Regarding close-in phase noise effects, it seems to have been most seriously investigated for its effects on radar. Some discussion here:https://www.qorvo.com/products/d/da007548
Since I have never seen it described for audio data converters, I will take a rough stab and say I think its probably the same basic problem as for radar: A small signal close in frequency to a larger signal may be buried in noise. IME, the effect is audible.
Since I have never seen it described for audio data converters, I will take a rough stab and say I think its probably the same basic problem as for radar: A small signal close in frequency to a larger signal may be buried in noise. IME, the effect is audible.
This isn’t a good way to put it. They expect the data to be interpolated already in this mode, as you seem to know. The maximum rate the modulator runs at is 12.288 MHz regardless of how you get there. This is a 192 kHz chip no matter how you slice it, and if it performed well at a real Fs of 384 kHz or higher I’m sure they would have taken credit for it. Maybe it works fine, but I doubt the performance is good. Just because you can technically set that register for 4x and stick in 768k externally doesn’t really guarantee it will work well or correctly since I’m sure the 4x mode was designed for the opposite use case of an external DF that didn’t even do 8x at 192 kHz. Remember the time in which this chip was in development.
Last edited:
No radar aspects in the linked document what I could see. No relation described between pn and precision in detecting an object distance and direction. So just as any other oscillator producer techno babble really
in this case a light brief in phase noise... //
If you want more about radar you could try: https://purehost.bath.ac.uk/ws/port...hase_Noise_Analysis_in_FMCW_Radar_Systems.pdf
Not sure how much insight it might lend for audio work.
There are books on radar too, if you are willing to buy one.
Not sure how much insight it might lend for audio work.
There are books on radar too, if you are willing to buy one.
Last edited:
The ADS127L01 does well ("as expected", I should say) @384k with 12.288MHz and x32, see above. Bit clock at 24.576MHz.
Side note for those interested, Shaw Electronics flatly refused to sell me an ESS ES9038PRO eval/development board they distribute and have in stock (and not for cheap!). According to the sales rep. when hearing I am looking for a private sale, "ESS eval boards are not for sale other than for design development", WTF are they thinking is the board for, using it as kitchen board for slicing onions? Probably the same ESS crap the Sabre chips are wrapped in, since the very beginning, although lately one can get them from Mouser.
I guess I'll wait for AKM to resume production in 6 months and meantime resume playing with SAR ADCs and (this time) FPGAs and/or CPLDs.
BTW, I don't give a flying **** if the phase noise/jitter is audible in ADCs or in radars; this thread is about instrumentation and what matters is what the measurements show. If one wants to design an instrument by ear, it would maybe be a good idea to open a new thread.
Last edited:
There are voices raised that the very low frequency (0-10 Hz) sideband of the used oscillator has a real negative impact on both ADC as well as DAC. The "close-in" term means "close to the carrier" i.e. phase noise components of a few Hz - hence "close-in". High end oscs exhibit say -125 dBc/Hz at 1 Hz i.e. a very clean carrier.
The 12-20k "thing" is that most oscillators existing now are derived from the telecom era and the standard specification describing jitter/phase noise in that band and close-in was of no importance to these applications. It does make a difference in the very tight rf coding in mobile systems and is there a real world performance issue. I have chased dropped calls due to passing trains in base stations exposed to vibration in my days
Hope that helped a bit
Thanks, now it makes some sense. I'm still struggling to see the relevance for audio, but hey, that's just me
.
What order of magnitude are we talking about. If its well below ear masking thresholds how can it be audible?
Yeah, might be some confusion, I was talking about PCM1792A in response to Terry.
WRT interpolated -> what else would it be? **External DF** mode is pretty self explanatory.
Yep.
It depends what you mean by 'well'. For audio playback at 384kHz SR with ext DF it may perform as well as many other audio DS DACs. As I have already stated have to suck it and see.
For 200kHz sine wave - that's another thing altogether.
You seem pretty confused about a lot of the functionality of this DAC. It's all in the data sheet and the rest is knowing how various people have implemented the DAC. All I can say is these chips are used in some very high
performance DACs but they are not necessarily using the 'go to' settings.
