Yes, it's a simulation.
I usually start with simulation, then I check in the real world.
Simulation is often a good start point to predict the result, overall in analog circuit, and the final stage of a DAC (the switches) is pure analog.
Simulation clearly shows the effects of the difference between the rise and fall time of the mosfet pair, and also shows the effect of the correction circuit.
Just one resistor and one capacitor and the inrush current (mosfet gate and power rail) almost disappears.
Finally, IMHO, simulation helps to save time and several proto board.
And a few times I agree with your statement… It doesn't matter how you load the registers, overall if you stop the bit clock as soon as the registers are loaded, avoiding the bit clock to create interference with the latch enable. Sigma-Delta aside, the crucial timing is the latch of the switches, and obviously the analog behavior of the switches themselves.
What mosfet pair are you referring with 1 ohm Rds On, the Rohm EM6M2?
I did not know that MSB uses mosfet pair as the switch, the only picture I saw shows a battery of 374 or similar (attached).
Attachments
Last edited:
Don't remember, just remember I have seen them.... I wouldn't use the EM6M2, find one with a higher Vgs to reduce the shoot through.
I have a picture of a board, no labels, though it was MSB, it seems to be Rockna, which started out using MSB modules, then designed their own, their RD-0 use mosfets, their RD-1 use logic chips....
Wavedream DAC | Rockna Audio
Btw, can't we agree on that MSB Tech is the reference ? And they use logic chips. And btw, Rockna is using the Si570....
Last edited:
Using a word 'subjective' as opposed to 'objective' leads to the presumption that measurements give objective results. From one point of view it is true, from the other is false. Unfortunately our brain process sound differently. We hear things that DSP engine ignore completely and in reverse. What is objective for DSP is subjective to our brain and in reverse. Personally I will never buy a product from a designer who claims that our hearing is subjective. Take seriously this note, you will sell more.
SBAF administrators have a nasty habit of assigning 'funny' names to some users who are not in line with their pack dog mentality, so I am trying to stay away from such community. Their agenda is actually inline with my opinion about Topping products, so they took my attention. In result I was also following discussion about Audio GD before I decided to purchase R2R11. You will be really disappointed what I tell you now.
There is one guy who had previous experience with the first version of A-GD DAC module. The guy is always dominating discussion about newer A-GD products like R2R11/R1/R28 and there are members who have to remind him that he had absolutely no experience with these products. Read carefully or trust me, you will better spend time on making money.
As for now you are spending so much time trying to convince us that A-GD product is inferior to yours, but you failed to bring a proof for your accusation that A-GD website is "full of incorrect statements and mixups".
Regarding your opinion on the CPLD, it is a blanket statement. You will be more convincing if you bring more information how you would like to implement it. For now it is just unsupported claim.
Parallel processing is a term used on the A-GD website. Not sure what exactly it means. In scope of my visibilty it is just drivers. If you say that 8 or 9 times faster clock required for shift registers do not affect the ladder, I have no much experience to argue on this, not an important issue.
Using a word 'subjective' as opposed to 'objective' leads to the presumption that measurements give objective results. From one point of view it is true, from the other is false. Unfortunately our brain process sound differently. We hear things that DSP engine ignore completely and in reverse. What is objective for DSP is subjective to our brain and in reverse. Personally I will never buy a product from a designer who claims that our hearing is subjective. Take seriously this note, you will sell more.
SBAF administrators have a nasty habit of assigning 'funny' names to some users who are not in line with their pack dog mentality, so I am trying to stay away from such community. Their agenda is actually inline with my opinion about Topping products, so they took my attention. In result I was also following discussion about Audio GD before I decided to purchase R2R11. You will be really disappointed what I tell you now.
There is one guy who had previous experience with the first version of A-GD DAC module. The guy is always dominating discussion about newer A-GD products like R2R11/R1/R28 and there are members who have to remind him that he had absolutely no experience with these products. Read carefully or trust me, you will better spend time on making money.
As for now you are spending so much time trying to convince us that A-GD product is inferior to yours, but you failed to bring a proof for your accusation that A-GD website is "full of incorrect statements and mixups".
