DAC AD1862: Almost THT, I2S input, NOS, R-2R

Disabled Account
Joined 2019
thanks for the BOM as I use Mouser :)


the oap NJM4556 is a dual, but a special need for an other pcb, it will not suit on the Miro1360's.
It is advised to stay dip package-I will try the LM and directly solder these high speed video amplifier amp.


I'm going to try also on soic adapters with extra decoupling these ones : opa209 or opa211; opa810 ; ada4610-1 ; AD8610 but if someone experienced is saying there are better choices of course.
 
Last edited:
thanks for the BOM as I use Mouser :)


the oap NJM4556 is a dual, but a special need for an other pcb, it will not suit on the Miro1360's.
It is advised to stay dip package-I will try the LM and directly solder these high speed video amplifier amp.


I'm going to try also on soic adapters with extra decoupling these ones : opa209 or opa211; opa810 ; ada4610-1 ; AD8610 but if someone experienced is saying there are better choices of course.

Yes... there are some parts to delete... wrong file sorry 😐
 
Disabled Account
Joined 2019
I'm the first to mix between sing/dual channel/ input stage tech cmos, fet,bipolar, etc.. too much choice nowadays at many prices ! opa627 is just fine but too much expensive...and I will have certainly a use for the non selected oaps at the end in the Miro's for others devices to be tweaked for me or for friends...
 
I found quite difficult to find on mouser the 2.5mm pitch / 5mm diameter audio grade Nichicon capacitor. (6.3mm could be ok?)


If you planning to use ultra low noise LDOs these big electroyltics caps will only deteriorate the sound of these regs, you only need the small 100n decoupling caps and it will sound best this way.
edit: better yet, replace 100n with larger value film cap (1uf or so).

Ultrabib is also negatively effected in the same way, they didnt have much of an effect on the Super regulator but nor did they have much benefit.
 
Last edited:
Save money and buy 1656..:)
It's dual - be creative!!

I have tried, not in I/V but dac output position:
opa1612(=1611, opa211) ; (anyway a good chip) ;
Obviously 5532, 5534; (not bad at all)
4562, 49710, 49720
Opa1662; opa1652;
Opa1642 - good impression!
Ad8620- uhm... (+expensive..)
ADA4627 (a, b) - very fine
OPA1688 - very good very lively

Other places: ADA4898, - nice, OPA827 - standard choice for me; OPA828 : presently used in MyRef;

OPA1656 - presently doing hard lifting I/V job in the 4499 dac

Ciao, George

Ah forgot AD797, OPA627 yeah ok not bad
 
even not a 1 to 2 uF near the 0.1 uF with some speedy oaps ?

Actually yes, the 100n are the box type film caps I just noticed, I would ditch the 100n and use ~1 uf polyesters (or whatever largest value that will fit in that footprint).

Ideally the reg output should be so close to the load that you only need the 100n but 1u is good compromise for external reg boards.
 
Last edited:
Ive heard the opposite, at least the lm317 can oscillate with too low ESR output cap and LT3045 and TPS7A4700 require low ESR ceramic output caps.

But you are talking about the output caps for the reg, which should be as close to the regulator output as possible for stability and will already be part of any regulator board you buy (hopefully). The decoupling caps are seperate and should be as close to the load as possible.
 
Disabled Account
Joined 2019
Save money and buy 1656..:)
It's dual - be creative!!

I have tried, not in I/V but dac output position:
opa1612(=1611, opa211) ; (anyway a good chip) ;
Obviously 5532, 5534; (not bad at all)
4562, 49710, 49720
Opa1662; opa1652;
Opa1642 - good impression!
Ad8620- uhm... (+expensive..)
ADA4627 (a, b) - very fine
OPA1688 - very good very lively

Other places: ADA4898, - nice, OPA827 - standard choice for me; OPA828 : presently used in MyRef;

OPA1656 - presently doing hard lifting I/V job in the 4499 dac

Ciao, George

Ah forgot AD797, OPA627 yeah ok not bad


Hi, the last two I avoid because the priiiiice :yuck:


I'm focussed on the I/V purpose but have you tried the opa2228 as a buffer in front of the opa1656 as an I/V ?
opa 1656 .... creative, creative, lifting leads, wirering, adaptator hack ... thi drive me 20 years back... pouah :clown: . I'v been creative by "selecting" the AD8610 instead (the single version of the 8620)... But your "uhm" is scarring me :yikes:...:(


You mean you use the opa820 in buffer in front of the opa1656 when you say "MyRef" - I understand your AK4499 from the guy than write japanese ?


I'm very attracted by the opa810 as well cause it's single channel- opa828 is dual ... the creative factor again :D-


Thanks for that nice list KJoseph. ... we should try the opa227/228 instead the opa861 as a buffer in the aya... but it needs an adaptator because the board layout ! Well off topic here anyway.
 
Disabled Account
Joined 2019
Ive heard the opposite, at least the lm317 can oscillate with too low ESR output cap and LT3045 and TPS7A4700 require low ESR ceramic output caps.

But you are talking about the output caps for the reg, which should be as close to the regulator output as possible for stability and will already be part of any regulator board you buy (hopefully). The decoupling caps are seperate and should be as close to the load as possible.


Yes in case of the 78xx & 79xx and the LM : better a 100 to 220 uf cap at its output with a bad esr...but a 47uF with a good ESR near the load if the bad cap at the outputt pin of the 78/79xx with or without the 0.1 uF -with the buffers oaps-never harmed my ears... Ah, and a good ESR cap at the opposit at the inputt position of that reg ; Pan FC, Pan FR, even Elna SII works not bad here.



Certainly different in I/V purpose maybe. Anyway Miro is pretty right, all these modern oaps have a good noise/ripple management... so I will maybe go towards Nazar fellow discrete reg LM317 based, or a single 78/79xx reg - but the one I have must be NJR if I remember-


One can imagine also to putt no reg on the 5V but instead two LiPoFe4 in serie for a 6.6 V... could be funny while there will be always the noise of the I2S front end !
You know what I think about the ceramic at the outputt of the Ti reg ;) throw them all at the garbadge and change them for something else than X7r ! :D I use it for the clock board, fifo, front end of my ref DAC and tunned that chip extensivly around Mavrica member pcbs for it.
 
Last edited:
Disabled Account
Joined 2019
Actually yes, the 100n are the box type film caps I just noticed, I would ditch the 100n and use ~1 uf polyesters (or whatever largest value that will fit in that footprint).

Ideally the reg output should be so close to the load that you only need the 100n but 1u is good compromise for external reg boards.


For instance a good 1 uF MKT Wima with a 2.5 mm pitch :)
 
create your own adapter or test this one for errors :D
use online-gerberviewer to check the files
 

Attachments

  • dual-to-single-opamp_2020-07-17.zip
    30.7 KB · Views: 85
  • dual-to-single-opamp.jpg
    dual-to-single-opamp.jpg
    47.4 KB · Views: 398
  • dual-to-single-opamp_PCB.jpg
    dual-to-single-opamp_PCB.jpg
    75.1 KB · Views: 407