Yes, an inverting nfb topology doesn't have that problem (the 1+ term).
Ever seen JLH's inverting RIAA circuit? Pretty cool. The rumble filter is optional.
http://www.angelfire.com/sd/paulkemble/sound3.html
Last edited:
Brad, yes.
In an all active RIAA (note John, all active around a single gain stage) you cannot get a perfect fit in the top end with the Lipshitz non inverting network.
Note, you can also shoot for s perfect fit at HF, but you will then have to deal with pole zero location problems lower down in frequency.
I suspect you may be able to add the HF correction pole across the main feedback network, but have not explored that option.
You can download and read Lipshitz's paper - he discusses this point. You can get it for $20 from the AES store ('On RIAA Equalization' is the title IIRC).
As already mentioned, you do not need this for passive or split EQ.
In an all active RIAA (note John, all active around a single gain stage) you cannot get a perfect fit in the top end with the Lipshitz non inverting network.
Note, you can also shoot for s perfect fit at HF, but you will then have to deal with pole zero location problems lower down in frequency.
I suspect you may be able to add the HF correction pole across the main feedback network, but have not explored that option.
You can download and read Lipshitz's paper - he discusses this point. You can get it for $20 from the AES store ('On RIAA Equalization' is the title IIRC).
As already mentioned, you do not need this for passive or split EQ.
Last edited:
"This is bogus. It all depends on the phono eq approach. My circuits do not do this, for example. However, 1dB AT 20KHz is the least of the problems with hi fi reproduction."
It's not bogus and yes it does depend on the EQ approach. I thought I made that clear: I am referring to all active topologies which are especially applicable to opamp based solutions where you want decent overload capability on +- 15 V rails - say > 30 dB.
It's not bogus and yes it does depend on the EQ approach. I thought I made that clear: I am referring to all active topologies which are especially applicable to opamp based solutions where you want decent overload capability on +- 15 V rails - say > 30 dB.
John:
Don't be so defensive. You don't use those topologies and no one is accusing you of having this issue. My point was that the angst about the Wright correction for the Neumann lathe HF protection comes for free in simple cheap opamp based preamps, should anyone care. Your right, this comes after a number of other issues are resolved. The error at 15-20 KHz may be audible. Whats happening at 50 KHz is more of academic interest for now.
Yes, and John Linsley Hood does mention this drawback in one of his books, after presenting such a circuit.
I think it's a small issue - and much less of a problem than active-passive schemes using Opamps where there are significant trade-offs wrt noise and or overload capability.
I went through Lipshitz's paper very carefully - for opamp based solutions, I really think all active is best. For tube and certain discrete designs, where you may have the luxury of higher supply rails, active/passive solutions can offer good performance.
Here is my write- up http://hifisonix.com/riaa-equalizers/
I went through Lipshitz's paper very carefully - for opamp based solutions, I really think all active is best. For tube and certain discrete designs, where you may have the luxury of higher supply rails, active/passive solutions can offer good performance.
Here is my write- up http://hifisonix.com/riaa-equalizers/
Last edited:
Demian, I USED to design common op amp type circuit configurations, 40 years ago, but no more! EVEN if I were to recommend the 'cheapest' phono stage that I would want to be identified with, I would use a 2 stage phono design. They don't have the problem that Bonasi has, AND they sound better.
I just hope to keep the design standards on THIS thread up to date, and not those of 40 years ago or more.
Now, you Demian, of all people know to what extent I will take a phono design. In fact, my last effort sort of got away from me, it was so complex (it confused you too) but it is now out in the real world. My next phono design, through Audible Illusions, will be simpler but still just as refined, and up-to-date.
I just hope to keep the design standards on THIS thread up to date, and not those of 40 years ago or more.
Now, you Demian, of all people know to what extent I will take a phono design. In fact, my last effort sort of got away from me, it was so complex (it confused you too) but it is now out in the real world. My next phono design, through Audible Illusions, will be simpler but still just as refined, and up-to-date.
Op-amps 40 years ago were nothing like the kind of devices available today.