Regarding your opinion on the CPLD, it is a blanket statement. You will be more convincing if you bring more information how you would like to implement it. For now it is just unsupported claim.
Parallel processing is a term used on the A-GD website. Not sure what exactly it means. In scope of my visibilty it is just drivers. If you say that 8 or 9 times faster clock required for shift registers do not affect the ladder, I have no much experience to argue on this, not an important issue.
Parallel processing should mean that they drive directly all the DAC switches by the CPLD, if there are enough I/O ports, to avoid using shift registers.
IMHO, if the bit clock is stopped before the DAC latching there are no differences. The issue could be with continuous bit clock due to the interference on DAC latching.
Parallel or serial loading aside, IMHO the best practice should be isolating the bit clock from the DAC, I would avoid any signal coming from FPGA/CPLD to feed directly the DAC switches. In other words, only the latch enable should come directly from the master clock using appropriate divider.
Yes, MSB should be the reference, although I have not yet listen to one of their DAC.
The problem using tiny logic, IMHO, is that they are too fast, with a serious risk of ground bounce and ringing, that finally means noise.
I remember Nicolae bought a pair of Laptech crystals from me, I don't know what he is working on.
About the Si570 I can agree if you need absolutely to sync audio and video avoiding a large FIFO buffer, but since I don't need so I'm happy to avoid a programmable oscillator with a poor phase noise. I would avoid any PLL in digital to analog conversion, overall with multibit R2R DAC, otherwise the reference becomes Sabre rather than MSB.
Thanks for clarification. Shift registers version brings complication to the clock design, it has to be stopped before latching, so a simple 8x clock division no longer applies. And it has to be isolated from the master clock. Solution by Kingwa is actually appealing to me more because of simplicity.
In both cases we have a lot of switching noise of the fast logic and your idea of using slower MOSFETs makes lot of sense. I think a difference in the on/off switching times can be compensated.
Few discrete r2r in the world till today, has anyone had enough cyriosity to listen to and/or study discrete R2R Dac of ECdesigns and TotalDac, especially according the way they use clock.
An interesting thing as well is TotalDac still use NOS while being in the said list of the few commercial DACs sounding goods. Don't want to be unpleasant, Soekris core DAC used in some commercial products are not yet in that short list, even with Lampizator that are always took for their "tube sound" which put it not in the world of hih fidelity...
My two cents... Denafrips, I'm not impressed after a fast listening... though always to be pondered with what devices are around...
There is no doubt there are different price spent in the parts, Dam dac were made to be cheap and from my experience many diy dac I made from here and elswhere sounds better according to my tastes and some of musicians thread. Denafrips, well bench of caps and low esr never proved to be an efficient way to make a device sounding good, nore the Lipo cells when putted everywhere in the dac but few rare areas. But who I am, I don't like Sabre ESS chip sound either nore I developp Dacs by listening on Sheinezer Tibet mountains spl curve either....
At the end they are not, whatever the technic, EQed the same way, so... the quality is in the eyes of the beholder, or the eyes, or maybe the ears.... Well, subjective thoughs fro subjectives ears... just two cents, YMMV as usual...
An interesting thing as well is TotalDac still use NOS while being in the said list of the few commercial DACs sounding goods. Don't want to be unpleasant, Soekris core DAC used in some commercial products are not yet in that short list, even with Lampizator that are always took for their "tube sound" which put it not in the world of hih fidelity...
My two cents... Denafrips, I'm not impressed after a fast listening... though always to be pondered with what devices are around...
There is no doubt there are different price spent in the parts, Dam dac were made to be cheap and from my experience many diy dac I made from here and elswhere sounds better according to my tastes and some of musicians thread. Denafrips, well bench of caps and low esr never proved to be an efficient way to make a device sounding good, nore the Lipo cells when putted everywhere in the dac but few rare areas. But who I am, I don't like Sabre ESS chip sound either nore I developp Dacs by listening on Sheinezer Tibet mountains spl curve either....
At the end they are not, whatever the technic, EQed the same way, so... the quality is in the eyes of the beholder, or the eyes, or maybe the ears.... Well, subjective thoughs fro subjectives ears... just two cents, YMMV as usual...
May I ask what actually satisfies you as the list of those DACs which dont work for you is pretty long ? Seriuously, just curious.
