trafoless microphone preamp wish list
- Thread starter bcarso
- Start date
Help Support GroupDIY Audio Forum:
I think what people here might want to check would be the Gordon preamp. I have not seen on or heard one but the web site writeup is interesting
Film caps, high input inpedance etc.
I think CJ has some pictures.
Maybe PRRs design should be looked at again.
Film caps, high input inpedance etc.
I think CJ has some pictures.
Maybe PRRs design should be looked at again.
tk@halmi
Well-known member
Gus,
Are you referring to that dual stage using the doubled up BJT cascode front of Rod Elliot's Project 66 circuit?
Tamas
Are you referring to that dual stage using the doubled up BJT cascode front of Rod Elliot's Project 66 circuit?
Tamas
the gordon stuff
http://www.gordonaudio.com/
It kind of sounds like the gain settings of each gain stage are controlled and set(ratio varied) for headroom and best S/N for the gain setting maybe via a microcontroller gain control?
Maybe if you don't have a variable gain control via a pot but have set gains that you can optimize the feedback network and maybe even bias points, currents in parts of the circuit for what one wants for the overall gain wanted.
http://www.gordonaudio.com/
It kind of sounds like the gain settings of each gain stage are controlled and set(ratio varied) for headroom and best S/N for the gain setting maybe via a microcontroller gain control?
Maybe if you don't have a variable gain control via a pot but have set gains that you can optimize the feedback network and maybe even bias points, currents in parts of the circuit for what one wants for the overall gain wanted.
bcarso
Well-known member
The Gordon looks interesting. It is probably a trifle expensive, but so it goes.
Paul Stamler, have you had a chance to play with this beast?
It sounds as if Gordon does some local feedback etc. and then develops the voltages across a variable resistance. This was actually the approach I was going to use with the H*rm*n subjective eval facility computer-controlled attentuator system, although it got too expensive even for their deep pockets and I defaulted to a AD797/AD811 etc. based system, which in the end was still too expensive and was never completed
(It was finish the system or fire a secretary, as it was presented to me [and not, I might add, by the secretary, whose name was not Chloë]).
As far as the variable input R, I would only think of doing that with some fixed R's at this point across an otherwise fairly high-Z input. I have some other ideas but they are far from being sufficiently developed to discuss at this point, and frankly would have little benefit at these typical source impedances.
Paul Stamler, have you had a chance to play with this beast?
It sounds as if Gordon does some local feedback etc. and then develops the voltages across a variable resistance. This was actually the approach I was going to use with the H*rm*n subjective eval facility computer-controlled attentuator system, although it got too expensive even for their deep pockets and I defaulted to a AD797/AD811 etc. based system, which in the end was still too expensive and was never completed
(It was finish the system or fire a secretary, as it was presented to me [and not, I might add, by the secretary, whose name was not Chloë]).
As far as the variable input R, I would only think of doing that with some fixed R's at this point across an otherwise fairly high-Z input. I have some other ideas but they are far from being sufficiently developed to discuss at this point, and frankly would have little benefit at these typical source impedances.
tk@halmi
Well-known member
[quote author="Gus"] the gordon stuff http://www.gordonaudio.com/ [/quote]
No kidding! It looks like a thousand bucks worth of parts are inside that box.
Yeah, I think he is talking about varying the open loop gain on demand so the available feedback is always the same. If the remaining gain is always the same it should sound the same at all gain settings, at least in theory.
No kidding! It looks like a thousand bucks worth of parts are inside that box.
Yeah, I think he is talking about varying the open loop gain on demand so the available feedback is always the same. If the remaining gain is always the same it should sound the same at all gain settings, at least in theory.
bcarso
Well-known member
"In the Gordon preamplifier, the gain in the signal path is variable, equal to the actual setting of the gain control, with no loop feedback, no attenuators, nor the associated redundant gain." (from the website)
If we take Gordon at its word it sounds like no overall feedback.
If we take Gordon at its word it sounds like no overall feedback.
tk@halmi
Well-known member
Then you have to cascade multiple stages to get enough gain, don't you. Perhaps, that is why distortion specs are not published. Can you elaborate on this misterious idea?
Thanks,
Tamas
Thanks,
Tamas
NewYorkDave
Well-known member
I talked to the Gordon preamp's designer (nice guy) at the 2003 NYC AES. If I remember correctly, he described the circuit as a number of cascadable gain stages that are switched in or out depending on the required overall gain. It's an approach that makes sense in a lot of ways, but it does result in a very large box with many parts inside!
EDIT: I just dug around a little on Usenet and found these tidbits from the designer himself. This should clarify a bit.
EDIT: I just dug around a little on Usenet and found these tidbits from the designer himself. This should clarify a bit.
bcarso
Well-known member
Kewl Dave. That's sort of what I figured.
So really at most only two stages it sounds like, with only local feedback within each and d.c. loop feedback for d.c. stability.
The output load compensation is interesting too---I hope it makes up its mind before a session and not in the middle of one!
