CS8420 is lousy?

[inatightspot]

Other than the "glitch" while using this ASRC in software mode what exactly is the problem sound-wise with this chip?
It sounds okay to me and from a theoretic standpoint I find the engineering backing this device to be beyond reproach.
Can anyone here confirm the reasons for the lacklustre performance of this convertor with hard evidence?

Just curious,
Jim

[JohnW]

Originally posted by inatightspot
It sounds okay to me and from a theoretic standpoint I find the engineering backing this device to be beyond reproach.

If Your happy with the sound quality and performance - then why be concerned?

[gmarsh]

The CS8420 is unbelievably useful... AES transmitter/receiver and SRC integrated into a single chip, and the cost is excellent.

The audio performance isn't that good by today's standards - the cheap AD1895/SRC4190 chips beat it, and the AD1896/SRC4192 chips completely humiliate it. It also has an upper limit of 96KHz... but for CD quality applications, it's adequate.

Now the bad part: there's so many stupid little bugs in the CS8420 that I won't design it into anything. Here's what I've personally encountered:

- the AES transmitter won't start up unless you connect an AES input, even for a moment.
- unplugging/plugging an AES input can cause the chip to put out heavily distorted audio
- the reset workaround for the unplugging/plugging trick doesn't work 100% of the time. Sometimes resetting the chip can throw it into the distorted state.

And I learned these things by designing the CS8420 into a board then having to scrap the boards and redesign without the chip, which leaves me a bit bitter. And a couple weeks ago, I was on the phone with someone from another audio equipment company who shared this gem: "I had to get on a f---ing plane because of a CS8420."

For the home experimenter, the CS8420 bugs are tolerable. When you're designing professional audio equipment which has to work 100% of the time with no user interaction, the chip is unacceptable.