Not having AD1865 DAC in the collection?
Here is one ... the same PCB changed for AD1865 DAC.
It is tested by @Vunce.
Power supplies are slightly different!
analog part: +5VA, -5VA
digital part: +5VD
Have DIY fun
@miro1360 I was wondering in your PCB design you have used single ground plane for both DGND and AGND, would it improve the noise factor if both ground planes are drawn separately as polygon and then joined at one single point through a link or 0R resistor, the way it is mentioned in the datasheet of PGA231x.
It depends on your PCB. Splitting the GND has to be done under DAC chips. It can be very tricky and if it is done wrong, it will do more trouble than benefit. Solid GND is a better way in this case. Route it so that signal and analog traces are not crossed
But this offset may need to be trimmed also because the opamps do not see exactly the same input at Left & Right by the dac chip outputs .
A bit more extended listening to my setup. Still with LM6171. I have an order from JLPCB on its way with some of Vunces dual to single oak adaptors so I can try my op1565 then.
It is a very nice sounding dac. I am not very good at comparisons. But knowing the costs and efforts gone into my Ian canada dac, the AD1862 is a no brainer. Straight out of the box with nothing particularly 'esoteric' apart from maybe the Salas Ubib on the 12v rails, it sounds amazing. I know I wouldn't have built the ess dac if I'd have built this first. I may shoehorn the ad1862 into the Ian streamer, using the Fifo and Rpi to feed it. It would really simplify the psus in there.
I've noticed one thing which maybe someone can comment on.
I'm using the AK4118 spdif input. On my initial test of this I think I reported it sounded 'better' set to the Left justified setting. It definitely had higher output so maybe the 'better' was confused with 'louder. The only other setting that worked was 'I2S 24b'.
I have noticed now that it does distort heavily in the Left justified setting, as if it is clipping or something. If I lower the volume it helps alleviate it but still sometimes distortion on transients in the music.
So I leave it on I2S 24 now.
I have another couple of boards I intend to populate and swap the dac chips over to compare. I dont wish to be pulling out and refitting the 1862s too often though. I found it tricky installing them in the sockets.
I'll try some different caps and psus on this one. I have built a VRDN regulator and I also have Miros PSU1 boards coming in the JLPCB order. I shall populate these with tps7a4701 and neg version 3pin regulators from Aliexpress. So I shall split the jumper points and feed separately 12v to AD and oaps.
Also I shall test different i2s source. I have a spare Rpi with Ian canada isolator so I can feed the dac isolated i2s (not reclocked or fifo'd here). Shall be interesting.
Thanks again Miro for the design and Vunce for the testing
It is a very nice sounding dac. I am not very good at comparisons. But knowing the costs and efforts gone into my Ian canada dac, the AD1862 is a no brainer. Straight out of the box with nothing particularly 'esoteric' apart from maybe the Salas Ubib on the 12v rails, it sounds amazing. I know I wouldn't have built the ess dac if I'd have built this first. I may shoehorn the ad1862 into the Ian streamer, using the Fifo and Rpi to feed it. It would really simplify the psus in there.
I've noticed one thing which maybe someone can comment on.
I'm using the AK4118 spdif input. On my initial test of this I think I reported it sounded 'better' set to the Left justified setting. It definitely had higher output so maybe the 'better' was confused with 'louder. The only other setting that worked was 'I2S 24b'.
I have noticed now that it does distort heavily in the Left justified setting, as if it is clipping or something. If I lower the volume it helps alleviate it but still sometimes distortion on transients in the music.
So I leave it on I2S 24 now.
I have another couple of boards I intend to populate and swap the dac chips over to compare. I dont wish to be pulling out and refitting the 1862s too often though. I found it tricky installing them in the sockets.
I'll try some different caps and psus on this one. I have built a VRDN regulator and I also have Miros PSU1 boards coming in the JLPCB order. I shall populate these with tps7a4701 and neg version 3pin regulators from Aliexpress. So I shall split the jumper points and feed separately 12v to AD and oaps.
Also I shall test different i2s source. I have a spare Rpi with Ian canada isolator so I can feed the dac isolated i2s (not reclocked or fifo'd here). Shall be interesting.
