Hi,
sorry, but thats´s not true.
The only reason for single stage designs is due to cost, and effort.
All other technical aspects can be dealt with and improved by split stages.
The most important aspects beeing amplitude response/precision of RIAA and noise.
The highest precision is gained if every time constant is made up from its own filter-network and using highest possible feedback ratio.
Splitting the required gain between multiple active stages improves noise behaviour.
Each active stage´s topology and type (OPamp, non-OPamp, bipolar, JFET) may be chosen to provide for optimum performance.
Also the impedance values of each filter network may be tailored for optimum noise performance.
As most OPAmps for DIY are chosen as 8-pin-devices (either DIP or SOIC) a single OPAmp stage doesn´t even necessarily build more compact or cheaper.
Regarding global feedback only two ways may be used.
- either no global feedback but just local feedback in form of degeneration etc.
- or massive amounts of global feedback
´Soft feedback´ is probably the worst one can choose, neither flesh nor fish.
Similar to HighOutput MCs, that neither reach the qualities of MMs nor MCs.
Regarding clicks and hiss.
Basically it is good to put some HF-filtering as close to the input as possible.
From this point of view a first linear gain stage is certainly not the optimum.
In split-gain stages this doesn´t matter as much as the signal level is still rather small after the first gain stage, some 10-20dB below the overall gain, keeping sufficient headroom left.
It´s not that easy to get the required network to the input without compromising on other specs too much.
Think alone of differences in noise performance if using a inverting input stage, even if it were the best regarding all else signal handling.
Now a typical noninverting input stage with the complete RIAA in series feedback connection reduces the HF-gain to 1, while RIAA requires a gain of 0.
SideNote: The NJM2122 not only requires extra compensation for gains below 30dB, but the DS specs that it "must not be used as follower". This rules out its useage in single stage RIAAs.
Such a single OPamp stage required a second filter stage (LP), either passive or active, to correct for the error in amplitude response.
If not used it results in a raising response against HF.
jauu
Calvin
sorry, but thats´s not true.
The only reason for single stage designs is due to cost, and effort.
All other technical aspects can be dealt with and improved by split stages.
The most important aspects beeing amplitude response/precision of RIAA and noise.
The highest precision is gained if every time constant is made up from its own filter-network and using highest possible feedback ratio.
Splitting the required gain between multiple active stages improves noise behaviour.
Each active stage´s topology and type (OPamp, non-OPamp, bipolar, JFET) may be chosen to provide for optimum performance.
Also the impedance values of each filter network may be tailored for optimum noise performance.
As most OPAmps for DIY are chosen as 8-pin-devices (either DIP or SOIC) a single OPAmp stage doesn´t even necessarily build more compact or cheaper.
Regarding global feedback only two ways may be used.
- either no global feedback but just local feedback in form of degeneration etc.
- or massive amounts of global feedback
´Soft feedback´ is probably the worst one can choose, neither flesh nor fish.
Similar to HighOutput MCs, that neither reach the qualities of MMs nor MCs.
Regarding clicks and hiss.
Basically it is good to put some HF-filtering as close to the input as possible.
From this point of view a first linear gain stage is certainly not the optimum.
In split-gain stages this doesn´t matter as much as the signal level is still rather small after the first gain stage, some 10-20dB below the overall gain, keeping sufficient headroom left.
It´s not that easy to get the required network to the input without compromising on other specs too much.
Think alone of differences in noise performance if using a inverting input stage, even if it were the best regarding all else signal handling.
Now a typical noninverting input stage with the complete RIAA in series feedback connection reduces the HF-gain to 1, while RIAA requires a gain of 0.
SideNote: The NJM2122 not only requires extra compensation for gains below 30dB, but the DS specs that it "must not be used as follower". This rules out its useage in single stage RIAAs.
Such a single OPamp stage required a second filter stage (LP), either passive or active, to correct for the error in amplitude response.
If not used it results in a raising response against HF.
jauu
Calvin
Yes, I remember that unit, and the beautiful sales brochure which I think I still have in my files somewhere.
I had one of those for a few months, loaned to me while I was doing a preamp project. It was quite good. If memory serves, it got very positive reviews from the Gordon-Holt-era Stereophile.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.