Why do you keep wheeling out this outdated stuff? It's like me saying all jets are crap because the de Haviland Comet was a sub optimal design and fell out the sky. Hell, the 5534 only came out in '77 so just what were you using on your opamp design? A 741?
There's nothing 'unsophisticated' or 'common' about an equalizer that delivers -84 dB noise ref 3 mV, <0.2 dB conformity and > 30 dB OL margins - see Self's designs for example.
Anyway, you and I will surely get into a unproductive exchange on this issue, so let's not discuss it.
Why do you keep wheeling out this outdated stuff? It's like me saying all jets are crap because the de Haviland Comet was a sub optimal design and fell out the sky. Hell, the 5534 only came out in '77 so just what were you using on your opamp design? A 741?
There's nothing 'unsophisticated' or 'common' about an equalizer that delivers -84 dB noise ref 3 mV, <0.2 dB conformity and > 30 dB OL margins - see Self's designs for example.
Anyway, you and I will surely get into a unproductive exchange on this issue, so let's not discuss it.
Last edited:
They were not so usual in the studios i worked in. Don't know why. Because they had low output level ?
In fact, I only used one, the Beyer M-160, mostly in PA or movie shootings. Not enough to get an idea of the "character" of the technology. It was nice for strings, if i remember well ?
John, this means you'll need a high gain line stage (gain around 15 ~ 20dB).
Do you prefer this combination or isn't it, today in times of digital audio, better to move some of the gain into the phono stage and use öower gain line stages?
The common practice is to use dedicated input stages that bring any signal to the same line level (1V in digital age). At this point, you will have, usually, 15/20dB of margin. it is the designer problem to design his input stage in such a way that it will have, at any frequencies, the same dynamic margin.
Not so obvious when RIAA curve has to be applied ;-)
Reason why J.C. seems to prefer a phono input stage with active EQ for the basses, followed by a passive EQ for the trebles, followed by a 15dB linear gain stage (If i understand well) ?
Right ?
Last edited:
It's not over your head, I'm just not capable of articulating what I'm trying to say in a language most can relate to. My lack of formal training in some areas can be an asset or a curse.
It's the temporal issue vs the frequency, how they can be related. The frequency response up to a perfected 200khz is not required, what is required.... is the jitter free temporal accuracy (in either digital or analog systems)of a perfected 200khz, with respect of micro transient relationships with one another --wideband, dynamically, on the fly. It actually needs to be better than that, the 200khz is a 'minimum'.
An example is going to a higher voltage rail (relatively speaking) in a discrete phono preamp, this gives a greater linearity to complex draw considerations. Let's say the circuit requires 20V to operate correctly but can function with a 40 volt rail. the 40 volt rail, in that analysis, as a projected outcome (all things being equal-which they are not) should have a better micro transient wideband (complex interrelationships, dynamically) character.
the issue is one of micro-transient phase, placement, shape and level disturbance of each and all four components of that micro-change/transient as a set, as we hear via transient function leading edge, and not much else. It is illustrated in how cdom as a individual component and in relationship and balance, becomes so critical.
Again, traditional weighting of waveforms in classical engineering analysis as applied to audio....is 90% wrong, thus there is little correlation to hearing. The other 90% of the measurement has nothing to do with what we listen for or hear with our ears.
Last edited:
Yes, phono preamps not only need decent headroom, but decent recovery from saturation due to clipping arising from micro transients such as tick/pop noise. Check out Barney Oliver's design approach in his HP phono preamp to fast saturation recovery.
Know your op-amps and keep them out of saturation. Or if transient headroom is the answer, just use valves...........
And I venture that no-one on this thread can hear 20kHz at any reasonable programme level, or remotely like. Based on demographic of age. Check it out, it's shocking. I first noticed during a calibration on an analog tape deck, and saw the VUs move but heard nothing................I believe George Martin had the same experience by accounts?
True, but EQ an instrument at 40kHz, even with very few and even no change at 20kHz and you will hear the change. Why ?
(I used-it often with female voices)
- Status
- Not open for further replies.