So really at most only two stages it sounds like, with only local feedback within each and d.c. loop feedback for d.c. stability.
The output load compensation is interesting too---I hope it makes up its mind before a session and not in the middle of one!
Svart
Well-known member
you can tell that it has tons of cascaded stages, just look at the sheer number of relays in that box!
tk@halmi
Well-known member
bcarso
Well-known member
Thanks for that before-my-time link to the earlier thread.
Still waiting to hear from Stamler about how much above 4th harmonic is considered too much, but that seems to be one of the most worrisome constraints on the design at the moment. The differential compound pair approach is elegant and straightforward, but there's plenty of higher order products, and they are not going to go away with FETs.
But there are ways and they will emerge.
Still waiting to hear from Stamler about how much above 4th harmonic is considered too much, but that seems to be one of the most worrisome constraints on the design at the moment. The differential compound pair approach is elegant and straightforward, but there's plenty of higher order products, and they are not going to go away with FETs.
But there are ways and they will emerge.
pstamler
Well-known member
[quote author="Rossi"]Another cool possibility IMHO would be to have a pretty high fixed input impedance. If impedance was about 10k, we could use smaller input caps and would thus be able to use e.g. polyprop at relatively low cost. How's that?[/quote]
If the inputs are FET-based that'll work fine (although there are a few microphones, like the Shure SM57, that really prefer a lower impedance, but they also want it to be a transformer, so never mind). If they're bipolar-based, though, the input noise current may cause problems at infrasonic frequencies.
Peace,
Paul
If the inputs are FET-based that'll work fine (although there are a few microphones, like the Shure SM57, that really prefer a lower impedance, but they also want it to be a transformer, so never mind). If they're bipolar-based, though, the input noise current may cause problems at infrasonic frequencies.
Peace,
Paul
The Calrec UA8000 switches between transformerless preamp for high levels and transformerbased preamp for low levels:
http://www.gyraf.dk/schematics/Calrec_UA8000_mic_input.gif
I think this is probably "the best of two worlds" - and if you look closely, you'll see that they do it with a single-deck (but wierd) gain switch.
Jakob E.
http://www.gyraf.dk/schematics/Calrec_UA8000_mic_input.gif
I think this is probably "the best of two worlds" - and if you look closely, you'll see that they do it with a single-deck (but wierd) gain switch.
Jakob E.
BYacey
Well-known member
The Valley circuit has an input impedance of 50K to 100K depending on which app notes you look at. pstamler, why would you say the SM57 requires a low impedance load? I have used SM57s with the Valley circuit, and they sound great. As well, the SM57 has an internal transformer between the voice coil and the XLR of the mic.
pstamler
Well-known member
[quote author="bcarso"]The Gordon looks interesting. It is probably a trifle expensive, but so it goes.
Paul Stamler, have you had a chance to play with this beast?[/quote]
I haven't. Some folks whose ears I respect have, though, and they've universally reported that the thing is head-and-shoulders above any preamp they've heard for pure clean neutrality of sound.
I'm told it has multiple low-gain amplifiers; you switch in and out however many you need for the gain you want.
[EDIT:] And evidently I was told wrong. Oh well.
Peace,
Paul
Paul Stamler, have you had a chance to play with this beast?[/quote]
I haven't. Some folks whose ears I respect have, though, and they've universally reported that the thing is head-and-shoulders above any preamp they've heard for pure clean neutrality of sound.
I'm told it has multiple low-gain amplifiers; you switch in and out however many you need for the gain you want.
[EDIT:] And evidently I was told wrong. Oh well.
Peace,
Paul
pstamler
Well-known member
[quote author="bcarso"]Thanks for that before-my-time link to the earlier thread.
Still waiting to hear from Stamler about how much above 4th harmonic is considered too much, but that seems to be one of the most worrisome constraints on the design at the moment. The differential compound pair approach is elegant and straightforward, but there's plenty of higher order products, and they are not going to go away with FETs.
But there are ways and they will emerge.[/quote]
Basically, I like to see the fourth harmonic just above the noise and everything higher below the noise floor. Rather surprisingly considering their reputation, there are several IC opamps that will do this nicely when treated right; so will some tube circuits. I haven't done a lot of messing with FETS but I gather they can do it too -- some of Erno Borbely's circuits, for example.
Peace,
Paul
Still waiting to hear from Stamler about how much above 4th harmonic is considered too much, but that seems to be one of the most worrisome constraints on the design at the moment. The differential compound pair approach is elegant and straightforward, but there's plenty of higher order products, and they are not going to go away with FETs.
But there are ways and they will emerge.[/quote]
Basically, I like to see the fourth harmonic just above the noise and everything higher below the noise floor. Rather surprisingly considering their reputation, there are several IC opamps that will do this nicely when treated right; so will some tube circuits. I haven't done a lot of messing with FETS but I gather they can do it too -- some of Erno Borbely's circuits, for example.
Peace,
Paul
