ES9038Q2M Board

Can I seek help from the group?
I have ver 1.04 of the DAC board, with xmos u8 usb to I2S connected to it. However, been using spdif input as I find it SQ better than usb - jumper J3 is shorted per instruction. I ported the DAC board to another enclusure, WITHOUT the xmos us decoder. Now there is no sound coming from it.
Is my understanding correct that the DAC board using spdif input will work even without the usb decoder board? Anyone tried it before? Any clues as to why no sound is coming from it?
Thanks and I appreciate any input.

J2 for SPDIF input ;)
 
There is a blue led ( Lock Led ) :The led is blue when a signal ( SPDIF) is present at the Coax input

Checked all wiring and sources, everything seems to be in order.

Source computer and usb/spdif (Halide) is okay, as it works on another DAC. Changed shorting caps between J2 and J3 - nothing. There are 2 LEDs,one is lock the other is dsd, no lights. Power source +/-15V is okay.

I tested outside my system, I.e just computer, DAC and earphones nothing coming out. I could hear the sound when power is switched on the earphones.

The only thing I haven’t done is to solder back the xmos usb decoder.
 
Hi Guys,
I'm back after kind of taking a sick day off. Thanks to those that provided updates on where we are all at.

dv187736, Could it be that you when you unsoldered the XMOS board you left a solder bridge or some short? The dac board should run fine without XMOS, but there is 5v over there among the I2S pins that could get shorted to something else. If you have a voltmeter you might check that the big regulator on the dac board drops the +15v to +5, and smaller regulator then drops the +5v to +3.3v

If power is okay, then for most of the boards I am aware of J1 selects optical input, J2 selects SPDIF, and J1 and J2 both on or both off selects I2S.

If power looks okay, you have a valid SPDIF signal with the appropriate jumper selected and no shorts anywhere, then the board should work and the lock LED should light up.
 
@eslei,
You can probably get much better sound quality if you were to decide to go with the recommended mods. However, if you are happy with what you have then I guess that's fine too.

@Mikett,
Not sure you need to do so much pin bending. I would probably just use wire wrap jumpers for a lot of what you are trying to do. For short lengths there should not be any penalty. Also, if you break off a pin from too much bending, you may be able to scrape off some epoxy from the case and expose enough of the stub to solder a little piece of wire wrap wire to. I broke one once from too much fiddling and was able to patch it as described.
 
@Ddenoo,
Good to hear LTC6655 worked out well for you. The whole output stage is very good too.

@Serge,
Thanks buddy for proving to me that people really wanted a pdf with some instructions to give some the confidence to go forward with modding. I didn't believe it at first, but I think you were really onto something with that. By the way, I hope you will decide to do the full mods, I think you will be happy if you do.
 
hi Mark
I have built the avcc circuit, but in the simple form, exactly as in the ess file, and im using the TPS7A4700 3.3V circuit as a source.
for some reason though when i connect everything up the voltage i get on the output is 3.9V instead of 3.3V:)
will recheck the conections on the weekend, as i have time to work on the dac only on the weekends
 
hi Mark
I have built the avcc circuit, but in the simple form, exactly as in the ess file, and im using the TPS7A4700 3.3V circuit as a source.
for some reason though when i connect everything up the voltage i get on the output is 3.9V instead of 3.3V:)

Hi Paul,
In that case I would check the voltage at the output of the 3.3v source, the opamp inputs, and their outputs. Unless the problem is at the source with its voltage programming, something sounds very wrong.

Also, my guess is that the 3.3v source is more noisy than would be best, so if stuck with that you might want to use more stages of filtering after it, although it will take longer to come up to full 3.3v after you turn on the power that way.
 
Hi Mark,

Here is a revised copy of the AVCC circuit, using standard schematic symbols. Let me know if there are any issues with it.

It is based on the schematic and BOM from bih.

I've also attached the original BOM from bih with only the AVCC sheet revised to reflect the capacitor types that you recommended. I used 47uf Panasonic low ESR Electrolytics at 1/10th the cost rather than the Organic Polymers you mentioned.

I'll probably do the I/V stage schematic next and then post a revised copy of the BOM from bih before going onto the PCB layout step of either circuit.

Also, where do the 22uF Wima's fit into the circuit, I didn't quite figure that one out?

John
 

Attachments

  • ESS9038AVCC-V0.2.jpg
    ESS9038AVCC-V0.2.jpg
    281.3 KB · Views: 659
  • bom OutputStage-V02.zip
    12.3 KB · Views: 95
Hi MrSlim,
Thank you for the nice work!

Actually, the 22uf Wima caps replace the 47uf output caps, C8 and C9.

