ES9038Q2M Board

Very little progress today, except LTC6655 soldered to small adapter board. Hope to start making good progress on AVCC board. Will post pictures when I have something to show. I will not get pushy about trying to get layout drawings out of people unless people say that want do any. Were we to have interest in that, I would be happy to accommodate and support however I can. If no interest, then I will just get on with it, I think. Want to keep progress happening in some way, and there wasn't any to speak of here today.
 
Hi guys,
I am about to order an ak4137 to convert pcm signal to dsd256 to feed the 9038q2m board.
I just saw on ebay that there is different models with differents clocks.
Could you tell me which one is the better for what we are doing?

A Standard Edition Crystal Configuration 22.5792M 24.576M

B Low frequency version of the crystal configuration 11.2896M 12.288M

C 768K version of the crystal configuration 22.5972M 49.152M

AK4137 I2S/DSD Sample Rate Conversion Board Supports PCM/DSD Interchange | eBay

Thank you.
 
Hi terry22,

You could get A or C.

Have the A version here.

Have not tested version C, which gives highest PCM sample rate output (768k), but as a result doesn't allow for sample rate conversion to 48k PCM.

DSD sample rates are the same for both A and C.

Very low price here: AK4137 I2S/DSD Sample Rate Conversion Board Supports PCM/DSD Interchange Supports DOP Input for hifi amplificatore-in Amplifier from Consumer Electronics on Aliexpress.com | Alibaba Group

Also, in the current descriptions of this board where it says: "NOTE: ...The DSD256 cannot be output. "

When I bought a board it said: "NOTE: ...DSD128 can't output DSD256. Because the output 768K signal needs to use a higher frequency crystal, the default output of this board is only 384K."
 
Last edited:
Hi terry22,

You could get A or C.

Have the A version here.

Have not tested version C, which gives highest PCM sample rate output (768k), but as a result doesn't allow for sample rate conversion to 48k PCM.

DSD sample rates are the same for both A and C.

Very low price here: AK4137 I2S/DSD Sample Rate Conversion Board Supports PCM/DSD Interchange Supports DOP Input for hifi amplificatore-in Amplifier from Consumer Electronics on Aliexpress.com | Alibaba Group

Also, in the current descriptions of this board where it says: "NOTE: ...The DSD256 cannot be output. "

When I bought a board it said: "NOTE: ...DSD128 can't output DSD256. Because the output 768K signal needs to use a higher frequency crystal, the default output of this board is only 384K."

Thank you mark, you're very helpfull as always.
If I understand it well, to output dsd256 I have to upsample all signal at least at 176.4K. Is that right?
(That is not a problem, I can upsample it in moode player using SOX)
 
Thank you mark, you're very helpfull as always.
If I understand it well, to output dsd256 I have to upsample all signal at least at 176.4K. Is that right?
(That is not a problem, I can upsample it in moode player using SOX)



My board arrived late yesterday and with any luck I may find time to connect it today.

From what I read files at 44.1 to <176.4 can be converted by the board to DSD64 or DSD128.

Files from 176.4 up can be converted to DSD256
Files below 44.1 can't be converted to DSD at all.
 
Last edited:
Both types of AK4137 boards here can convert 16/44 to 11.2MHz DSD, the highest DSD rate.

Just remember how to wire them up, which is a bit odd for the low cost AK4137 board. Pictures here: ES9038Q2M Board - Page 204 - diyAudio

Also, I added some extra ground pins to that board.

Hi Mark, Terry, I've just connected the AK4137 and it works first time ootb...

For testing I'm using the "b" system of 'oh whoops so what' devices .
Moode Audio on raspberry Pi B (yes Terry, the single core B) > Xmos U8 > AK4137> Ebay 9018K2m dac.

Moode is playing webradio and output to usb ( no sox resampling) shows as."Encoded at: 32 bit, 44.1 kHz, Stereo, 2.822 mbps"

AK4137 shows input as I2S and output as DSD256. so confirms Marks comment above.

Later today I hope to connect the AK4137 into the main system and do some listening.

Mark, what was the digital filter setting you recommended for this board ?
 
Last edited:
I am using the ShortSharp filter. There are four to choose from and two of them cut a lot of HF out of the top of the audio band which I saw feeding them pink noise. So, there are two more or less flat filters to choose from. If your dac is modded enough I think ShortSharp will sound best and most accurate. If it sounds too bright or something then you probably still have some distortion to get rid of and or are missing some bass. Everything matters, including harmonic distortion compensation and DPLL bandwidth in the dac registers. Don't stop until you get to the end would be my advice. It is all doable. The pin lifting is not even a challenge, its easy and reliable with a little practice. Please lets talk if any qualms about doing any of the mods.
 
