I have had this problem with the V1 board, saw this when I as experimenting with various DC power supplies, were you using the default transformer setup or providing it with an external DC supply?
Hi there,
I am running v.99 with amanero USB with synology nas, no unlocking issue from 44.1 to 384. DAM v2 board, no mod, factory filters. Powered by 2x7vac or +/-12vdc.
The only way to really know whats going on is to put a scope on the I2C bus. The Amanero has been flawless with the dam in my setup.
Actually Paul states to use a lesser value for the compensation cap with such a high capacitance value per rail, between 0 and 75R. 499R just happens to work best with less capacitance per rail and is an overall compromise to not requiring large capacitance in the stock config. The LowResMod is officially 499R/0R/0.01R (feedback/compensation/series)
Last edited:
Finishing up the DAC
Hi, I have been using the Soekris DAC since this summer and am quite happy with it and now it's time to finish up this project. Move it from a plywood board to a nice cabinet. Add some nice DC supplies and so on.
I just wanted to have some feedback to the remaining things I plan to do:
1. Remove the 5V regs and feed +- 5VDC from a JSR-06 supply at J2
2. Feed 6-7VDC from a Placid supply for the digital regs at J1
3. Remove diode bridge
4. Possibly add caps to 3.3 and 1.2V at J2 (under the board)
The current state of my rev1 board:
Opamps removed, 16-cap mod, Factory mod, C135,142 removed and off-board Polystyrene used instead.
Any thoughts??
By the way, I am quite happy with this DAC. My best one yet.
Ulf
Hi, I have been using the Soekris DAC since this summer and am quite happy with it and now it's time to finish up this project. Move it from a plywood board to a nice cabinet. Add some nice DC supplies and so on.
I just wanted to have some feedback to the remaining things I plan to do:
1. Remove the 5V regs and feed +- 5VDC from a JSR-06 supply at J2
2. Feed 6-7VDC from a Placid supply for the digital regs at J1
3. Remove diode bridge
4. Possibly add caps to 3.3 and 1.2V at J2 (under the board)
The current state of my rev1 board:
Opamps removed, 16-cap mod, Factory mod, C135,142 removed and off-board Polystyrene used instead.
Any thoughts??
By the way, I am quite happy with this DAC. My best one yet.
Ulf
Reporting back...
Testing (non)locking issue, confirming sample rates via RS232 output on PC...
1. Optical & coax both okay up to 96k & 192k respectively.
2. USB from PC okay up to 384k.
3. USB from NAS problematic at 96k and above.
Synology NAS with DAM via USB appears to be the problem then. The NAS Media Server has an option to 'Enable Audio Transcoding' for various formats. Struggling to see the need to use it as I play flac files, but worth a try... Better if not 100% fixed, wish I knew why?! Synology have announced a Version 6 Beta release, will wait for the full release then see if it improves. (Apologies for heading off-topic a little.)
Paul's recommendations... I thought he originally recommended 0.01/499/499R with 4000uF, although looks like he is suggesting replacing the Comp Cap with 200R or 150R works well in simulations (17 Oct). Guess I have missed a post somewhere?
Testing (non)locking issue, confirming sample rates via RS232 output on PC...
1. Optical & coax both okay up to 96k & 192k respectively.
2. USB from PC okay up to 384k.
3. USB from NAS problematic at 96k and above.
Synology NAS with DAM via USB appears to be the problem then. The NAS Media Server has an option to 'Enable Audio Transcoding' for various formats. Struggling to see the need to use it as I play flac files, but worth a try... Better if not 100% fixed, wish I knew why?! Synology have announced a Version 6 Beta release, will wait for the full release then see if it improves. (Apologies for heading off-topic a little.)
Paul's recommendations... I thought he originally recommended 0.01/499/499R with 4000uF, although looks like he is suggesting replacing the Comp Cap with 200R or 150R works well in simulations (17 Oct). Guess I have missed a post somewhere?