TCD
No, I understand perfectly how this DAC works and how to use the external filter mode. I suspect you are misreading my post, oh well. Anyone who feeds it with true 384 kHz or 768 kHz Fs data (via external filter) should not be surprised if they get poor performance.
The entire purpose of the external filter mode is to allow you to input data that has been through your own 8x interpolation filter, it's not to feed in > 192 kHz data at a lower external oversampling ratio. Might be fine for audio / listening, probably not for instrumentation, which is the purpose of this thread.
The entire purpose of the external filter mode is to allow you to input data that has been through your own 8x interpolation filter, it's not to feed in > 192 kHz data at a lower external oversampling ratio. Might be fine for audio / listening, probably not for instrumentation, which is the purpose of this thread.
Last edited:
I don't think what people report hearing when lower phase noise clocks are used is below masking thresholds. Also as you may recall, the thresholds are estimates of an average value for a population, not a hard limit.
I would suggest you see if you can find someone who can swap the clocks in a dac in your presence and you can find out for yourself.
Has anybody seen the datasheet for the new CMedia CM6635 chipset? For me, as an XMOS architecture and power "hater", it appears as a viable alternative as an USB Audio 2.0 controller, and apparently also has flash memory on the chip. But without the datasheet and an updated SDK it is not an option for us mortals... Or does the same CM66xxA SDK and documentation apply to the new CM6635?
I also see a new EVB board is available at Symmetry, anybody seen the specs and schematics for this board? I have the CM6631 EVB, what are the differences?
I also see a new EVB board is available at Symmetry, anybody seen the specs and schematics for this board? I have the CM6631 EVB, what are the differences?
The SDK does not apply, as far as I am aware. I'm sure it's quite similar, but there is no openly available firmware framework like the CM6631/2A. It does look ideal for a 2ch in/out system without the driver licensing and power consumption annoyances of XMOS. I was thinking of using it for a project and I messaged forum member tdtsai, who had an inside contact, but sadly they are no longer with C-Media.
After scouring the Internet, I found on an obscure server the CM6635 Eval Board schematics (they use an ESS ES9038Q2M DAC for line output, a TI PCM1792 DAC for speaker output, a Cirrus CS5346 ADC for microphone in and a Cirrus CS5381 ADC for line in), and the Eval Board manual (pretty much nothing useful inside). Also a so-called CM6635 datasheet that has nothing inside except the chip DFN64 pinout.
Since I don't know if the information is subject to an NDA I won't put the information here yet... Anyway, the crucial part which is the software framework is still missing...
Since I don't know if the information is subject to an NDA I won't put the information here yet... Anyway, the crucial part which is the software framework is still missing...
Have you asked IVX about the issues as that is where the schematics probably comes from? https://www.diyaudio.com/forums/dig...20-choice-usb-audio-bridge-2.html#post6144221
^^^
Don't think so, at least not us mere mortals. It is possible ESS has shipped some samples to potential major customers in the industry (I would wonder who needs such an ADC in their audio products, though). However, the ESS previous record shows they are very slow to make their products available through distributors, begin by requiring an NDA to get even a datasheet, keep evaluation boards as a tight secret available only to the industry, etc... It could be years before we'll see these ADCs @Mouser, and a meaningful datasheet a click away, if at all.
It could also be just a "me too" product that would in fact never reach mass production, it happened before with AKM, which obsoleted their best ADC product.
Don't think so, at least not us mere mortals. It is possible ESS has shipped some samples to potential major customers in the industry (I would wonder who needs such an ADC in their audio products, though). However, the ESS previous record shows they are very slow to make their products available through distributors, begin by requiring an NDA to get even a datasheet, keep evaluation boards as a tight secret available only to the industry, etc... It could be years before we'll see these ADCs @Mouser, and a meaningful datasheet a click away, if at all.
It could also be just a "me too" product that would in fact never reach mass production, it happened before with AKM, which obsoleted their best ADC product.