For the record, and I hope not deleted, Parasound schematics, especially older ones, are not secret. I do not publish the JC-1 or JC-2 schematics, but everything else is up for grabs, so to speak. Now, why should I keep 'anything' confidential? Well, need to know, is important. What do you 'need to know' about a JC-1 power amp, unless you want to: clone it, modify it, or fix it. It is still under warranty, so the need to fix it is almost out of the question. I do not recommend that you modify it, and it is ALREADY tweaked and modded. That is what makes it so much more expensive than the HCA-3500, now discontinued. They are essentially the same circuit. I did that on purpose, to show what doing the subtle things in making a power amp did to changing and usually improving the sound. The HCA-3500 was tested and reviewed by 'Stereophile' and did not even get a rating. (Not good) The JC-1 got an 'A' rating. All, or mostly from tweaking. It is the 'tweaking' that I try to teach here, but to little avail.
john curl said:
I wish i would have kept my first attempt at my current riaa for demo purposes. It is basically the same circuit (emitter input, all bjt complementary folded cascode followed by a jfet/bjt complementary differential folded cascode) as it was from he start. It has tranformed from fair to really good through many hours of trying different devices, changing bias points, different dc servo implementations etc etc. I think we usually give up to fast on a new design before we have fully explored it's potential.
John, i would really like to hear your comments on the sonical differences between using a bjt vs a mosfet as the second device in your folded cascodes.
TIA,
Anders
Hi John,
Audio is mostly a hobby these days... and mostly tube. The only SS gear I have is one of your Mc Kinnie pre-preamp and my own active x-over. But I DO enjoy this thread, so don't ask me to leave. 🙂
Zung
Audio is mostly a hobby these days... and mostly tube. The only SS gear I have is one of your Mc Kinnie pre-preamp and my own active x-over. But I DO enjoy this thread, so don't ask me to leave. 🙂
Zung
OT Naim stuff split off to its own thread.http://www.diyaudio.com/forums/showthread.php?s=&threadid=125941
I may have missed some, so my apologies in advance.
And snoopy, the word "Jap" is considered highly offensive and racist in the US (and probably other places), at least since 1946. Please do not use that term again.
Zung, I think your choice is good. Tubes rule! I just don't design with them, but the McKinnie
(Or Vendetta, as the guy who built it, was my tech, long ago) is a very good choice for a low noise front end. I know that audio just won't pay the bills for a family, and I hope that your family is well.
(Or Vendetta, as the guy who built it, was my tech, long ago) is a very good choice for a low noise front end. I know that audio just won't pay the bills for a family, and I hope that your family is well.
john curl said:
Hi John,
Let's discuss choices for volume controls, including traditional pots, stepped attenuators, and certain kinds of high-quality IC volume controls, maybe like those available from THAT. I know that many may consider IC volume controls to be a compromise, and maybe they are, but they have gotten much better and it would be interesting to see people's opinions on how close they have gotten to the best achievable traditional approaches.
I'm not pushing for IC-based volume controls, and have not used any, but I think a discussion of all the available options might fit well into this thread.
Bob
Regarding Volume Control:
Could it be an option to place the pot / stepped attenuator as per the attached schematic (U1)?
In my opinion it would
- not degrade cmmr as we do not have two pots per channel (in a balanced design)
- no interactions with the source output impedance and possible input offset currents (in a bjt input)
Regarding the interface with the power amp:
Would it be an option to place the gain resistors R16+R17 (loading the transconductance stage) at the power amp? The transconductance stage driving the cable with a high output z...
Of course, for both questions, I'm only talking about open loop designs.
Thanks, Tino
Could it be an option to place the pot / stepped attenuator as per the attached schematic (U1)?
In my opinion it would
- not degrade cmmr as we do not have two pots per channel (in a balanced design)
- no interactions with the source output impedance and possible input offset currents (in a bjt input)
Regarding the interface with the power amp:
Would it be an option to place the gain resistors R16+R17 (loading the transconductance stage) at the power amp? The transconductance stage driving the cable with a high output z...
