No, just let it NOS, but work a little on the passive parts both caps & resistors. And of course the PS, putt more bad esr lytic caps in front of the 78/79 regs if used - and a faster before it of course.
I hope to tell more the time I need to command from Miro1360 gerbers and a pair of AD1862 I bought to compare t my genuine Rochester.
The design is simple but good thing is through hole was chosen, if not error made on the pcb, there is no reason it sounds bad, and I have no doubt it can eat all the ESS chips on Red book materials. A good TDA1541A already does that.
I hope to tell more the time I need to command from Miro1360 gerbers and a pair of AD1862 I bought to compare t my genuine Rochester.
The design is simple but good thing is through hole was chosen, if not error made on the pcb, there is no reason it sounds bad, and I have no doubt it can eat all the ESS chips on Red book materials. A good TDA1541A already does that.
Last edited:
Abraxalito is a very knowledgeable guy, but for some reason he wants to ignore some details that are necessary for a dac to sound better, especially with these old chips and I talk about caps and resistor and power supplies tipology, according my experience only. We fight each others unluckily because he doesn't want to see or try to follow his own philosophy that a bit strange to me when simple tips may work, smd x7r in the analog path is a no go for me, sorry. But he is a reliable technician I believe, not as open I believed, but reliable and knowledgeable. You only know by try the both boards. But still frankly think the Rogic TDA1541A digs them all. I wish John from ECdesigns made a diy tda1541a board as well !
The one Abrax is using is imho under the 1541A, and the fact he uses multiples of tda 1387 certainly adds problems of concistency as stated John from EC designs when he also tried the technic of stacked dac chips for increasing outputt curent !
Anyway, nothing bad with the ad1862, best pcm chip ever with the tda1541A, I believed the tda1545a had a better treble than the 1541A, but at the end it's all about the design, power supply, layout, BOM and marriage with the rest of your hifi devices. For Red Book they beat them all AK or ESS, YMMV as usual.
Frankly, ad1862 for an headphone system is a RollsRoyce choice, coooool ! I just will add some care on the digital front end not to waste th ecool pcb. (and a care to the supplies as usual...c-core advised for the smooth but the snap yet)
The one Abrax is using is imho under the 1541A, and the fact he uses multiples of tda 1387 certainly adds problems of concistency as stated John from EC designs when he also tried the technic of stacked dac chips for increasing outputt curent !
Anyway, nothing bad with the ad1862, best pcm chip ever with the tda1541A, I believed the tda1545a had a better treble than the 1541A, but at the end it's all about the design, power supply, layout, BOM and marriage with the rest of your hifi devices. For Red Book they beat them all AK or ESS, YMMV as usual.
Frankly, ad1862 for an headphone system is a RollsRoyce choice, coooool ! I just will add some care on the digital front end not to waste th ecool pcb. (and a care to the supplies as usual...c-core advised for the smooth but the snap yet)
Last edited:
Hi,
In comparison to this:
AUDIOPHONICS DAC I-Sabre ES9038Q2M KALI EDITION Raspberry Pi / I2S & SPDIF / PCM DSD - Audiophonics
Review and Measurements of AUDIOPHONICS DAC I-Sabre ES9038Q2M | Audio Science Review (ASR) Forum
Do you think your design is better?
Can I connect it directly to a power amp?
What is the output vrms ?
What is SNR value? THD?
What is the output impedance?
Thanks
In comparison to this:
AUDIOPHONICS DAC I-Sabre ES9038Q2M KALI EDITION Raspberry Pi / I2S & SPDIF / PCM DSD - Audiophonics
Review and Measurements of AUDIOPHONICS DAC I-Sabre ES9038Q2M | Audio Science Review (ASR) Forum
Do you think your design is better?
Can I connect it directly to a power amp?
What is the output vrms ?
What is SNR value? THD?
What is the output impedance?
Thanks
Just a couple of adjustments, added a low 1k+4n7 filter, made the sound a whole lot worse with very little benefit.
The IV is probably most sensitive to out of band noise (or maybe filtering is just not that important?).
Changed supply, previously TPS7A for digital supply and SR reg for OPA and analogue supply, now super regulator powering everything,
Sounds cleaner/smoother this way.
isolation of digital and analogue supply not as important as power quality?
The IV is probably most sensitive to out of band noise (or maybe filtering is just not that important?).
