you have the guarantee the JLS is working, have a good layout, support from the vendor, are compatible with several pcm chips, and are stackable on several pcbs given in this thread ! (no messy wires and shortcuts)
Sometimes less expensive is more expensive !
And you avoid the mess of shift registers and its layout related.
Sometimes less expensive is more expensive !
And you avoid the mess of shift registers and its layout related.
Last edited:
The JLsounds board allow to configure for separate PSU to feed the USB ( dirty) side and the clock ( clean) side for better isolation.
I do not know which Xmos board you are referring to in AliExpress, they have quite some variety of Xmos boards. But most of them are built to be powered from the USB bus just like the legit Amanero Combo384 board. Powering them from a low noise PSU instead of your computer USB bus usually results in better sound. You need to cut some trace or remove some small components for most of these boards.
I used to have two of these AliExpress boards and they all work on all the DACs by Miro. But I prefer the JLSounds board, which sounds better. I am not good with my language to describe the differences but it is immediately discernible in my system.
I ve got several items together with chips from Tamra, not sure if the board is stackable. Not critical. So I need to plan for another psu and transformer to feed isolated clean power.
My shift registers are not mess. It would be if signal lines cross each other and GND below them would be interrupted, but it is not in my layouts
He can of course choose design without shift registers to fit jlsounds directly, everything is shared.
From my experience, the most important thing is the internal design of the DAC chip, which we cannot influence in any way. If the internal switches are not "smart enough" to deal with digital/analog interference, DAC will not be suitable for audio. You can use even the cleanest power supplies or USB-I2S, but if the DAC chip is poorly designed, the digital will be transferred to your output. Luckily the AD1862 is designed perfectly
It's I2SoverUSB stackable.
I also send you with shift resistor one.
You need +-12v and +-5v 〜 +-11v for main PCB.
Clean usb +5v and another usb +5v.
+-15v if you use Patrick's cenIV through hole.
I also send you with shift resistor one.
You need +-12v and +-5v 〜 +-11v for main PCB.
Clean usb +5v and another usb +5v.
+-15v if you use Patrick's cenIV through hole.
Stackable or not, is simply physical attributes. They can be used in the same way with the JLSounds. 1. Skip the shift registers on the Dac board, feed the pcm signal from the JLSounds and 2. JLsounds i2s signal to the i2s inputs of the Dac.
That is the flexibility of JLSounds board. You will have lots of fun.
I cannot imagine Miro doing a messy job right at the entry point.. 🤣My shift registers are not mess. It would be if signal lines cross each other and GND below them would be interrupted, but it is not in my layouts
My shift registers are not mess. It would be if signal lines cross each other and GND below them would be interrupted, but it is not in my layouts
He can of course choose design without shift registers to fit jlsounds directly, everything is shared.
From my experience, the most important thing is the internal design of the DAC chip, which we cannot influence in any way. If the internal switches are not "smart enough" to deal with digital/analog interference, DAC will not be suitable for audio. You can use even the cleanest power supplies or USB-I2S, but if the DAC chip is poorly designed, the digital will be transferred to your output. Luckily the AD1862 is designed perfectly
Don't get me wrong, it is just an illustration to say more than a layout with shift registers has in plus to the FGPA a far more complex layout, another active traces arounds, passive parts too, longer I2S bus... I have not saud you did it wrong or bad.
The JLSOUNDS is less messy bacause you can just plug it, it is also simplier for the people that are not easy with smd soldering. You also save the cost of the shift registers, the NPO caps and doesn' share the +5V with a cap that modulates also the "digital" positive power rail of the ships... That's what messy is saying for me, maybe not the accurte word in english due to my bad rulling of this language.
Edit, btw : you have the I2S line that cross the 5V line whatever you have Jumpers. You did because no other possibilities with a simple 5V feeding both dac ic with 2 layers, whatever they cross orthogona, you have a messy node here, in fact with each ICs l. I made a 4 layers in order there the jumpers see a clean ground no polluted by the return current of the "dirtier, edger" I2S . Can I hear it is another story, lol ! You can measure it if you want I gave you my gerbers to thanks for your help with all the good files 🙂
Last edited:
I'll be placing an order with JLS. Should I add an oscillator board too if I am going to play with upgrades down the line? How does AD1862 respond to the external clock?
That s okay, focus now on the parts to buy as the female header that is not provided, a 4k7 metal resistor. And choose your outputstage.
Read the thread on the sand at the beach for advices on parts and opa. Rewritting the thread each time makes it even longer. Read the shortcuts in miro blog linked in the first post.
Read the thread on the sand at the beach for advices on parts and opa. Rewritting the thread each time makes it even longer. Read the shortcuts in miro blog linked in the first post.
The 4.7k resistor (R13 on stacked DAC board) is no longer needed for the current JLSounds FIO board.
https://www.diyaudio.com/community/...s-input-nos-r-2r.354078/page-365#post-7438473
https://www.diyaudio.com/community/...s-input-nos-r-2r.354078/page-365#post-7438473
Micro project of the day.
I wasn't entirely happy with some aspects of the sound of my 1862 (unofficial) board. Today, decided to replace the OPA1641 I/V with a quick and dirty attempt at EUVL's CEN. The little protoboards fit the DIL opamp socket, so the experiment was really easy and preserved the same Shinkoh IV resistor already on the dac board. Obviously with a floating PS.
The tonal balance is finally normal with all sorts of music. There are other improvements too. Very impressed and a big thank you to Patrick.
I wasn't entirely happy with some aspects of the sound of my 1862 (unofficial) board. Today, decided to replace the OPA1641 I/V with a quick and dirty attempt at EUVL's CEN. The little protoboards fit the DIL opamp socket, so the experiment was really easy and preserved the same Shinkoh IV resistor already on the dac board. Obviously with a floating PS.
The tonal balance is finally normal with all sorts of music. There are other improvements too. Very impressed and a big thank you to Patrick.
Attachments
At the time JLS got in business i had a little collection of USB boards. EXA, WaveIO, Amanero. JLS sounded better than any of the above, the Amanero by a significant margin, being the worst. In more recent times i bought a very highly specced diyinhk board which sounded ludicrously poor in comparison. So, i guess i am sticking with JLS. Same as many high end manufacturers.
Fwiw i have a slight preference for the older Mk2 board.