - class2 (the simple generic carbon are better w/o doubt but maybe burning in is lacking there, go figure).
Row2 : Mills MRA05 - 5W, wirewound Ayrton Perry layout. Often seen in loudspeaker filters, I had some on hands for the test. My expectations were high with them but hard is the fall. Bass are subjectivly short and weaker. Bad decay. If highs are not particulary bad low mid lack of texture. A little flat sounding. Colored, notes are somewhat simplified. Subjective note 2/5- Class I (=better to avoid, I can not live with, it waste the design).
Row3 : Bulk Vishat LTO-030F : a not too much expensive bulk I had to try. It is a strange resistor here. texture on indian tablas and bass strings are among the best. Females voices are not nice and drilling on attacks while below all is a little muddy. It lacks bass (weak or shortened) that are too round as well. It is a little flat without meat/weigth while dark background, there is also a veil feeling. Artificial, a very strange resistor, somewhat colored. If they are better that the simple metal films, they lack of clearness. Subjective note 2.5/5 - Class 1 (better to avoid, I can not live with).
Another way to read it : if your hifi equilibrium is different from mine at home, defaults could be qualities for you... but I doubt ! My loudspeaker is a 3 ways with metal mid 125 hz to 2600 hz . aluminium treble - sealed bass 35 hz (f3) to 125 hz.
Here the simple carbon films at 2 cents are the clear winner of this first round.
Row 4: the expensive Amtrans carbon comp. Nice, soft, borring. Short on bass. Very different fromthe Allen Bradley that are more clear. A sorta of analogic sound in the bad way because not too much detailled and lacking of life. Maybe it will fit a compression driver with metal diag. and big bass ? Voice are nice but a little veil. you have climbing trebles, that's the one you need. lacking of life/juice and snap in the upper bass. Subjective note : 3/5 - class 2 (I could live with if I was listening Ramstein or electronic but not jazz or classic or pop- you want a relax life they are for you. Although they have not this drills on attacks like the Allen Bradley, but this last is better everywhere)
Row 4 : Vishay wirewound green cement. Wow (the winner of the bench). there is something very rigth here. The most detailed and best decay while not fatiguing. Not dark background, clear but without artificial ligthness. Neutral while musical (imho the only one with another one here). Voices are very good : male and female have the average texture and weigth. Dynamic range is good. Claps in venues are rigthly sized, a difficult test. The strings are the best here, close to my ref DAC that has a more complex front-end and analog stage (TDA1541A S1 and late Taiwann). There texture is good and detailled with decay and texture. Bass strings, idem. Piano : check. The background is quiet. Non fatiguing music, analogic in the good side of the word while detailled (LP killer). Notes have the average weigth. You stop to ask question. Subjective note 4.5/5 . Class 3 . Winner of the test (smell of high end I can/want live with). You should try it if you have an aop and a simple metal film resistor: it becomes something another dac and what we are figthing for with boring diy experiments !
Row5: Vishay SFR16 : Flee! Veiled, colored, muddy bass and too much round everywhere. Artificial and fatiguing high mid and treble. It wastes the DAC. people say it is less noisy than a generic carbon, but here all the opposit. It makes noise and no music. Very short decay if I read my notes; climbing trebles; lack of bass (short)cheap hifi sound, fatiguing, especially with voices. Subjective note 2/5- class I (waste all the design, you made efforts elswhere but no luck you putted a metal film here and it sounds bad ! Now you know).
