I thought feedback was never a good thing. It is used to lower distortions, but brings with all sorts of bad "chit"
Why use it if it could be done without it, or at least a high degree of it? Has something changed with regards to negative feedback? Since when have IC's been considered a good way to make great sound?
Excuse me! I thought that I was asking a question that pertained to the circuit.
Exactly this is, what I often note by my modifications. But I would say, global negative feedback with more than one voltage gain stage in the feedback loop. Mostly there are two voltage gain stages in the NFB loop, both operational amplifier ICs and discrete pre- and power amp commercial and diy versions. If I use NFB by only one voltage gain stage, there are no problems.
In other case it is to observe a sine wave deformation to a saw tooth (in direction to higher frequencies), both by Pi-spice simulation and measuring in real life. But I want to see clearly low pass character (i. e. reducing of the level in direction to higher frequencies) without visual deformation of the sine wave, even arround 1 MHz.
Thus, the main work by amp modify is to reduce the open loop gain of the first gain stage to "unity gain" though enhancing the emitter resistors of differential stage (LTP).
If you look at my simulation results of the Pass X Series, you understand, what I mean
http://www.diyaudio.com/forums/pass...m-results-pass-x-series-us-pat-5376899-a.html
also this thread is of interesting regarded the (dis) advantages of global NFB:
http://www.diyaudio.com/forums/solid-state/114865-densen-amp.html
For me the measurement method, that gives advices about the expected sonic transmission is clearly those, that shows, whether there are low pass character or a deformation of the sine wave shape is visible. The last is an indicator for harsh and metallic sound
Last edited:
i just came back from vienna where we listened for two phono stages-
one was an open loop design with fet-bipolar cascode and the other an op amp design with feedback. people (all professional audio designers) preferred the op amp version although they did not know what was playing. The comments i got was that the op amp version had a more extended bandwidth in the bass and treble and that it had more dynamics and more information content without being less "musical" then the open loop one. So it is possible to design a good sounding circuit with high feedback in my experience although i use an unusual approach in that design we listened too. Anyway it measures very well and has high feedback and sounds good people tell me.
one was an open loop design with fet-bipolar cascode and the other an op amp design with feedback. people (all professional audio designers) preferred the op amp version although they did not know what was playing. The comments i got was that the op amp version had a more extended bandwidth in the bass and treble and that it had more dynamics and more information content without being less "musical" then the open loop one. So it is possible to design a good sounding circuit with high feedback in my experience although i use an unusual approach in that design we listened too. Anyway it measures very well and has high feedback and sounds good people tell me.
Interesting. I have always been told that IC's and high negative feedback cause many issues. I guess there is more than one way to skin a cat. I would like to personally listen to a similar circuit for an extended period of time though. Sometimes initial reactions of something being better, often fail over time as what was deemed to be better on a quick A/B, in time is a distraction and causes listening fatigue. Just my observation from past experiences.
Last edited:
Short feedback loops of one stage only are called 'LOCAL feedback' and can solve many problems in audio design. I always use some 'local feedback' to linearize a design. It is the 'open loop bandwidth' that is the difference. Local feedback usually has a very high open loop bandwidth, so FM does not become an issue from the feedback, itself.
Curley, not all discrete designs sound extra good. If they had a bandwidth, distortion or another problem, then the op amp might well sound better. In that one case.
Curley, not all discrete designs sound extra good. If they had a bandwidth, distortion or another problem, then the op amp might well sound better. In that one case.
in fact it is a transimpedance instrumentation amplifier made of "seperate" op amps.
so i use shunt and parallel feedback. i come from the high tech industry so i look at things a little different. i have seen circuits like that in brain wave scanners but not in hifi.
dc precission is so good that i use no capacitor and no servo. it is just precission op amps and precission (0.1%) resistors it´s a head amp without riaa so that was a head amp shootout in vienna
so i use shunt and parallel feedback. i come from the high tech industry so i look at things a little different. i have seen circuits like that in brain wave scanners but not in hifi.
dc precission is so good that i use no capacitor and no servo. it is just precission op amps and precission (0.1%) resistors it´s a head amp without riaa so that was a head amp shootout in vienna
I am addressing a specific concern. What is your input impedance, the effective load resistance that will load your MC cartridge? IF you use one kind of feedback, you will reduce the input impedance below the cartridge resistance. Is this what you did? Or did you supply conventional feedback, like Syn08 tends to use. This might be an important factor in ANY phono stage comparison.
i make two versions, one has low z and the other high z. in vienna we tested the low z version with lyra and MFSL carridge. that version has 2x10 Ohm to amp and 2x220 Ohm to ground. no compression or bandwidth limiting was observed in that setup. maybe different with Benz or Clearaudio that have much higher internal impedance and may not like that load as much. i really have to test my Clearaudio Signature with that amp but agjusting a cartridge mechanically for optimum performance is a pain in the butt, even for me after all those years setting up 100´s of sytems at shows, dealers and customers.
the basic problem by such listening sessions is follow:
I don't hear the differences only between a circuit with NFB and without NFB.
There are always a wide range of other influences (e. g. the character of other components such as speakers, power amplifiers, the individual prefering in sonic character and so on).
The most important question for me in this case is follow:
- What about the power supplies of this both different phono riaa preamp's (please note, residual hum components on the voltage lines are more critical by topologies without global NFB)?
- What quality standart and what position in the circuit are there by the RIAA capacitors/components?
- How much sensitivity there is by the output stages of both phono RIAA preamps to different cable load?
One could ask more questions when I do something thinking.
Unfortunately it is so, that we can take only a fairly safe proposition when both topologies in such a perfect condition is, and irritation/confusion effects through design and construction errors can be ruled out fairly safe. But such commercial RIAA preamplifier devices are very hard to find, independend of the price class. Please note, most phono preamps haven't perfection like this:
Lyra Connoisseur 4-2 P SE世界公èª�é ‚å°–å”±é 放大器之一-myaudio 家家音響
But even by this device I would make some things in other way.
Last edited:
as i said maybe not here but somewhere else i own and have owned several versions of the Connoisseur and have listened to it in docent of setup´s so i know where i am standing. the pre-pre´s in vienna where supplied by batteries. the fet with a lead battery and the op amp type with alkaline. the fet version already got great praise for soundquality and i know about volume diffences. anyway, good observations, they are true. the single BEST phonostage for all ocations simply does not exist or i have not heard it sofar.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.