Thanks again Miro for the design and Vunce for the testing
IMHO you made the correct choice. The difference between LJ and I2S is that the timing of data bits is shifted by one BCLK period relative to LRCK. Selecting LJ instead of I2S should cause premature clipping since the MSB on one channel will be moved over time time so as to be interpreted as the LSB of the other channel. As a result MSB-1 becomes the new MSB for each channel. One can measure with a scope or logic analyzer, but you seem to have figured out the audible symptoms of an incorrect choice.
Just to say I think I have reached the end of the line for this little 3a5 stage from AliExpress with Miro's DAC. It's taken a bit of work to get it here, but it sounds wonderful, and I'm very happy with it - it should be a great complement to a solid state stage.
I've received a lot of help with this, so much thanks to the fellas over in the tube forum and Rod Coleman who not only supplied the kits for the filaments but also advice along the way.
So, it's a 3a5 kit from AliExpress, choke input B+, CCS on the plates, Coleman filament supplies. Yet to be added is a time delay relay to allow muting on power up and down. Very happy with how this sounds, it's really very very good.
I've received a lot of help with this, so much thanks to the fellas over in the tube forum and Rod Coleman who not only supplied the kits for the filaments but also advice along the way.
So, it's a 3a5 kit from AliExpress, choke input B+, CCS on the plates, Coleman filament supplies. Yet to be added is a time delay relay to allow muting on power up and down. Very happy with how this sounds, it's really very very good.
Attachments
Good result Fran. I have to say that I didn't clock that the 3a5 is a DHT - no wonder you had a lot of hum and I can see how the Coleman regs fixed that.
You might find this of interest/use;
3A5 DHT – Bartola(R) Valves
I'm currently working on a project to build one of Ale's DHT preamp designs, using the 2P29L, that is featured on his Bartola site - prioritised to facilitate testing of my MoFo (need the additional gain) - that's why I've been AWOL from this thread for a while.
BTW, I can let you have a time-delay board if you need one?
Last edited:
Thanks for the comments all - its all about letting this DAC shine through!
@samoloko - Comparison between SEN and 3a5 stage - well, they are different in lots of ways. The SEN is more precise, and I'd think its probably more "correct". Its very sweet, very stable, and there is no start up shut down issues like there are with the valves (as the output caps charge etc). The other thing that can't be overlooked is that SEN is much much simpler, other than the floating supply. But if you use 2 x rechargeable 20V drill batteries that issue is largely put away. I think ultimately that both sound excellent, and it is a matter of personal preference.
@nautiboy - Ales website is the bible on this valve (and all things DHT)- everyone refers to it! I have some 01A valves on the way to try out at some stage, but that's a longer term project as I gather the bits over time. I have some delay relay boards on the way, so I should be Ok. I could just wire them up on a bit of perf board but I'm lazy like that, and much prefer working off a PCB in general (more laziness!). Rod was really great to deal with and would 100% recommend his work.
Do you know when you build something, and you just kinda know its rock solid. Nothing is varying, or needs to be tweaked, or that you need to be careful about? Well, its like that with this build, its one you'd have high confidence in. The last thing is to add the delay relays and its good. Its a question then of a box to put it in to protect fingers from the B+
Fran
@samoloko - Comparison between SEN and 3a5 stage - well, they are different in lots of ways. The SEN is more precise, and I'd think its probably more "correct". Its very sweet, very stable, and there is no start up shut down issues like there are with the valves (as the output caps charge etc). The other thing that can't be overlooked is that SEN is much much simpler, other than the floating supply. But if you use 2 x rechargeable 20V drill batteries that issue is largely put away. I think ultimately that both sound excellent, and it is a matter of personal preference.
@nautiboy - Ales website is the bible on this valve (and all things DHT)- everyone refers to it! I have some 01A valves on the way to try out at some stage, but that's a longer term project as I gather the bits over time. I have some delay relay boards on the way, so I should be Ok. I could just wire them up on a bit of perf board but I'm lazy like that, and much prefer working off a PCB in general (more laziness!). Rod was really great to deal with and would 100% recommend his work.
Do you know when you build something, and you just kinda know its rock solid. Nothing is varying, or needs to be tweaked, or that you need to be careful about? Well, its like that with this build, its one you'd have high confidence in. The last thing is to add the delay relays and its good. Its a question then of a box to put it in to protect fingers from the B+
Fran
Starting to build Miros miracle DAC?
Hi Miro,
I am close to be convinced that following your design might be a good decision. I was reading many of the 180+ pages but sure not all of em so may I have a few questions before I go ahead:
* Is this thread still up and alive and no foreseeable problem to finish it?