Row 5 : Vishay wirewound black. It doesn't sound like the other cement of the group. Very clear in the sense it adds ligth. This is the one if your hifi is very dark. Although it is not brigth. It is detailled but over simplified VS the green wirewound. Lack a little of weigth and attacks on voices can be sometimes a little fatiguing, but again it climbs without particular brigthness. Good to wake up a muddy too weigthy system. If bass are invading your listening room, this I/V resistor is a good choice according my taste and setup. No dark nore quiet background. Maybe a little too much for jazz while still good for Classic and Rock. Good details, on par with the first row carbon here. A little more thin and sun ligth on the notes... less neutral than the carbon. Not sounding at all like the green wirewound. But a good one. subjective note 3/5. class 2
Row 5 : Vishay metal film precision 5 or 10 PPM (the dark blue one). Just a little better than the SFR16. same feeling than the RN50 from Dale not tested here but i always avoid as it is mellows the sound on my system. OK no breaking in, after two hours and a re test, they go back to the Mouser bag for another futur task. Most of what is writted for the SFr16 above is more or less the same here. subjective note 2/5. Class I (better to avoid, it wastes the system)
Row 6: CCF07 metal film (the upper red of the row) : A little above the other metal films but still bad. Better to avoid. Subjective note 2.5/5 . Class I (better to avoid, it wastes the DAC). Seems clear to me the metal films are not well suited for I/V with aops, at least without breaking in, and in my system. But the others are way better I doubt you had better chance with this category (Class I). I mean, buy another resistor than such metal film as soon as possible !
Row 6: the brown CPF2 (2W) wirewound, cement again. Surprise as different from the two other cements (the black and the green). Very clear like the black one but with less sun (I prefer). Very neutral and non fatiguing like the green while less mat. A little less good texture than the green (difference is not subtle though). With Nomex like cone it may be a winner. but the green are more musical. This brown one has a bigger soundstage maybe due to the ligth, maybe more life while non fatiguing yet. But it doesn't pass the test against the green wirewound if you listen all music styles and acoustic instruments and voices. It could be the good one if you only listen rock and electronic maybe over the black wirewound. Subjective note 4/5. Class 3 (smell of high end and second best). Try it as benchmark wiith the green ?
Row 6 : the famous Rhopoint wirewound Ayrton Perry style. Lol : non tested, leads too short cause too much bulky body for Miro pcb layout. This is the resistor I use in my reference dac, but the other green gave me the envy to challende the Rhopoint.
While it is all subjective, I had fun with this experiment. Difference are big enough to make the DAC sounding from I do not want to listen and I am happy to have it cause it is musical, like with different oaps and caps or different power supplies. On the test pcb most of the caps are Panasonic FC, Pan FR on the 5V of the chip. Acrylic cap as HF decoupling for the dac chip. Vishay SAL and Pan FC cap as dac chip filter directly soldered on the adaptator pins. resistors on the pcb but this I/V resistor are Yageo 1% that are less noisy than the dale RN55 often seen.
ref of the green Vishay is Mouser 594-AC01W1K500J : warmly recomended for the official BOM
! A cheap mod/tweak that should make smile all the guys with basic metal films I/V resitors. https://www.mouser.fr/ProductDetail/Vishay-Draloric/AC01000001501JA100?qs=R4/Aj8xQbdWbMGxRCrZj/w==
Notice : all the cements have also a brother with Ayrton Perry winding I have not tested. Funny enough : a simple carbon film is among the best ! Resistors again were not burning in. Of course it is subjective. but surely, other people in other threads needing a cheap I/V resistor should try...
I should had tested also a Sussumu smd thin film RR resistors on the cap pads of the I/V layout at the bottom of the pcb. Better inductance is said to be important IIRC in the voltage feedback loop of oaps. But the different length of leads with the through holes resistors gave not concistent results, good or bad, nore the smal ohms variation of the 1k5 ohms between resistors.
.
Row 4 : Vishay wirewound green cement. Wow (the winner of the bench). there is something very rigth here. The most detailed and best decay while not fatiguing. Not dark background, clear but without artificial ligthness. Neutral while musical (imho the only one with another one here). Voices are very good : male and female have the average texture and weigth. Dynamic range is good. Claps in venues are rigthly sized, a difficult test. The strings are the best here, close to my ref DAC that has a more complex front-end and analog stage (TDA1541A S1 and late Taiwann). There texture is good and detailled with decay and texture. Bass strings, idem. Piano : check. The background is quiet. Non fatiguing music, analogic in the good side of the word while detailled (LP killer). Notes have the average weigth. You stop to ask question. Subjective note 4.5/5 . Class 3 . Winner of the test (smell of high end I can/want live with). You should try it if you have an aop and a simple metal film resistor: it becomes something another dac and what we are figthing for with boring diy experiments !
