Does the C84 require an op-amp?

[allbaldo]

The descriptions aren't posted yet, and I don't really read schematics, so I can't tell if the C84 needs an op-amp. Does anyone know? I think I may start with this pre, since it's a fair bit cheaper than the others, but if I need to but an op-amp to go with it, I may rethink that. Any help would be great! Thanks!

[tele_player]

It uses two Burr-Brown OPA2604AP dual op amps, and they're included in the full kit. This saves a LOT of money (and board space) over using discrete op amps. The Burr-Browns are under $3 in quantities of 100 or more, and remember, that's TWO op amps. Compare it to $50-$75 for a single discrete op amp like the Hardy 990 or the Avedis 1122.

From the parts list, I'd say the full kit is truly a full kit. Going transformerless really keeps the cost down.

[EptihetMusic]

Could you excange the op amps for other discrete op amps like the Hardy 990?

[dabo]

An adventurous solder jockey would probably say it's "possible" to replace the IC Op-Amps on the C84 with discrete devices, but if you download the assembly instructions and check out the photos you'll see that it's not a simple thing.

[tpryan]

First, you'd need four of them. Not only is there no space on the board, they'll cost a bunch of money.

Second, why would you want to? The problem with most IC op-amps in high performance audio applications is that they 1) aren't quiet enough, 2) don't have enough headroom, and 3) can't drive low impedance loads to high levels. The C84 solves 1) wiith a pair of LM394 super-matched transistors on the front end, 2) by specifying OPA2604s with their +/-24V maximum supply voltage, and 3) by adding a pair of discrete Z buffers on the output.

[Erik Sikkema]

Is the 2604 really the best opamp to think of. Noise 11nV/root hertz, there are better ones, even a 5532 is less noisy, not that I prefer that one, but how about the AD797 ?

Erik Sikkema

[tele_player]

There's more to it than that. Tim's designed things using the 797, I'm sure he's got good reasons for his choice in this application.

[Erik Sikkema]

Probably the fact the 2604 can run on a higher voltage and has a higher slewrate, but the noise of the 797 is soo low, about the lowest you can get in opamp land. I would like to hear this from Tim, because he is not the only one using the 2604, as far as I have seen myself the Prism MM4 uses the SSM 2022 paired transistor opamp and the 2604. But the 2604 has a much higher distortion than the 797.

Erik

[tpryan]

Is the 2604 really the best opamp to think of. Noise 11nV/root hertz, there are better ones, even a 5532 is less noisy, not that I prefer that one, but how about the AD797 ?

If you care to read the AES paper describing the circuit, you'll see that in the input stage, op-amp noise is dominated by input transistor and feedback resistor noise, especially at higher gains. What you need from the op-amps are enough voltage swing to accommodate the maximum output swing and enough output current to keep the LM394s working at the optimal point on the noise curve. Also, you want closely matched input offset voltages, since the whole thing is direct coupled, all of which strongly suggests a part like the 2604.

There are certainly others to try, however, and there's no reason not to experiment to see if there's something that works better for you. That's one big reason to build it yourself, right?