ADCs and DACs for audio instrumentation applications

Member
Joined 2004
Paid Member
The AKM chip AK5394A sold quite well as these go for a long time, its been around for over 15 years. I was told its demise was due to the package being discontinued. It seems they would have needed to redo the chip to fit an alternative package. The AKM fire is having a major impact and we can't predict what will happen next there but my guess will be the car products getting priority. Toyota will be applying incredible pressure among others.
 
evals at the Audio Chips - Ismosys too... on backorder :D

If I could have a dime for each time I've seen "Available on back order" from ESS I would be rich for now. Let's hope they'll change their stupid business model to keep everything under wraps. Anyway, $10 a pop, even the price sounds ridiculous, given the current high end (audio or industrial grades) ADC prices. Perhaps $10 in 1000's quantities :D.
 
The AKM chip AK5394A sold quite well as these go for a long time, its been around for over 15 years. I was told its demise was due to the package being discontinued. It seems they would have needed to redo the chip to fit an alternative package. The AKM fire is having a major impact and we can't predict what will happen next there but my guess will be the car products getting priority. Toyota will be applying incredible pressure among others.

Sorry, doesn't really hold water... The AK5394A came in the standard 28pin SOP, which is widely used in the industry (and with other AK products as well, for example the AK4112 audio receiver).

Occam's razor principle suggest that this product did not have a market large enough to justify the cost of keeping it in the catalogue. Same as for so many other demised chips, including discretes.
 
Member
Joined 2004
Paid Member
AKM told me there was something odd about its specific package that had been discontinued, I forgot what. Most of these premium audio chips are barely break even once you factor in the mask. The volume on the chip was still pretty good but it was old and a much larger die and expensive. I gather AP gave them an earful over the EOL as well.

I have some dead EMU boards with a handful of those chips I'll harvest.

In any case time to move on.
 
It's a larger pitch than the others, though, it's actually a SOP not even SSOP according to the datasheet. Not saying it's a big factor, but it has a 1.27 mm pin pitch, double that of CS5381. The long dimension of the chip is 18.7 mm! The CS5381 is only 7.8 mm. Power consumption is 2x-4x also. Those factors are probably why it didn't get more design wins in pro interfaces. It did have a decent number, but CS5381 was definitely more popular.

On the AD1955, certainly AD has a very good track record of keeping old parts alive. I wonder if AD1955 internally used a lower voltage and had level translators and an LDO built in? There are no signs of that on the block diagrams, though. You're right on the sales - I have no idea who would still be buying that chip. AD1955 saw very little adoption in the pro audio industry as far as I can tell, and it wasn't even that popular in the Hi-Fi market. Benchmark opted for the cheaper AD1853 in the DAC1.
 
Last edited:
I have one of those EMU boards, it is fully functional, one of the best sound cards I've ever seen (Juli@ was at par, while the Lynx was a smidge better) and if memory serves it was $150 only, however it is also useless today because of the lacking drivers for Win x64. They released some beta drivers years ago, but they were an abomination. Thanks for the suggestion, I'll harvest the parts, although I am not about to design anything new around them, just keeping them in my spare parts repository.
 
Member
Joined 2004
Paid Member
I didn't do in-depth measurements, but when I turned on phantom power, a lot of ICs smoked, so I recapped the entire 1820m and replaced some of the JRC2068 with OPA1612, except on the headphone output. That box got really hot, probably greatly contributes to the demise of the cheap caps.
 
Actually anyone can order from ISMOsys and the price is per unit. I had 4 ES9822 Pro arrive this morning and they took about a week to get here from the date they were ordered. I too was surprised at how cost effective these ADCs are given their reputed performance.

Whether or not you are eligible to sign an NDA for the data sheets is another thing but that's up the people at ESS. It doesn't hurt to try. If you have a company email/name or website, of your own, you can go through that would make it easier.
 
I will have to contact the ISMOsys representative and ask what kind of data I'm allowed to show. I can't see why my own measurements would be against the NDA as stereophile can perform measurements of gear containing ESS parts.

Without rereading the NDA I do remember there being some section on this and it wasn't promising. Needless to say if it performs well then I can surely say it performs well, go out and get them for yourselves.

What I can say is this chip is very new, even to ESS. I get the impression that, following the fire at AKM, they rushed to get this to market to fill the sudden gap.
 
What I can say is this chip is very new, even to ESS. I get the impression that, following the fire at AKM, they rushed to get this to market to fill the sudden gap.

Really promised since x years a new ADC to come..

Stereophile is not under NDA, while testing official gear.. while developer is on NDA and may not able to share the good or even ugly behavior :eek:

Funny will be the Hyperstream modulation against the SDM of AKM. While the noise on ESS is almost a step lower. This means also it is the inverse function of the ESS Hyperstream DAC!

I had 4 ES9822 Pro arrive this morning and they took about a week to get here from the date they were ordered. I too was surprised at how cost effective these ADCs are given their reputed performance.

So the question rises: Do you build your own ADC PCB (for measurement or recording) as based on Evaluation, somewhat else and/or may share the purpose of..

Some are already nervous, while AKM ADC get's rare on sky :goodbad:

Hp
 
Indeed, the ADS127L01 looks almost to good to be true :D. I would think nobody in the audio world looks at the industrial grade ADCs, since they cannot claim any outlandish performance they could get due to all sort af audio grade tricks, like noise shaping out from the audio band, etc... Plus the lack of a nice "audio grade" story about TI, log story short, the TI parts look boring and bland.

I have to admit, after quite some experimenting with, I don't really understand how this ADC is working in frame mode, in particular at 384KHz. At 384KHz it works perfectly with the master clock set to 12.272MHz, oversampling set to x32 and the bit clock set at 24.576MHz. Question is, who's feeding the output at that bit rate, doesn't make much sense to me, I would expect the master clock to be larger (double) or at least equal to the bit clock. Must be some sort of double buffering inside that is not mentioned in the data sheet.