Row5: Vishay SFR16 : Flee! Veiled, colored, muddy bass and too much round everywhere. Artificial and fatiguing high mid and treble. It wastes the DAC. people say it is less noisy than a generic carbon, but here all the opposit. It makes noise and no music. Very short decay if I read my notes; climbing trebles; lack of bass (short)cheap hifi sound, fatiguing, especially with voices. Subjective note 2/5- class I (waste all the design, you made efforts elswhere but no luck you putted a metal film here and it sounds bad ! Now you know).
Row 5 : Vishay wirewound black. It doesn't sound like the other cement of the group. Very clear in the sense it adds ligth. This is the one if your hifi is very dark. Although it is not brigth. It is detailled but over simplified VS the green wirewound. Lack a little of weigth and attacks on voices can be sometimes a little fatiguing, but again it climbs without particular brigthness. Good to wake up a muddy too weigthy system. If bass are invading your listening room, this I/V resistor is a good choice according my taste and setup. No dark nore quiet background. Maybe a little too much for jazz while still good for Classic and Rock. Good details, on par with the first row carbon here. A little more thin and sun ligth on the notes... less neutral than the carbon. Not sounding at all like the green wirewound. But a good one. subjective note 3/5. class 2
Row 5 : Vishay metal film precision 5 or 10 PPM (the dark blue one). Just a little better than the SFR16. same feeling than the RN50 from Dale not tested here but i always avoid as it is mellows the sound on my system. OK no breaking in, after two hours and a re test, they go back to the Mouser bag for another futur task. Most of what is writted for the SFr16 above is more or less the same here. subjective note 2/5. Class I (better to avoid, it wastes the system)
Row 6: CCF07 metal film (the upper red of the row) : A little above the other metal films but still bad. Better to avoid. Subjective note 2.5/5 . Class I (better to avoid, it wastes the DAC). Seems clear to me the metal films are not well suited for I/V with aops, at least without breaking in, and in my system. But the others are way better I doubt you had better chance with this category (Class I). I mean, buy another resistor than such metal film as soon as possible !
Row 6: the brown CPF2 (2W) wirewound, cement again. Surprise as different from the two other cements (the black and the green). Very clear like the black one but with less sun (I prefer). Very neutral and non fatiguing like the green while less mat. A little less good texture than the green (difference is not subtle though). With Nomex like cone it may be a winner. but the green are more musical. This brown one has a bigger soundstage maybe due to the ligth, maybe more life while non fatiguing yet. But it doesn't pass the test against the green wirewound if you listen all music styles and acoustic instruments and voices. It could be the good one if you only listen rock and electronic maybe over the black wirewound. Subjective note 4/5. Class 3 (smell of high end and second best). Try it as benchmark wiith the green ?
Row 6 : the famous Rhopoint wirewound Ayrton Perry style. Lol : non tested, leads too short cause too much bulky body for Miro pcb layout. This is the resistor I use in my reference dac, but the other green gave me the envy to challende the Rhopoint.
While it is all subjective, I had fun with this experiment. Difference are big enough to make the DAC sounding from I do not want to listen and I am happy to have it cause it is musical, like with different oaps and caps or different power supplies. On the test pcb most of the caps are Panasonic FC, Pan FR on the 5V of the chip. Acrylic cap as HF decoupling for the dac chip. Vishay SAL and Pan FC cap as dac chip filter directly soldered on the adaptator pins. resistors on the pcb but this I/V resistor are Yageo 1% that are less noisy than the dale RN55 often seen.
ref of the green Vishay is Mouser 594-AC01W1K500J : warmly recomended for the official BOM
! A cheap mod/tweak that should make smile all the guys with basic metal films I/V resitors. https://www.mouser.fr/ProductDetail/Vishay-Draloric/AC01000001501JA100?qs=R4/Aj8xQbdWbMGxRCrZj/w==Notice : all the cements have also a brother with Ayrton Perry winding I have not tested. Funny enough : a simple carbon film is among the best ! Resistors again were not burning in. Of course it is subjective. but surely, other people in other threads needing a cheap I/V resistor should try...