Changed supply, previously TPS7A for digital supply and SR reg for OPA and analogue supply, now super regulator powering everything,
Sounds cleaner/smoother this way.
isolation of digital and analogue supply not as important as power quality?
slightly yes, a warmer chip can sound softer
@surfparadise
I don't know the design or the sound, it's a completely different system (delta-sigma, where every detail matters) ..
... yes you can connect directly to PA, output vrms can be set by the resistor on I/V (R8, R9)
... SNR and THD depends on used opamp (otherwise it will be similiar to values from AD1862 datasheet)
... set the output impedance with R10, R11
1k+4n7 is not good choice, avoid that capacitor on the output (it also depends on the type, it's parallel on the analog way, it affects the sound), if necessary, use a smaller value with correct R combination (or LC filters from Abraxalito) ...you do not need an output filter for AD1862 (something is integrated inside)
MC78/79 regulators from onSemi are fine with a gently sound
separating power supplies has benefit, that is why AD1862 has +-5V for digital and +-12V for analog ... separating them for each chip separately no longer has such a great benefit, only a little
AD1862 has Zout around 2kohm so one of my filters would work yes. The filter does have downsides - for example the noise of the I/V stage becomes much more critical due to the noise gain being increased by the filter's capacitor to GND. So its not necessarily a plug and play solution - don't let that discourage you though
Using bisesik transformer for IV actually sounds amazing with the AD1862. It's the first time preferring it over OPA IV in all the different DACs it was tested with, because for first time (miraculously) it is more resolving than OPA IV, not just more pleasant/natural sounding.
I think the main reason is unrelated to the DAC, I am using a composite HP amp with an op amp on the input now instead of the simple discrete designs without global feedback used on all other occasions, so input impedance is many many times higher now.
Funny since the the trafo was intended to eliminate all op amps and turns out they can have great synergy when used in the right places.
adding a filter (5nf to ground) before the OPA IV had a noticeable benefit, more liquid, less haze, but did not really change the overall flat/grey character of the op amp IV.
Maybe it is not fair to compare a single cap to proper reconstruction filter but I cant see it transforming the sound (pun intended) like the trafo. Abraxalito, you have some actual experience with that, so I would like to hear what u think.
I think the main reason is unrelated to the DAC, I am using a composite HP amp with an op amp on the input now instead of the simple discrete designs without global feedback used on all other occasions, so input impedance is many many times higher now.
Funny since the the trafo was intended to eliminate all op amps and turns out they can have great synergy when used in the right places.
adding a filter (5nf to ground) before the OPA IV had a noticeable benefit, more liquid, less haze, but did not really change the overall flat/grey character of the op amp IV.
Maybe it is not fair to compare a single cap to proper reconstruction filter but I cant see it transforming the sound (pun intended) like the trafo. Abraxalito, you have some actual experience with that, so I would like to hear what u think.
My boards arrived today, Yahoo!! I’ve got all the parts ready to build except........ the AD1862’s
I’d like to give Bisesik’s transformers a try also
Attachments
My boards arrived today, Yahoo!! I’ve got all the parts ready to build except........ the AD1862’s
I’d like to give Bisesik’s transformers a try also
hold on, 5v for digital supply?
I thought it was a typo when miro wrote +/-5V for AD1862 digital supply (ad1865 does use 5V).
The datasheet recommends +/-12V for the digital supply, and the diyinhk PCB is marked as +/-12V.
@Vunce, very nice
@laserscrape use +/-5V for digital, it is not a typo, really use +-5V for digital and +-12V for analog
Is there technical reason for this or from listening you concluded that?
edit: quick search showed this has been asked before, so the digital circuits operate faster at lower voltage, interesting.
I guess then they suggest 12v for convienence/cost saving (sharing the analogue supply) rather than performance
Last edited:
it is advice from one very experienced engineer, named @dohnalik
I personally don't know the exact details behind it, but that shift can theoretically help to better separate the digital part ... critical part of an R-2R is reference, in AD1862 is powered from analog supply, so there is no reason to +-12V for digital
I personally don't know the exact details behind it, but that shift can theoretically help to better separate the digital part ... critical part of an R-2R is reference, in AD1862 is powered from analog supply, so there is no reason to +-12V for digital
My boards arrived today, Yahoo!! I’ve got all the parts ready to build except........ the AD1862’s
I’d like to give Bisesik’s transformers a try also
Vunce, I found a small error
electrolyte in negative supply for I/V is rotated wrong, in the schematic and on the PCB: C27, C32 ... please rotate it correctly, it is very important