Did anybody ever built a power amplifier (30w min) with the RIAA network embedded in the feedback network able to do the whole gain from mm or mc cart to speakers ?
Why would anyone with any sense do something like that? Noise, PSRR, stability (power amps are hard enough to get stable with all kinds of loads as it is), distortion issues etc.
For fun?!... I think that actually the RIAA network applied in the feedback path can help a lot to the stability problem.It can also be split, nested...
Interesting question, never seen that amplifier. Maybe because power amplifiers have much lower gain than phono preamplifiers.
Maybe it is possible, many commercial japanese amplifiers use RC network in feedback of power amplifiers for tone controls, maybe it could be modified to phono RIAA network.
Maybe it is possible, many commercial japanese amplifiers use RC network in feedback of power amplifiers for tone controls, maybe it could be modified to phono RIAA network.
Of course...we loose the possibility to use other sources than phono cartridges, but why wouldn't we have a perfect combination in one? Most audiophiles have more than one amp in their homes anyway...
The main reason for doing it was to run the phono preamps at the highest possible rail voltage to get the lowest distortion figures and widest dynamics , scrap the eq section as no op-amp was going to work over 22.5v anyway ,while discrete eq were expensive and unnecessary this way!
I think that the only reason they didn't built the riaa network in the feedback path was to leave the user the option for using other sources as the usual user wouldn't buy a one use only amp.Otherwise it seems to me like the best way for the best phono setup.
Last edited:
No I very much doubt it - the RIAA network is inherently fixed gain. You need a volume control.
Trying to stabilize something like 140dB of voltage gain via a strongly frequency-dependent network is unlikely to be tractable without a lot of work.
You can get world class performance for a phono amp from as low +/-5V rails, +/-15V is plenty.
The dynamic range is not an issue if you provide jumper selection of gain to match to different cartridges, vinyl has modest dynamic range!
Its unworkable to have >120dB of loop gain that would require vast amounts of open loop gain which would be beyond a nightmare to compensate to stability and require careful shielding too.
And a volume control is "quite useful"!
The traditional approach of a separate phono preamp / RIAA network with its own filtered supply feeding a generic preamp and volume control into a generic power amp is far superior in many respects, its flexible, its works, it can be made to work extremely well without compromize as each component has one task to do.
As said by Mark in #7 and #10, amplification must be from 100µV to say 16V (30W/8Ω) 160 000x (104 dB), but for the low frequencies an extra 10x (20dB) is needed: 1.6 million times / 124dB. Without feedback included.
Let's take 26 dB (20x) feedback.
Open loop ampification must be 32 000 000x / 150dB.
We're not discussing hf-stability, temperature stability, power supply issues, components (choise, selection), topology, circuit board, and the rest (what not).
And volume control (active in the feedback?)
Even if you're downscaling to a headphone amplifier, the reduction is only say 100x (40dB).
Do you know any commercial enterprise that have ever wrought such thing?
Let's take 26 dB (20x) feedback.
Open loop ampification must be 32 000 000x / 150dB.
We're not discussing hf-stability, temperature stability, power supply issues, components (choise, selection), topology, circuit board, and the rest (what not).
And volume control (active in the feedback?)
Even if you're downscaling to a headphone amplifier, the reduction is only say 100x (40dB).
Do you know any commercial enterprise that have ever wrought such thing?
Last edited:
I think I got my figures wrong, 100dB loop gain, not 120dB, but you get the point its a vast amount of gain for a single amp.
Good low-noise rail-to-rail output opamp, something like the OPA1641/2, for MM, add a ZTX851 MC stage if needed. Stick in Doug Self's RIAA network of choice, voila! Just be careful to have enough preset gains due to headroom constraints.
Soon the low voltage CMOS opamps will have low enough 1/f knee to do an even better job...
So not exactly a world class...I know only one phono preamp that perform really well at +-15v, but it's current feedback.Once you have dust on your vinyl nothing lower than this work good enough, and none of the voltage feedback sort with less than +-25v on its rails.You can tell me otherwise, but i tried so many phono preamps, both diy and commercial that you can't simply change my mind in any way. I didn't try Scott's jfet only design to benefit its soft compression behavior as it's too expensive for me...
By the way, I have said that before, unless i bought a faulty batch of opa1642, its noise performance on high impedance source stinks...
Last edited:
An assumption of prooven realisation could have answered the question before posting it. Nevertheless a nice jump of thought and enlightning calculus.
The said amount of amplification is needed whatsoever, but practically divided over several stages for control, stability and volume adjustment.
Could you please expand on this. Why would a CFB amp be more tolerant of lower supply voltage? I assume this was an active RIAA, right?
Pick a better opamp if you want (the 1642 isn't the greatest for voltage noise, true, but its JFET so no current noise to speak of) - but you perhaps should define what you think world-class is for a phono preamp, because it sounds like 0.0001% distortion disappoints you, despite the fact people cannot detect distortion below around the 0.03 to 0.1% level.
My point is the supply voltage isn't a notable factor in performance with opamp circuits so long as its high enough for the headroom - no point clipping the power amplifier, 3V rms should be enough for anything, and if you need to step up to +/-9V using batteries that eliminates PSU noise too, which is my preferred approach.
This batch of opamps with the noise issue - what did they measure? (You did decouple according to the datasheet?)
Last edited:
I didn't say that a current feedback is necessarily more tolerant of lower supply voltage...I just said i know one preamp of this sort that sounds well.