Last edited:
Starting to work on AVCC board layout. This time instead of paper, I needed to play around with the physical board and the major components. What I came up with is shown in the pictures below.

A red line across the edge of the board indicates the approximate area where the ground plane will be connected to the dac board ground plane. There are also a couple of smaller red marks showing where the 10uf film cap leads will go.

Ground plane will run on the solder side of the board again like we did with the output stage board. The ground plane will run underneath the film cap body, so ground plane needs to go in first with notches at the edge where the film cap leads go. There will also be some ground plane run around the edge of the board to fully surround the LTC6655 adapter board.

This time opamp electrolytic filter caps will be located at the sides of the opamp socket rather than off the ends. Thinking about some holes in the the ground plane for where the opamp 10uf input filter cap and 10k resistor get joined together. Also need to figure out what to do with creating the two Vref signals needed by the output stage. I sure would like to use SMD resistors for that, although metal film could be used too. Have to see what I have in stock because I don't want to place an order for just four resistors.

The AVCC board will be oriented with respect to the dac board such that the AVCC opamp pins point towards the 47uf AVCC filter caps in the dac board and also point at the output stage board. It will be located right next to the 47uf filter cap pins.
 

Attachments

  • AVCCstart_layout1.jpg
    AVCCstart_layout1.jpg
    258.5 KB · Views: 432
  • AVCCstart_layout2.jpg
    AVCCstart_layout2.jpg
    233.1 KB · Views: 430
  • AVCCstart_layout3.jpg
    AVCCstart_layout3.jpg
    286.2 KB · Views: 405
  • AVCCstart_layout4.jpg
    AVCCstart_layout4.jpg
    324.4 KB · Views: 411
Last edited:
Hi Deenoo,
The 47uf AVCC caps in my existing modded dac board are Panasonic Aluminum Organic Polymer, part number EEH-ZK1E470R. They are actually intended to be SMD, but I removed the plastic spacer on the bottom and straightened out the short leads in order to solder them into the cap mounting holes.

Thing is about those caps is I decided to try them because they are 47uf like ESS specified, but low ESR. I don't know if they are better or worse than other capacitor choices for AVCC. I figured if there were any problems I could try some other ones, but I never got around to trying any others. Although my AVCC on that board seems fine and produces bass with the same balance to other frequencies as DAC-3 and Allo Katana, I don't know if other caps might have made the job easier or performed better in some way. For that reason, I am still open to trying some different caps to see if any difference in SQ, or anything else.

Speaking of Katana, I think I mostly know how they designed the AVCC supply for it, and it is very different from the opamp circuit ESS recommends. Sometime later when I write a review of Katana I would like to describe how I think it works in more detail. Maybe describe some other things of interest to the group as well, since we are a pretty technical bunch and may have more than average interest in such things.

In the meantime and since we are using the opamp circuit for our dacs now, I think it would be fine if people want to experiment with different caps. It's just if they do they should follow through to swap them out and make sure they actually go through with some testing. If they don't want to get into that, then using the ones I used before worked okay then and will probably work okay again.

That being said, I was thinking of trying some slightly different ones this time, but no final decision yet. They are still polymer, pretty similar to the ones I used before, but intended to be through hole. They may be different in other ways too, I will have to look at the data sheets.

If I wanted to try a bigger change compared to last time around, it would probably be some good quality standard aluminum electrolytics. They might work okay too.

For more extreme testing, I also picked up some very low ESR super caps. Never used any of those before, so that would be a whole new experiment to play around with. Might not use the opamp circuit anymore in that case, have to think about it some more and or maybe just try it to see.
 
Last edited:
Hi Mark, I'm certain you're wondering why some of us are not reporting our build results and questions...... well it may be a geographical hiccough.... I for one am awaiting a 'finished' instruction and BOM.

Down here at the 'rs end of the world we are poorly served by parts suppliers....and the vagaries of international shipping mean that often the cost of shipping exceeds the cost of parts...!

Needless to say if a part's specified value changes or the part is not commonly available here (your 0.1% resistors for example) then the cost to order individual items is dwarfed manifold by the postage.
That's assuming suppliers will send one-offs...most want a minimum order that can be US $50 or more before they will ship here.

So, while the build is still a moveable feast I am happy to observe, follow and query..... once everything falls into place and a final "recipe' (similar to Serge's ?) is available I'm very keen (and fearless !) to build...

Cheers,
Bob
 
@Mark @Drone,
thank you for all the info.