I went last year to the High End in Munich and listened to MSB etc....well, I went home and enjoyed my modified Soekris DAM1941 with DHT outputstage a lot. I respect the work of Soekris to make stuff available for the DIY community.
It would be nice, if there be a unit for the hardcore diy-guys where you give more options which resistors to use or clocks etc...just a bit more modular on the hardware side would be nice...for people who can go the last mile themself...and to reallocate costs on those parts, which comes with board...but anyway, I really appreciate the ability to work with R2R Dac modules and give Soekris a lot credit for making up the way for many DIYlers.
I went last year to the High End in Munich and listened to MSB etc....well, I went home and enjoyed my modified Soekris DAM1941 with DHT outputstage a lot. I respect the work of Soekris to make stuff available for the DIY community.
It would be nice, if there be a unit for the hardcore diy-guys where you give more options which resistors to use or clocks etc...just a bit more modular on the hardware side would be nice...for people who can go the last mile themself...and to reallocate costs on those parts, which comes with board...but anyway, I really appreciate the ability to work with R2R Dac modules and give Soekris a lot credit for making up the way for many DIYlers.
I believe one have to find the trade offs he much likes. Nothing much to complain if you like the dac you daily uses.
Btw, were there dacs, no necessarely r2r, you liked more than the MSB in the M. Show ? Too bad all the hifi events are cancelled all over the world (but it's ok for health of course)...
Well all the different are in the nature and make a world...
Btw, were there dacs, no necessarely r2r, you liked more than the MSB in the M. Show ? Too bad all the hifi events are cancelled all over the world (but it's ok for health of course)...
Well all the different are in the nature and make a world...
Last edited:
Which one do you use? I know the game of trade offs and built by now 10 dacs or so.
Its always difficult to judge as you listen on a show to a whole chain, so its not clear which component does what.
I found a dac for 20k quiet good coming from switzerland but using akm4497 i believe... than i learned a new chip is coming akm4499, but i did not saw any boards for those yet, the prototype board is now 1000 euro, which is not pleasant anymore...for an experiment too much.
Its always difficult to judge as you listen on a show to a whole chain, so its not clear which component does what.
I found a dac for 20k quiet good coming from switzerland but using akm4497 i believe... than i learned a new chip is coming akm4499, but i did not saw any boards for those yet, the prototype board is now 1000 euro, which is not pleasant anymore...for an experiment too much.
I am sorry refering to the old post, but it looks like the Audiophonics website is wrong. According to the picture from manufacturer website (I checked also some reviews), it is a bigger chip, XC2C128.
???响
Your contribution is valuable, but the above is a blank statement made repeatedly. There is a DIY DA-8 module, number of commecial DACs from Audio GD that sll sound very good R1, R8, R7, R2R11, R28, R27, a proof of a concept. All, even the most expensive ($3k) model use the same CPLD chip.
If you are unable to do it yourself, it doesn't mean it cannot be done, but you continue to post this completely unsupported claim here in this and also in the other threads like this: R-2R DAC Kit
If you are unable to do it yourself, it doesn't mean it cannot be done, but you continue to post this completely unsupported claim here in this and also in the other threads like this: R-2R DAC Kit
sajunky, you're clearly not an engineer. As I said, the DA-8 module by itself clearly do not have the resource to do any calibration, but a systems using it can have additional logic, ADC and memory resources on the mainboard, enable it to do calibration.
If you keep insisting that the DA module itself can do it, please provide docs, or explain how it do it.
And it's you that keep posting unsurported claims, or more like repeating some chi-fi company's wild claims everywhere. I'm just trying to correct the wrong claims with facts....
If you keep insisting that the DA module itself can do it, please provide docs, or explain how it do it.
And it's you that keep posting unsurported claims, or more like repeating some chi-fi company's wild claims everywhere. I'm just trying to correct the wrong claims with facts....
@soekris. This is to confirm that what you say is true, my bad sorry. DA-8/DA-7 modules do not do calibration itself. Therefore my R2R11 DAC has no auto calibration. All other models have additionally FPGA chip (R1, R8, R7, R28, R27) and do auto calibration through FPGA. It is stated clearly on website.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.