* Is the board layout page 136 ending with "-2021-0-14.zip" still latest version and the one to go ahead with?
* Is the AD1862 chip still available via Paddy Garcia?
* Is there a way of buying the PCB via Paddy or anybody else on this thread as well?
* Any group buys I can join?
* May I abuse you, Miro, and the other guys for good advice when struggling with problems? I have done a few pass labs amps but the 1862 DAC would be my first DAC.
Hi Miro,
I am close to be convinced that following your design might be a good decision. I was reading many of the 180+ pages but sure not all of em so may I have a few questions before I go ahead:
* Is this thread still up and alive and no foreseeable problem to finish it?
* Is the board layout page 136 ending with "-2021-0-14.zip" still latest version and the one to go ahead with?
* Is the AD1862 chip still available via Paddy Garcia?
* Is there a way of buying the PCB via Paddy or anybody else on this thread as well?
* Any group buys I can join?
* May I abuse you, Miro, and the other guys for good advice when struggling with problems? I have done a few pass labs amps but the 1862 DAC would be my first DAC.
just one difficulty : to get two good dac chips
It's less difficult than an amp. though chips don't want any shorts. That's all you have t care and respect the instruction by reading the thread. No difficulty and few parts !
As far I know there are no dac chips grup buys anymore and the the pcb is very cheap if you follow the adress of the post 1. Go the the GB question to ask.
The only thing you have to matter is to take care not putting a double aop as it's a single channel aop if you play with the ref we talked about. Other than single channel oap chips need tweaking or further pcb, nicely offered for your own printing pleasure.
Member may help you or not if any problem but no official support, it's a friendly thread for friendly people.
It's less difficult than an amp. though chips don't want any shorts. That's all you have t care and respect the instruction by reading the thread. No difficulty and few parts !
As far I know there are no dac chips grup buys anymore and the the pcb is very cheap if you follow the adress of the post 1. Go the the GB question to ask.
The only thing you have to matter is to take care not putting a double aop as it's a single channel aop if you play with the ref we talked about. Other than single channel oap chips need tweaking or further pcb, nicely offered for your own printing pleasure.
Member may help you or not if any problem but no official support, it's a friendly thread for friendly people.
Last edited:
Got mine fired up last night....could not wait to tidy up so all a bit sketchy but superb sound out of the gates.
Connected it to another SD Player with older A Morei clocks and great soundstage and separation.
Have the new A Morei clocks to test next
Thanks Fran for all the help and support!!
Connected it to another SD Player with older A Morei clocks and great soundstage and separation.
Have the new A Morei clocks to test next
Thanks Fran for all the help and support!!
@Zatopek
>>Is this thread still up and alive and no foreseeable problem to finish it?
- this thread is alive and often on the first page, it would be sad to let such a good DAC die
>>Is the board layout page 136 ending with "-2021-0-14.zip" still latest version and the one to go ahead with?
- you are correct, _2021-0-14.zip is the latest
diyAudio_AD1862_DAC_v1.3_pad_PPY3_2021-01-14.zip is for THT opamp (DIP-8)
diyAudio_AD1862_DAC_v1.3_pad_PPY3_SO08_2021-01-14.zip is for SMD opamp (SO-08)
>>May I abuse you, Miro, and the other guys for good advice when struggling with problems? I have done a few pass labs amps but the 1862 DAC would be my first DAC.
- my thread connects people who like to help others, just a humble philosophy
- Paddy Garcia replied
>>Is this thread still up and alive and no foreseeable problem to finish it?
- this thread is alive and often on the first page, it would be sad to let such a good DAC die
>>Is the board layout page 136 ending with "-2021-0-14.zip" still latest version and the one to go ahead with?
- you are correct, _2021-0-14.zip is the latest
diyAudio_AD1862_DAC_v1.3_pad_PPY3_2021-01-14.zip is for THT opamp (DIP-8)
diyAudio_AD1862_DAC_v1.3_pad_PPY3_SO08_2021-01-14.zip is for SMD opamp (SO-08)
>>May I abuse you, Miro, and the other guys for good advice when struggling with problems? I have done a few pass labs amps but the 1862 DAC would be my first DAC.
- my thread connects people who like to help others, just a humble philosophy
- Paddy Garcia replied