NAS Issue - problem found?
Had option checked: 'Restrict access to Media Server for newly detected UPnP/DLNA devices'. Uncheck and looks to be working 90-95%. Switch off the transcoding option and still working. Problem solved? Operator error? Looking good, whatever! Next, will try 330R in parallel with the 499R on Comp Cap (198R effective) and 220R in parallel (153R effective).
All good fun!
Andrew
Had option checked: 'Restrict access to Media Server for newly detected UPnP/DLNA devices'. Uncheck and looks to be working 90-95%. Switch off the transcoding option and still working. Problem solved? Operator error? Looking good, whatever! Next, will try 330R in parallel with the 499R on Comp Cap (198R effective) and 220R in parallel (153R effective).
All good fun!
Andrew
Last edited:
With the buffer opamps removed there is no need for a "minus" supply on the input. I am using +5.5 volts. (DDDAC1795 shunt reg supply)
The Vref needs plus/minus 5 volts - any more than that is wasteful heat for the Vref components.
Adding caps to the 3.3 reg has not been universally acclaimed. moreDAMfilters/Paul found it to be inconsequential. You can experiment. I am leaving it alone for the time being though I will try a better regulator.
Probably worth the trouble for the dc-dc converter.
If using a good regulator for the +5/6 volts supply at the "front" remove all of the on board caps.
I am using the cap pads after the 5 volts regs (pre-reg for Vref) as the input for the replacement regulators.
a bit worrying
its a bit worrying that these 4 13 ohm resistors would have so much influence
after further investigation approx half the msb goes through these 13r's
although these 13r's have the largest influence there is many other resistors contributing to an increasingly accumulative sq loss
it would seem from the pics from the rmaf that there's a different type of resistor on the most significant portion of the r2r ladder
and by the look of the dk website there will be a diy version and an oem version along with a newer version of the 1021 called 1121
is all the creases ironed out?
i'm doing my testing on a 0.05 v1 board so i'll likely not bother changing any resistors on this board maybe the "final" one
i removed the lpf cap for me it sounds much better without
but maybe there's a cap that will sound better
its a bit worrying that these 4 13 ohm resistors would have so much influence
after further investigation approx half the msb goes through these 13r's
although these 13r's have the largest influence there is many other resistors contributing to an increasingly accumulative sq loss
it would seem from the pics from the rmaf that there's a different type of resistor on the most significant portion of the r2r ladder
and by the look of the dk website there will be a diy version and an oem version along with a newer version of the 1021 called 1121
is all the creases ironed out?
i'm doing my testing on a 0.05 v1 board so i'll likely not bother changing any resistors on this board maybe the "final" one
i removed the lpf cap for me it sounds much better without
but maybe there's a cap that will sound better
replacing the 3.3v ldo with a lifepo4 battery was my first experiment
it was crystal clear that this is an area of great improvement
infact ive never seen a case where improving the power supply to a clock had an inconsequential effect
Hi Ulf,
first I wouldn't suggest you to feed +-5v trough J1. Reason is that if you look closer, (see attachment) board traces from J1 they are all way smaller compared to ones going from reg's to buffer's. You will be better off connecting +-5v to pads under regs (attachment 2).
Second, Placid supply is excellent VR, however keep in mind that positive rail need almost 3 x more current than negative.... remove diode bridge, yeah why not.
Third, it was already (by Paul) concluded that adding any more caps to 3.3v and 1.2v didn't make any difference! This is because DAM1021 is R2R dac and as such doesn't really (unlike sigma-delta) need high precision clock to perform at it's best.
cheers
Danny
Attachments
Nigel,
You know I completely respect your observations but I was thinking it was a long time ago in the evolution of this board, before the improved filters and the Vref fixes, along with the fact that moreDAMfilters/Paul added CAPACITANCE and heard and measured little difference in his set-up. I think he also tried an improved regulator and was less than overwhelmed. Of course, a cap is analogous to a battery, sort of, but in your application I figure it is just as much a regulator change as a simple addition of capacitance, which is why I suggested it be relegated to later in the modification cycle - adding capacitance, that is. Which is what the fellow asked about.
