The Lua language threw down the gauntlet years ago, making it really easy to call C libraries from Lua. The great thing about open source is that everyone can just borrow the idea and improve on it, so foreign function interfaces in Python have progressed pretty far. There are tools that automate a lot of the "glue" of wrapping Python calls around C library functions. You might want to google around for "python foreign function interfaces". Here's the top hit when I googled around for it:
Goals — CFFI 1.14.0 documentation
Getting back to dacs for a moment, looks like the dac subforum is about as slow as its ever been since I joined the forum. There is still some activity with older dacs but nothing approaching modern SOA.
What do others think, do we need a basic AK4499 board project that actually follows though to completion (given that there have been a few false starts so far)?
What do others think, do we need a basic AK4499 board project that actually follows though to completion (given that there have been a few false starts so far)?
I am a little confused, what exactly "basic AK4499 board" would mean? The 128-pin HTQFP is hardly DIY part for many of us, not mention the price.
Actually those "older dacs" are far from being outdated. IMHO there is a trend to build crappy implementations with very modern parts, and an opposite minor trend to develop novel implementations with not so recent parts. the later requires much higher degree of engineering skills, I guess.
sure everyone wants "an excellent sounding low priced DAC to happen"
is it possible after all? for sure it won't be 4499. for a while I am listening a DAC based on 2xUDA1334 in differential mode, designed by Nazar. it is an excellent sounding and relatively low priced DAC indeed.
Actually those "older dacs" are far from being outdated. IMHO there is a trend to build crappy implementations with very modern parts, and an opposite minor trend to develop novel implementations with not so recent parts. the later requires much higher degree of engineering skills, I guess.
sure everyone wants "an excellent sounding low priced DAC to happen"
is it possible after all? for sure it won't be 4499. for a while I am listening a DAC based on 2xUDA1334 in differential mode, designed by Nazar. it is an excellent sounding and relatively low priced DAC indeed.
Last edited:
Name it. I would like to know for my personal reference.
I’m interested in your subjective opinion (really). Objective measurements not required.
I agree. It must be a reason that we talk about ES9038Q2M in this topic and not about ES9038PRO.
Why target budget? - we all know our little DiY builds end up pimped to the limit... better start off with a the best we can get... then pimp out to the max no point trying to make a silk purse from a pigs ear...
I could talk with AKM to see if I can get hold of there pending 2 chip solution for a DiY kit - in fact we would use in dual mono...
Combine with XMOS 208 or maybe the 216 for future expansion...
I have a full SMT line inhouse so maybe it would be possible to run a few PCB's with the "hard parts " already mounted to the PCB....
We could have a simple API from XMOS over serial UART to allow a simple front panel MCU interface or drive leds directly...
Anybody here have experience with XMOS? (its a real pig) - my software guy is overloaded so I rather not distract him...
no point trying to make a silk purse from a pigs ear...I could talk with AKM to see if I can get hold of there pending 2 chip solution for a DiY kit - in fact we would use in dual mono...
Combine with XMOS 208 or maybe the 216 for future expansion...
I have a full SMT line inhouse so maybe it would be possible to run a few PCB's with the "hard parts " already mounted to the PCB....
We could have a simple API from XMOS over serial UART to allow a simple front panel MCU interface or drive leds directly...
Anybody here have experience with XMOS? (its a real pig) - my software guy is overloaded so I rather not distract him...
Last edited:
Ahem. I did say basic, not cheap, for obvious reasons. The dac chip is not cheap in small quantities, and its not the only thing needed.
By basic I was thinking in terms of a board with the dac chip, roughly 7 or 8 three terminal regulators, I/V opamps, and some caps and resistors. Such a board might only accept I2C input. The board might or might not include 2 or 4 Reference Voltage regulators (depending on stereo-only or 4-independent channels). Thing about Reference Voltage regulators, they do have a substantial effect on sound quality. Eval board Jungs are pretty good if fed clean input power, although IME slightly better is possible.
A board like that along with some external power supplies, say, maybe a JLsounds I2SoverUSB, and an Arduino would be enough to get something pretty good going. Probably good enough to easily beat the last generation of ESS dacs, including Benchmark DAC-3.
That said, IMO beating Topping D90 sound quality would require more than the minimum (Accusilicon clocks sound better than the NDK SDA that come with I2SoverUSB, might be the main thing, but I2SoverUSB supports external clocks too). In fact, no guarantee one could beat D90 sound quality for less than D90 cost. Maybe yes, maybe no. Don't know for sure at this point. Might depend on if one had to go out and buy all new external power supplies and stuff to use for an AK4499 project.
EDIT: Oops! Cross posted with JohnW.
By basic I was thinking in terms of a board with the dac chip, roughly 7 or 8 three terminal regulators, I/V opamps, and some caps and resistors. Such a board might only accept I2C input. The board might or might not include 2 or 4 Reference Voltage regulators (depending on stereo-only or 4-independent channels). Thing about Reference Voltage regulators, they do have a substantial effect on sound quality. Eval board Jungs are pretty good if fed clean input power, although IME slightly better is possible.
