I'm ashamed you don't have an AD8099 input + DSL composite design RIAA pre. 😉
Okay on second thought--the AD8099 probably not the right low-noise part for a RIAA. (And absurd regardless)
Last edited:
For phono, ditto.
For balanced input, finding 2 dual JFETs with 4 identical Idss values is bloody difficult. Much easier with parallel Idss, but a B doing by hand 40 years ago. Piece of having your cake with a few software lines.
(autocensorship deleted 'absolutely' and 'ugly', circumventing protocols was lots easier in the '70/'80s)
Last edited:
Yeah that is disgraceful, even with RR output op amps. The other thing people do is just pare back the overall gain, which I think is also cheating. I think at least John Atkinson is on to this now, and no longer oohs and ahhs over the overload margins when he notes that the gains are low.
Yes, I'm sure a lovely part but a lot of current noise for a half-henry cartridge.
> Lost track of my AD624/744 passive RIAA pre that Walt and I published in 1983 or so.
Google comes to the rescue.
http://www.analog.com/media/en/technical-documentation/application-notes/5866763300941AN245.pdf
Patrick
Google comes to the rescue.
http://www.analog.com/media/en/technical-documentation/application-notes/5866763300941AN245.pdf
Patrick
It is a lovely part though. The gain-bandwidth is absurdly good, and I notice the packaging had to be redesigned to allow us to take advantage of it. This does raise the question of stability when a gain stage is being fed from a reactive source, but that's why there are ferrite beads 🙂. This might be one time when one would find very good reasons for not using the conventional single-network RIAA compensation, besides potentially low loop gain. Or, locate the part right at the cartridge (that is for an MC cartridge with low inductance).
That's a valid point the 5534 is kind of a sweet spot on the current noise and no Ib comp. Take the ADA4841 on a modern SOI process beats the 5534 speed specs by a mile and has 2.1nV noise on only 1.1mA, but the current noise is still 1.4pA. Nice though for those battery powered mic circuits.
Last edited:
If you only knew the friction caused by re-assessing standard but bad pin-outs like electrometers with the input lead directly adjacent to a supply pin.
The latter points by you two do go nicely back to the original quote about "it's not all about audio". 😀
The AD8099 (as is the ADA4841) is an incredible opamp for sure--for the right application. (Another reason to shrug when folks discuss opamp rolling)
The AD8099 (as is the ADA4841) is an incredible opamp for sure--for the right application. (Another reason to shrug when folks discuss opamp rolling)
it is very strange, for me, to see some people, negating OPA's sound differences, discussing RIAA preamplifiers in 2015.
In an OEM three-piece system for Dell years ago (so I can talk about it, but it is a bit depressing that so many things I write include the phrase "years ago") the TSSOP uC was in the right satellite and doing a lot of work, and the satellite managed, over four wires, to get phantom power, power the speaker, power stereo headphones when they were plugged in, and transmit data bidirectionally over the phantom power conductor. Oh, and there was a shaft encoder control and selection of volume, bass, and treble adjustments and an LED display of the settings. I confess to being rather proud of that, and after I was kicked out Harman actually saw the patents past the office action and through the process to issue. My co-conspirator Mark Grzybek hacked out the firmware and did a great job. The development took all of four months. And although there was a mod or two before kicking off production, the system worked from the beginning. My boss at the time had an epiphany.
But the gate oscillator of that uC mentioned had its high Z input next to a power supply pin. The leakage when the washless flux process wasn't quite right caused the gate oscillator to be biased away from the midpoint and reduced the gain to below the threshold of oscillation. It could take about five minutes after power up for the thing to start. Some guarding would have helped, but between cleaning up the process and lowering the value of the input-output bias resistor, it became reliable.
Low noise is much less controversial than subtleties of "sound". And I'm not even going to mention stochastic resonance.
Did-you mean low noise from a vinyl, compared to 24/96 digital ? Or distortion ? Or bandwidth linearity ?
- Status
- Not open for further replies.