Changing the regulator itself, whether a simple replacement on the board or a
big battery remotely charged by a superior regulator would have to be better but, at this point, and in my case is purely speculative, I am thinking I would rather have all of the other stuff sorted before attacking the clock regulator. Especially after Paul's travails. Though I guess one can look at that two ways - better to blow up the board BEFORE doing all of that other stuff? Excuse my attempt at humour.
Just another opinion ...
the 3.3v will have no bearing on the filters or the vref fixes which i will admit are the primary improvements since first release i would say 3.3v replacement with lifepo4 was very evident to be an area of improvement despite not having new filters and vref mod/fixes
Pauls done great work no doubt about it but it would be be a boring world if we agreed on everything
I do have more cofidence in what people test themselves rather than quoting someone else
did a bit of cap vs battery earlier, i had thought with these supercaps we could forget about the complications of batteries but its looking like we've gone far with supercaps but still not even remotely close
(im sure you can read between the lines
)so your probably right just adding a cap to the ldo aint going to blow your socks off
theres always risk
i know ive had more than a few casualties with diy
Last edited:
We both have had a run at destroying stuff.
Education never comes cheap.
Our "argument" is about priorities. My point was that when you did that ancient mod (in SOEKRIS DAC years) I do not doubt it made a signifiant difference. My question, is with new filters and Vref mods, would it seem as significant as it did then? Obviously, I am leaning towards NO as the answer. I am not saying it makes no difference; I am speculating that it likely will not seem as starkly obvious as it was way back then.
I know they are not actually connected in any way and I am certainly not dismissing doing it. My suggestion is that it might not be as important as the other two things I keep mentioning. You know I enjoy a lively conversation especially with someone who knows more than I do.
I certainly appreciate your guidance on the battery/supercapacitor quandary. What i see writ large between the lines is that you saved me some money by not buying supercaps. THOUGH I have this suspicion than TOO LOW ESR is a problem within the Vref. I was looking at supercaps with unusually high ESR. Just a hunch ...
With the via it should be easy to sub the battery in there. I saw you had no fear of running the 3.3 volts caps at four volts. I guess at this low current draw the A123s should tolerate four volts? You had used long leads from the battery to the 3.3 reg pads so the lengths of wire required for the batteries in the Vref would be nothing in comparison. Stabilizing those batteries in relation to the PCB will require a little thought.
The SOEKRIS DAC will prevail!!!
Education never comes cheap.
Our "argument" is about priorities. My point was that when you did that ancient mod (in SOEKRIS DAC years) I do not doubt it made a signifiant difference. My question, is with new filters and Vref mods, would it seem as significant as it did then? Obviously, I am leaning towards NO as the answer. I am not saying it makes no difference; I am speculating that it likely will not seem as starkly obvious as it was way back then.
I know they are not actually connected in any way and I am certainly not dismissing doing it. My suggestion is that it might not be as important as the other two things I keep mentioning. You know I enjoy a lively conversation especially with someone who knows more than I do.
I certainly appreciate your guidance on the battery/supercapacitor quandary. What i see writ large between the lines is that you saved me some money by not buying supercaps. THOUGH I have this suspicion than TOO LOW ESR is a problem within the Vref. I was looking at supercaps with unusually high ESR. Just a hunch ...
With the via it should be easy to sub the battery in there. I saw you had no fear of running the 3.3 volts caps at four volts. I guess at this low current draw the A123s should tolerate four volts? You had used long leads from the battery to the 3.3 reg pads so the lengths of wire required for the batteries in the Vref would be nothing in comparison. Stabilizing those batteries in relation to the PCB will require a little thought.
The SOEKRIS DAC will prevail!!!