I should had tested also a Sussumu smd thin film RR resistors on the cap pads of the I/V layout at the bottom of the pcb. Better inductance is said to be important IIRC in the voltage feedback loop of oaps. But the different length of leads with the through holes resistors gave not concistent results, good or bad, nore the smal ohms variation of the 1k5 ohms between resistors.
.
Last edited:
I am about to do a jlcpcb order but will not have time to build for a long time.AD1862 Stop-Clock
I did this PCB half year ago and somehow forget to post it here and today found it
It incorporates the Stop-Clock system. What does that mean? Simply: The BCK (clock) is off while no data are present for DAC chip.
I don't know if it has any positive impact on the AD1862 chip, but only 4 digital chips are on the PCB instead of 6
Another "major" change is in the ground plane on the PCB. I divided the GND plane for digital and analog with thick cut lines. This helps eliminate potential digital interference and can have another positive impact on the sound 🤩
BOM example: https://www.mouser.com/ProjectManager/ProjectDetail.aspx?AccessID=c547c05bf3
note: I/V OpAmp in BOM is LM7171, you can buy another one
The PCB is not tested ...
If I have a few free boards to give out, is there anyone in Europe who would agree to build and test and compare with the original layout?
I cannot commit yet - but if this is something that appeals to you please let me know. There might be 2 or 3 boards available.
This is also subject to Miro being happy for me to do that - which I am assuming he is given that he has posted the info here.
I should have one...this one is 1w or 1/2w, I have to check the hifi collective bag. All the test was made without soldering. In my experience and same with caps, the soldering can waste a little VS contact. I surmise also thickness of leads matters in the experience. Try this green Vishay...for the price it is really above the AB I tested. Less than 1$. Notice the expensive Amtrans, Shinko copy was not so good. And with the 3 cement body wirewounds, wattage was a poor indicator of sounding quality. However as usual, the ranking changes from one circuit to another as you know...hence the utlity of trying on the real pcb. I also surmise your ML is short on trebles...while we are not young anymore, lol. Filter not needed with the green vishay... NOS ok without loss in details.
Last edited:
So I moved my last setup from wood back to the enclosure. I was getting some side eye looks from my girlfriend... Still need to sort out wiring. As to avoid drilling and hot glue I had to make 3d printed PCB holders that are adjustable because not all PCBs are made for 10x10mm holes in this tray. The long channel holds an M3 nut that gets screwed from the bottom end. At the end there is M3 nut recess from bottom and you can screw the plastic or brass spacer...
How do I set the bias on D1?
https://www.diyaudio.com/community/...st-tht-i2s-input-nos-r-2r.354078/post-6914268
https://www.diyaudio.com/community/...st-tht-i2s-input-nos-r-2r.354078/post-6914268
AD1862 Stop-Clock
I did this PCB half year ago and somehow forget to post it here and today found it
It incorporates the Stop-Clock system. What does that mean? Simply: The BCK (clock) is off while no data are present for DAC chip.
Miro, great work! I am impressed that you could again reduce the amount of ICs - but this circuit is not the one you posted half a year ago... I did test that one for you, see here.
The idea of stopped clock is to have all digital lines silent when analog output updates, plus a settling time to avoid all influences caused by switching tasks inside the DAC chip. If everything is completed, the chip waits for the next event: the power rails are unburdened, it's time for update of the analog output. I tested the AD1862 by feeding pulses slowly from an MCU to it so I can watch the exact moment it updates IOUT, see attached scope shot.
So if you want to improve sound quality with stopped clock you also need to delay LE, see attached diagram! Something as shown in
Just to show how it can be done here my simulations for similar purposes:
- stopped clock glue logic
- delayed LE for AD1862
- OT: solution for PCM1704
Attachments
Last edited:
No, I use Yageo metal film that are very quiet. I do not like here Dale CMF often seen. Koa speer carbon film and metal film are also liked by some audiophiles since Riken are NLA, or some old Holko metal films. I didn't like the Takman tantal in the area you are asking for. All this should be taken with a grain of salt, it is really hearable not on all hifi, you need to have a good enough level of transparency (all my loudspeakers caps are mkps, interconnects are Oyaiide, Cardas, etc : it is details that changes the game little by little when you work on your devices to get the sound you like in your own particular listening room, hence no universal tweak !)
