Hello Shinja,
thank you very much! But what is the benefit, if all other parts are still exposed to power sipply noise...? Does anyone have a datasheet?
thank you very much! But what is the benefit, if all other parts are still exposed to power sipply noise...? Does anyone have a datasheet?
I wonder whether cascading the two 1-bit DACs might be some clever means of achieving an net higher sigma-delta oversampling ratio, and a higher order noise-shaping.
By going analog in between? We did this this already 30 years ago,
when there was only ADAT as source and Dat as Master/target but no digital mixing consoles available...
when there was only ADAT as source and Dat as Master/target but no digital mixing consoles available...
By going analog in between? We did this this already 30 years ago,
when there was only ADAT as source and Dat as Master/target but no digital mixing consoles available...
Well, that's one of the interesting aspects of DAC outputs, they're analog, yes, but thery're discretely analog. Their analog outputs essentially exist as discrete levels. For examples, an 8-bit DAC has 256 discrete levels, a 4-bit DAC has 16, a 2-bit DAC has 4, and a 1-bit DAC has 2 discrete levels. Does 2 discrete levels sound familiar? Right, it's the basis of binary digital logic. So, the output of a 1-bit DAC is actually binary. Therefore, it's potentially compatible with following digital logic circuits, such as, say, the digital inputs of a second DAC.
... aaaah, thanks, the signal is "discrete" because it has not passed current to voltage/ a low pass filter yet?
We all know the cliché of the stairstepped curve representing a "digitalized"
sine wave, would this be an accurate description of the discrete levels of the currents leaving the first dac?
And why arent they synced? Wouldn´t it be helpful if the second DAC hit the change in step exactly? Or am I assuming something wrong again?
How would the signal keaving the first DAC look like? (but my scope is old and han an upper frequency of 15mHz)...
All the best,
Salar
We all know the cliché of the stairstepped curve representing a "digitalized"
sine wave, would this be an accurate description of the discrete levels of the currents leaving the first dac?
And why arent they synced? Wouldn´t it be helpful if the second DAC hit the change in step exactly? Or am I assuming something wrong again?
How would the signal keaving the first DAC look like? (but my scope is old and han an upper frequency of 15mHz)...
All the best,
Salar
Yes, essentially, the output reconstruction filter (also known as an, anti-image filter) converts the discrete output in to a continuous one. It does this by removing the repeating ultrasonic image bands. Which are what gives the unfiltered signal it's discrete appearance.
Assuming that I understand what you're asking here - which I feel unsure of. Yes, the stair-stepped DAC chip output, visible via an oscilloscope, shows some of the discrete levels I've been talking about. Regarding a 1-bit DAC output, it doesn't matter whether those levels represent voltages or currents, as far as them qualifying as binary. However, for potentially connecting said binary 1-bit DAC levels to any subsequent binary digital logic, it is important that the levels be physically formated to be correctly interpreted by any such logic.
I've not analyzed that Sony schematic. To be very clear, I'm not declaring that Sony is implementing any specific scheme. I only intended to offer some off-the-cuff conjecture about what they might possibly be up to.
Wow I thought it was synced, but it is not! Just two flip flops cost
was not affordable? Technics also developed similar current type 1bit DACs which is synced.(japanese model SL-P2000 SL-P860 SL-P770D) And no need to say philips DAC7 system was, though it was switched capacitor type.
picture is sl-p860 .
was not affordable? Technics also developed similar current type 1bit DACs which is synced.(japanese model SL-P2000 SL-P860 SL-P770D) And no need to say philips DAC7 system was, though it was switched capacitor type.
picture is sl-p860 .
Attachments
I think things might have been probably mixed up -
I assume the first DAC is multibit / Delta sigma (all the CXDxxxx - dacs in the schematics) whereas the
second dAC appearing in all schematics, the CXA 8042 is 1-bit ...?
I assume the first DAC is multibit / Delta sigma (all the CXDxxxx - dacs in the schematics) whereas the
second dAC appearing in all schematics, the CXA 8042 is 1-bit ...?
CXA8042 is not a DAC. It is a “summing” IC of an analogue nature & a filter to get maximum dynamic range and keep distortions to an absolute minimum. Those 2 X right {and [2 X left]} differential input signals should also make those CXA8042 IC’s quite immune to the power supply rail imperfections… in theory. That being said, I really like what LC audio proposed, but ONLY the resistor summing part... I do not like their ZAP filters at all.
In fact, I was going to suggest the same thing R1 + R2 (-) and R1 + R2 (+), fed to a simple differential to single-end stage (if you really need to have RCA instead of XLR at the output)…. but then you posted the LC Audio suggestion, so I refrained from repeating the same.
You really need to check what the oscilloscope says at CXD2562Q pins 16 and 18. You could try this even with your 15Mhz scope… or get a good one and start exploring….
Check this thread as well: The Best DAC is no DAC and try to get your head around it; this will help you understand what Sony did with CXD / CXA IC pair.
In fact, I was going to suggest the same thing R1 + R2 (-) and R1 + R2 (+), fed to a simple differential to single-end stage (if you really need to have RCA instead of XLR at the output)…. but then you posted the LC Audio suggestion, so I refrained from repeating the same.
You really need to check what the oscilloscope says at CXD2562Q pins 16 and 18. You could try this even with your 15Mhz scope… or get a good one and start exploring….
Check this thread as well: The Best DAC is no DAC and try to get your head around it; this will help you understand what Sony did with CXD / CXA IC pair.
Salar, did you see a patent and the pictures in this link? CXA8042 is mainly comprised of two long tail pair circuit, which can be referred current mode logic from digital circuit perspective.
In a 1bit system, the boundary of digital and analog is vague, but the input to output of CXA8042 is saturative, so analogous characteristics does not appear from input to output.Which is digital discrete value input characteristics I think. (less saturative than typical logic ICs but similar to 74HCU04 4049U unbuffered series, 1 stage amplifiers.)
In a 1bit system, you can use 74 logics as a DAC ,as some people and Signalyst did. Thus the boundary is vague.
- Status
- Not open for further replies.