A board like that along with some external power supplies, say, maybe a JLsounds I2SoverUSB, and an Arduino would be enough to get something pretty good going. Probably good enough to easily beat the last generation of ESS dacs, including Benchmark DAC-3.
That said, IMO beating Topping D90 sound quality would require more than the minimum (Accusilicon clocks sound better than the NDK SDA that come with I2SoverUSB, might be the main thing, but I2SoverUSB supports external clocks too). In fact, no guarantee one could beat D90 sound quality for less than D90 cost. Maybe yes, maybe no. Don't know for sure at this point. Might depend on if one had to go out and buy all new external power supplies and stuff to use for an AK4499 project.
EDIT: Oops! Cross posted with JohnW.
Last edited:
right, AK4191 delta-sigma modulator and the AK4498 DAC is kind of next winding in the developmental spiral, even though it is hard to imagine a DIY kit based on them.
IMHO this chase after the "very last and the best" DAC is simply pointless. one of the best DAC I have heard RME ADI-2 DAC is AK4493 based. can one build something like this at home? NO, but there are open source projects which can get you close to this. very well engineered and perfectly DIY. for instance one can get cheap and neat PCB for this and all the related info from :
DAC AH-D6 ver 2.1 - Share Project - PCBWay
the real charm is that the design is originating from a long development effort, exceeding data sheet or "manufacturers recommendations" by miles.
( I have neither a relation to the project above nor gain any benefit from it)
After having personally tried Amanero, LDOs, etc, personally can't believe that is in the same SQ class as what I have been talking about. Sorry, I mean no offense by that, just telling it the way it looks from here.
Also, it might measure great. I hope most of us here know better than to think that's all there is to it (sticking my neck way out again
).I would certainly be up for such a project! Of course, supposedly the eval board version is due out this quarter, although production quantity chips not for another year.
After having personally tried Amanero, LDOs, etc, personally can't believe that is in the same SQ class as what I have been talking about. Sorry, I mean no offense by that, just telling it the way it looks from here.
Also, it might measure great. I hope most of us here know better than to think that's all there is to it (sticking my neck way out again
Wrong. If you would care at least to follow the links and take a look at the schematic you would notice that at all the critical places there are discrete regulators, filtered reference and opamp based. Amanero has galvanic isolation and can be run as slave, unused clock forced down, an original LPF, etc...
The choice of part values and overall design has been verified in previous iterations with AK4490. No offense by that, but this thing is real and could be built, in contrast to something else. How it sounds? You would be surprised, or rather no?
If I propose to AKM I act as a "gate keeper" (so they don't get inundated with endless support questions) they might support a DiY project - it would give them good exposure...
The projects interesting if we use the AKM 2 / 3 chip solution as I can compare to my discrete DAC designs - always good to see things from different angles...
I'm thinking that the USB / DAC section would be mounted on a separate daughter board - allowing the DiYer to use there own analogue stage (Opamp / Tube or transformer etc)...
Connection via IDC headers to allow close connection to the Analog stage.
Only the XMOS firmware concerns me - I really cannot afford to use my own resources as they are overloaded with existing projects... I can do everything else...
The projects interesting if we use the AKM 2 / 3 chip solution as I can compare to my discrete DAC designs - always good to see things from different angles...
I'm thinking that the USB / DAC section would be mounted on a separate daughter board - allowing the DiYer to use there own analogue stage (Opamp / Tube or transformer etc)...
Connection via IDC headers to allow close connection to the Analog stage.
Only the XMOS firmware concerns me - I really cannot afford to use my own resources as they are overloaded with existing projects... I can do everything else...
Last edited:
Typical CD player from 2000 and on. Typical stereo or home theater receiver from 2000 and on. You may already have one.
Why do you care?
Hey guys, I have a question about building an AVCC supply with an opamp. What are the drawbacks/benefits of having a symmetric power supply vs Vee pin grounded for the opamp?
My reasoning is like this:
most opamps have a a little less PSRR on the negative supply pin than on the positive and also have no real ground reference through any of their pins (OPA1622 being an exception). So would it be better if the negative supply pin would be grounded instead of sitting at -15V?
Would like to hear some opinions on that...
My reasoning is like this:
most opamps have a a little less PSRR on the negative supply pin than on the positive and also have no real ground reference through any of their pins (OPA1622 being an exception). So would it be better if the negative supply pin would be grounded instead of sitting at -15V?
Would like to hear some opinions on that...
Normally you can use with grounded negative VCC on the Opamp, just depends on the voltages on the op-amps inputs (all being positive for starters) - and within the input operating range when used with Grounded Neg VCC rail....
If you post a circuit with the opamp you intend to use then I can confirm operation for you.
I'll take your word over what the schematics posted there show (looks like some errors in them, and legibility of text is not very good).
Don't want to argue about it, we have too much of that around the forum already.
Bottom line for me is that I like AK4499 a lot, but don't know exactly how good it can sound at best. Still working on trying to find out, but taking my time doing it. Another experiment is in the works only this time I have learn Circuit Studio first. Something to do while home bound I guess
Last edited: