I met with AKM today. They say the about to be released AK4497 should get to -120 dB THD+N, however the AP2722 even with tweaks from AP can't measure that low. They are awaiting an APX555 that may be able to measure its performance. All the harmonics are way down on both chips and its really limited by noise. Unfortunately the footprint is different (needed more pins for ground and references) so new PCB's. The plus (a really silly one) is support for 8X DSD.
Good news! Whats about the AKM ADC's part where the new AK5397 struggles as may the announced x channel Hires ADC's. Also whats on the missing measurements at 384/768 kHz 😱
A good THD's & SFDR performance is always the hen and egg source question (ADC or DAC)...
Hp
The newer AP is just a little better.... a few more db's (6?). And what level is AKM's reference and what filter is applied for this number? I hope it isnt an A weighted number.
And, when are they trying to get it out?
THx-RNMarsh
Last edited:
Good news indeed! Did they say anything about when they will release it? Their web page just says "coming soon". Although the AK4490 does perform quite well the AK4497 could probably improve both the distortion and the noise of the generator part.
Since my design is now made with a DAC + postfilter module for the generator part, it should be fairly easy to test the AK4497, when it becomes available.
Since my design is now made with a DAC + postfilter module for the generator part, it should be fairly easy to test the AK4497, when it becomes available.
Their input is the AK4497 will be as far as a voltage out DAC can go. the AK4499 will be current out, with all the attendant problems getting it to work.
The AK4497 should be sampling by the end of the year and production mid next year.
The AK4497 should be sampling by the end of the year and production mid next year.
Will they/do they plan to have a it available on a pcb demo board to test/evaluate? That will be the fastest way to see what it does or its potential. And, any 'funnies' that go on with interfacing etc etc etc.
THx-Richard
THx-Richard
Last edited:
Verifying the ES9018 performance is very challenging. its "better" than an AP2722, which makes verification extremely difficult. The AK4497 has the same problem so the ******* match doesn't have a real path to a winner. The AK4499 will also have the same issues.
Personally I prefer the AKM dac's and ADC's for a number of reasons but at this level its all academic.
Personally I prefer the AKM dac's and ADC's for a number of reasons but at this level its all academic.
I just ran across this device: HiFime UH1-Digital 384kHz USB to I2S/DSD/COAX interface (no DAC) based on this chip: http://download.esm-cn.com/ATTACHMENTS_201408_101635_20140829860_pdf/SA9227-v1.3a.pdf which is interesting if it makes the DAC/ADC interfacing easier. It claims to be isolated which may bring issues around the clock distribution. I think those can be overcome pretty easily myself. however, before I "invest" I would like to know if its a worthwile option to consider. It does have a Win 10 compatible driver that works up to 384 KHz.
Have you verified the performance of the ES9018? There seems to be very little objective measurements available. At least I haven't seen much.
I measured the ES9018 some time ago and cannot find any saved results. I do have measurements of the mobile version of the ES9018-2m I made (below). Its good but not outstanding. The source was Alex's Altor generator. The full ES9018 is very good but you need to do a lot of tweaking and fiddling to get results.
Attachments
Last edited:
I have isolated, up to 384kHz XMOS-based USB<->I2S, which is used in my ADC/DACs. Without any problem with clock distribution snd it's jitter.
Hi Damien,
As I see - you use notch filter to avoid ADC problems at high level signals. But the result are very closed to what I get with AK4490 -> AK59394 without the notch:
Attachments
can you guys pls add noise data to 100KHz to see amy added out-of-band noise as well?
THx-RNMarsh
THx-RNMarsh
The AK5394A has a noise corner that starts around 80 KHz, almost unique in ADC's. Which HF noise plots do you want (assuming they are in the collection)?
Hi Jens; when it comes to usb connection, the minidsp usb streamer is one choice but as you know we need to tweak it for injecting a clean master clock. Looking at this, it apears that the HifiMe UH1 interface is pretty cool and provide both input and output with isolated i2s (or spdif)... you may have a look. Also this is ran by the SA9227 which you can easily find on aliexpress with its eeprom ... then you might consider putting it on your product once you go in the final solution, just some cents 🙂
Hi maxidcx,
The SA9227 does look like an interesting alternative. I will take a look at the available documentation and contact the manufacturer for more information. I need to get the driver, EEPROM parameters and the component supply under control. If the analyzer becomes a real product I don't want to rely on aliexpress.
The SA9227 does look like an interesting alternative. I will take a look at the available documentation and contact the manufacturer for more information. I need to get the driver, EEPROM parameters and the component supply under control. If the analyzer becomes a real product I don't want to rely on aliexpress.
Saviaudio seems to have a poor customer service, or their product are not well documented. I don't understand the sense of their behavior.
If it could be of help, there is a distributor in Italy, Silicon Solution Stream.
Inviato dal mio YOGA Tablet 2-1050L utilizzando Tapatalk