current digital to analog converter chip compatible with cirrus logic CS4390KSEP JNAWEA9834?

[mhelin]

There seems to be difference in the value of the filter resistor from CS4290's VA+ to VD on those machines. DM Pro uses just a 2 ohms resistor, Lexicon 5.1 ohms but the datasheet specifies 10 ohms. I've seen also an small inductor used there on some other similar supply circuits where the digital supply is derived from the analog, that might decrease the noise that comes back from digital side to analog. Otherwise the CS4390 noise floor is below -135 dB so it is not the problem, you shouldn't be able to hear any noise because of that. Check at least that there aren't any cables that go close to the clock circuits.

 

[Newmarket]

My bet would be on the output opamps and associated components.
Are there protection diodes on the output ? If so I might suspect them of having gone noisy (if it's not the opamps themselves)...

 

[Bernardo]

Would it be of any help to replace those TL082 with TL072?

 

[Newmarket]

Would it be of any help to replace those TL082 with TL072?

They are quieter but since it's not a high gain stage afaik then small effect. If, however, the opamps are faulty then replacing with anything reasonable would be good. TL072 would be fine. My goto would be OPA2134. Depending on package / availability. If thru hole and not socketed then I'd say fit sockets.

 

[radardoug]

You say in your first post that you insert compresssers on all the outputs. If you run it without compressers, is it noisy?

 

[ccaudle]

If thru hole

Post #18 had pictures, all SOIC packages.

If you run it without compressers, is it noisy?

Post #17 showed the noise spectrum when connected directly to the A/D converter. Pretty drastic increase in noise with increasing frequency.

 

[Newmarket]

Post #18 had pictures, all SOIC packages.

Thanks. I missed that. Forget the sockets then ! So more choice wrt package with replacement but any decent audio device will prove (or not) the point. Before fitting new parts I'd take a look directly at the DAC output pins.

 

[radardoug]

It would still be better to have him put a noise meter on the output of the Alesis. And has he got his compressers in?

 

[Bernardo]

I changed the 072 for the new 082 and the noise went down a lot, there is still a little but the unit is now usable. Thank you very much for all the advice you gave me, I consider the case closed.

 

[abbey road d enfer]

I changed the 072 for the new 082 and the noise went down a lot, there is still a little but the unit is now usable. Thank you very much for all the advice you gave me, I consider the case closed.

That is weird, since the TL082 is noisier than the TL072. It probably means the TL072 was defective, and replacing it with a good 72 would almost certainly improve the noise performance.
And as Newmarket mentioned, an OPA2134 woud almost certainly improve.

 

[Bernardo]

Surely they were bad, I will try the change that you suggested and I will publish the result, there is still a little noise in all the outputs of the device, possibly it is from 082 as you say.

 

[Bernardo]

This device also has 2 JRC 4580, are they also compatible with the opa2134?

 

[abbey road d enfer]

This device also has 2 JRC 4580, are they also compatible with the opa2134?

They are largely, but they are of a different technology, using bipolar transistor inputs, when the TL0xx and OPAx134 use FET's. They would be incompatible in circuits with very high impedance.