Acording the DAC and in my hifi system I sometimes like a carbon comp here in relation to what is heard or not with a quiet metal film. On my ref DAC I use a 1/8W Allen Bradley here, the added noise (H3 ?) gives something more real on my system that is near the transparency of a planar loudspeaker (drived with a non colored Chord Amp). Maybe the added noise, a little H3, I do not know ! Sure it is not universal hack and everyone should try and proceed to his own experiences : it is a correction purpose but not an universal ranking test written in the marble. Though I doubt what I called class 1 is good and one should really try the two wirewound of class 3 winners of the bench. technically despite I use no dil_ bulky adaptator but just plug the pcb red oap adaptaor direct on Miro Pcb to do short, there are maybe still HF oscillation cause the 0.1 uF decoupling cap are a little far theorically accordinf Texas Instrument papers. Wirewound act maybe as a little inductor at HF, inductance increase as the frequencies climb, i.e. resistance : you have a low pass filter. Is this that which gives good result ? I dunno, I think it iis not in the ears frequency but who knows. Anyway, with all those wide frequency band oaps nowadays this test may help to thanks Miro design& work by avoiding class I resistors, imho (no burning in needed at least). And frankly at 0.75 euro the green resistor at Mouser, why not try it to the area you ask too
. I have just two so I will not try.Problem of the resistors you ask at the output with carbon film is the value drifting in time with potential mismatch in R values between the channel. but can we hear it? I doubt, here any 50R to 150R resistor is fine to avoid oap break in case of shorts, avoid osicilations at the output due to driving capacitance load and also is an adaptation voltage impedance helper. In my ref DAC, it is maybe the thin leads of this 1/8W resistors that helped, and I also think in the test for I/V area it can be one of the factors. If I had the drawing talent of Miro, I would do something with SMD oap pcb version nearest smd cap traces on the power oaps pins (0.1 uF cap). so cap pad rigth at the power pins, 0805 size for acrylic caps or 0603 for X7R (though I do not like ceramic but it is really subjective, they are the good dielectric for the task and the reduced size is the inductance wanted purpose, though CoG or acrylic are quieter but bulkier : so trade off : choose your poison). Notice all of that is a detail, the pcb is sounding good w/o all this hairs I cut in 4, lol ! Also the gnd pin of the bulky cap 4.7 uF to 47 uF/100 uF should be rigth near the smd pin gnd of the 0.1 uF to reduce loops (Pan FC 100uF/63V has a good inductance with 3.5 mm leads pitch and quiet neutral cap
). I think it can help theorically with the fast video amps as well... I dunno if one can do it with the gerbers given, I find Miro made enough work for the comunauty (thanks for that Miro ) !
Last edited:
AD1862 Stop-Clock
I did this PCB half year ago and somehow forget to post it here and today found it
It incorporates the Stop-Clock system. What does that mean? Simply: The BCK (clock) is off while no data are present for DAC chip.
I don't know if it has any positive impact on the AD1862 chip, but only 4 digital chips are on the PCB instead of 6
Another "major" change is in the ground plane on the PCB. I divided the GND plane for digital and analog with thick cut lines. This helps eliminate potential digital interference and can have another positive impact on the sound 🤩
BOM example: https://www.mouser.com/ProjectManager/ProjectDetail.aspx?AccessID=c547c05bf3
note: I/V OpAmp in BOM is LM7171, you can buy another one
The PCB is not tested ...
Guys, please don't build this PCB, I already discovered major mistake
Sorry for any inconvenience
I like Dale RN55 / CMF55.
They are one of the metal film resistors with the lowest distortion.
Of course one cannot argue against any subjective preferrences.
I also like Vishay S012, Caddock MK232, Resista (NOS), Nikkohm RP-44, Beyschlag MELF, Susumu thin film, ....
I don't use carbon film or composite.
But only my preference.
Cheers,
Patrick
They are one of the metal film resistors with the lowest distortion.
Of course one cannot argue against any subjective preferrences.
I also like Vishay S012, Caddock MK232, Resista (NOS), Nikkohm RP-44, Beyschlag MELF, Susumu thin film, ....
I don't use carbon film or composite.
But only my preference.
Cheers,
Patrick