Also, I have to say some of the mix of film caps and electrolytics doesn't seem entirely rational. If electrolytic leakage currents cause noise that is audible, which presumably they may at some very picky level, that is. I honestly don't know how much the 22uf film caps help because I haven't tried every variation of parts systematically. However, the AVCC load may also have current draw variations for which the film caps somehow work better. On the other hand, Vref supplies a fixed load, so some difference in that respect. It is entirely possible that most of the benefit of the recent changes comes from the new output stage, and LT6655 more than from film caps. Just don't know, but they aren't too expensive and it works. Maybe someone else can come along and do some more testing with different components. After I finished the first modded dac, with all the work I did on that one since it was the first, I said, "never again." Then I went ahead and did the second one to document how through hole components could be used. I feel like I have done my share of work in that regard. There are other things I may pursue that interest me, but I feel like I need a vacation from dac part swapping and listening tests for awhile. Not too worried though. It sounds good, if ever so slightly bright still but not objectionably so to me (some of that should diminish with HD compensation, I would expect). It sounds plausible enough although still not as good as DAC-3. It's good, enjoyable, non-fatiguing, non-clinical listening. Just not quite world class perfect.

EDIT: Actually, I may try a another circuit mod out of curiosity, which would be a pre-I/V anti-RF filter. I see other people are using them and feel they are important so I still have some curiosity about that for this dac. Will provide updates when I know more, likely in a few days.
 
Last edited:
Hi Mark,

Here is a revised copy of the AVCC circuit, using standard schematic symbols. Let me know if there are any issues with it.

It is based on the schematic and BOM from bih.

I've also attached the original BOM from bih with only the AVCC sheet revised to reflect the capacitor types that you recommended. I used 47uf Panasonic low ESR Electrolytics at 1/10th the cost rather than the Organic Polymers you mentioned.

I'll probably do the I/V stage schematic next and then post a revised copy of the BOM from bih before going onto the PCB layout step of either circuit.

Also, where do the 22uF Wima's fit into the circuit, I didn't quite figure that one out?

John

I think it's worth considering that the opamp is driving a capacitive load
directly on it's OP and as such will be subject to ringing etc depending on the
ESR of the cap.

Yes, there is no actual voltage swing on the OP but there is a transient
loading to some degree.

It's worth making allowance for this and there a many ways to do it. It may
very well affect the sound of the DAC.

Just sayin' :)

T
 
Understood. The caps being used for that are mostly inductive over 100kHz or so, and likely with increasing skin effect related resistance. If they were ceramic then it would seem much more likely an issue. In addition, the circuit is the one recommended by ESS. Examination with a scope during operation showed only stable DC, as well.
 
Last edited:
Hi Mark, Good to see that you've improved your build, although i'm unsure
what component changes you have made to achieve this, could you clarify?

Being a man on a budget i will be placing 2 or 3 orders a month or so apart, first up will be on board regs and all things AVCC. i'm about to pull the trigger on that.

By the way i found a board for the LTC6655 that takes a film TH cap over at EEVblog,
404 Not Found, the cost is 4$ for 3 boards from OSHPARK, I've ordered some up for myself.

As for interest in you carrying on, please do. this thread will remain relevant until the next gen ESS chips.

I'm not doing the O/P stage yet but have one question that's bugging me: is there a cap across the diff inputs as tn the forum schematic or not as in your hand drawn schematic?

@ serge, looking forward to your AVCC pcb, thanks for component refs.

Andy
 
Hi Guys,
I'm back after kind of taking a sick day off. Thanks to those that provided updates on where we are all at.

dv187736, Could it be that you when you unsoldered the XMOS board you left a solder bridge or some short? The dac board should run fine without XMOS, but there is 5v over there among the I2S pins that could get shorted to something else. If you have a voltmeter you might check that the big regulator on the dac board drops the +15v to +5, and smaller regulator then drops the +5v to +3.3v

If power is okay, then for most of the boards I am aware of J1 selects optical input, J2 selects SPDIF, and J1 and J2 both on or both off selects I2S.

If power looks okay, you have a valid SPDIF signal with the appropriate jumper selected and no shorts anywhere, then the board should work and the lock LED should light up.

Hi Mark,

I did as instructed. DAC is now working, many many thanks. The problem was that I wasn’t getting 3.3V after the regulator (mod, a discrete one). Then I realised my stupid mistake. Got the regulator installed in reverse! Many many thanks.

One thing peculiar though. My DAC spdif works with shorting cap on J3, not J2.

Anyway, back to music nirvana.

Cheers
 
Hi,
Good to hear back that the dac is now working. J3 rather than J2 only means that the firmware was a little different when your's was made.

By the way, the ES9038Q2M is still the same, which means you could mod the dac you have into being a very, very good one as we like to do here.