For now, I connect to the board via I2S only : rpi3->kali reclocker->dac Board

Do you think connection via usb and xmos like Drone will be better?

If yes, Do you think this xmos is ok?
XMOS +CPLD U208 DAC daughter card USB digital interface I2S DSD output Suitable AK4497 ES9018 ES9028 ES9038 DAC decoder board-in Amplifier from Consumer Electronics on Aliexpress.com | Alibaba Group

or dou you have better models to advise(in a reasonnable price range)?
 
Starting to work on AVCC board layout. This time instead of paper, I needed to play around with the physical board and the major components. What I came up with is shown in the pictures below.

A red line across the edge of the board indicates the approximate area where the ground plane will be connected to the dac board ground plane. There are also a couple of smaller red marks showing where the 10uf film cap leads will go.

Ground plane will run on the solder side of the board again like we did with the output stage board. The ground plane will run underneath the film cap body, so ground plane needs to go in first with notches at the edge where the film cap leads go. There will also be some ground plane run around the edge of the board to fully surround the LTC6655 adapter board.

This time opamp electrolytic filter caps will be located at the sides of the opamp socket rather than off the ends. Thinking about some holes in the the ground plane for where the opamp 10uf input filter cap and 10k resistor get joined together. Also need to figure out what to do with creating the two Vref signals needed by the output stage. I sure would like to use SMD resistors for that, although metal film could be used too. Have to see what I have in stock because I don't want to place an order for just four resistors.

The AVCC board will be oriented with respect to the dac board such that the AVCC opamp pins point towards the 47uf AVCC filter caps in the dac board and also point at the output stage board. It will be located right next to the 47uf filter cap pins.

For now I think, It will be doable for me.
I will order parts for AVCC this weekend
 
Hi Guys, I will have some more to say a bit later in response to the various issues raised in recent posts.

In the meantime I wanted to let you know I tried working on making a set of output board instructions with the goal of getting them into PDF format. I have pictures for most of the steps, but the few pictures I added to an rtf file came out looking very bad when it was saved. Problem is the original pictures are hi-res to make details clear and they won't fit into some file formats without being badly compressed.

An example file with some instructions and a few pictures can be found temporarily in my dropbox: Dropbox - OSB.zip
 
Last edited:
For now, I connect to the board via I2S only.
If yes, Do you think this xmos is ok?

The XMOS board you linked to looks fine. I have one using the same chip and it works well.

The AK4137 boards we use only have I2S output, and so I am using I2S input only now.

If you only need USB input, then the XMOS board and and low-cost AK4137 board should work fine.

If you require SPDIF and or TOSLINK inputs in addition to USB, then you might want to consider using the more expensive AK4137 board. Depends if you want to upsample and do DSD conversions on SPDIF and or TOSLINK inputs or not. If not then you could just use the SPDIF and or TOSLINK inputs on the dac board for those.
 
Hi Guys,

May I ask if anyone looked at the prototype instructions I posted a link to? If so, you may notice the picture quality is poor. That happens when trying to embed a hi-res image into a text type of format as .rtf is. Therefore, I have concerns about using pdf files of instructions. Maybe it would be better to have a list of instructions with the names of the picture files to refer to? That way you could use the hi-res pictures to better see how the build should be. What do you think?
 
More AVCC board progress. See pictures below.

Pics show bottom side of the board which will point towards the output stage board. The big 10uf film cap will stick out on this side. Other components go on the other side.

Third picture shows color outlines for various parts of the circuity. Blue areas are for the opamp socket and LTC6655 adapter board mounting header pins. Red is for the big 10uf film cap and the electrolytic opamp decoupling caps. Orange shows a few holes I left open for soldering together the opamp input filter, and Yellow shows where the Vref dividers and filter caps go.
 

Attachments

  • AVCCgp1.jpg
    AVCCgp1.jpg
    144.3 KB · Views: 384
  • AVCCgp2.jpg
    AVCCgp2.jpg
    179.5 KB · Views: 383
  • AVCCgp3.jpg
    AVCCgp3.jpg
    132 KB · Views: 378
Last edited:
The XMOS board you linked to looks fine. I have one using the same chip and it works well.

The AK4137 boards we use only have I2S output, and so I am using I2S input only now.

If you only need USB input, then the XMOS board and and low-cost AK4137 board should work fine.

If you require SPDIF and or TOSLINK inputs in addition to USB, then you might want to consider using the more expensive AK4137 board. Depends if you want to upsample and do DSD conversions on SPDIF and or TOSLINK inputs or not. If not then you could just use the SPDIF and or TOSLINK inputs on the dac board for those.

thank you Mark.