Of course, for both questions, I'm only talking about open loop designs.
Thanks, Tino
Attachments
zinsula said:
One problem I see with that is that the preamp will always have to handle the full output of the source component. That means that unless you're listening with the volume control turned all the way up, you'll have greater distortion and reduced signal-to-noise ratio than you would have with the attenuator placed at the input.
Also, if you use a conventional pot or switched attenuator, your output impedance will be much higher unless again you're listening with the volume control turned all the way up.
It's never made any sense to me to have a gain stage ahead of an attenuator.
se
Steve, I think you didn't see the location of the pot U1?
It is not at the output.
(Apart the fact that a pot at the output in fact does attenuate also the pre's noise....which a pot at the input doesn't)
This pot influences the degeneration of the input JFets.
R16+R17 are not for variable attenuation, but fxed resistors. In such an open loop design, you should load the transconductance stage, otherwise there is way too much gain.
Tino
It is not at the output.
(Apart the fact that a pot at the output in fact does attenuate also the pre's noise....which a pot at the input doesn't)
This pot influences the degeneration of the input JFets.
R16+R17 are not for variable attenuation, but fxed resistors. In such an open loop design, you should load the transconductance stage, otherwise there is way too much gain.
Tino
zinsula:
conceptually interesting ideas ...
does one need to be concerned about how the distortion spectral balance (is that the right phrase?
) might change as the '"volume" is adjusted?
mlloyd1
conceptually interesting ideas ...
does one need to be concerned about how the distortion spectral balance (is that the right phrase?
) might change as the '"volume" is adjusted?mlloyd1
zinsula said:
zinsula said:
Ah, sorry. "U" is typically used to denote an integrated circuit so my eyes immediately went to the opamp symbol in the schematic.
So now that I see what you're talking about, I have a question. Are you able to attenuate below unity gain with that arrangement?
se
It is usually better and easier to use a pot at the input. Unfortunately, most pots DON'T like current through the wiper. When you use the pot as a variable resistor (shunt) then more current flows in the wiper. Many pots distort badly at certain settings, and it can be audible.
While the noise floor can possibly increase, for line amps, the reduction in input to the line stage can be very useful for higher voltage sources.
While the noise floor can possibly increase, for line amps, the reduction in input to the line stage can be very useful for higher voltage sources.
John (and others), what's your opinion about a buffer before the pot to allow for a fixed input impedance and more well defined source impedance for the pot? I think I have seen Hafler using the usual single-ended JFET buffer for that purpose (and maybe it was even you who pointed that out to me once?). Of course the buffer adds some distorsion, so it's a compromise.
Christer said:
Well, unless the pot's feeding an unusually low impedance, then you'll have a fixed input impedance for all intents and purposes.
As for a more well defined source impedance for the pot, there can be some merit to that if you've got a source component with an unusually high output impedance, like some of the more asthmatic tube output stages.
Ideally you'd want to be able to bypass it for sources that don't need a rescue inhaler. 😀
se
Using resistors between the sources of Jfets as an input attenuator has been the standard practice in scopes since the '60s as this input circuit from a 5A18 Tek plugin illustrates 5A18 input Its really the only way to solve the problems in that application, with a high input impedance (1 MOhm/20 pF) and response to as high as 250 MHz without significant distortion. Distortion analyzers use similar tricks to manage the dynamic range requirements, over 120 dB for a decent distortion analyzer.
I have researched pots and measured the distortion of the wiper circuit on a number of them. And compared them to resistors, switched resistor pots and relays. The best sound I have found (and really unmeasurable distortion (less than -170 dBC) is the classic silver contact switch like Daven/Shallcross uses. And the most expensive imaginable. then a remote control for this is almost impossible to achieve. The relay comes right behind IF you use frame type relays, reeds are a compromise since the conductor through the reed MUST be magnetic to work. If you can get the miniature Teledyne type relays in TO5 cans Teledyne you are at the highest level of the relay art and paying at least $25 ea last I looked. For the truly obsessive here is an article on IM distortion in passive attenuators using relays IM in Attenuators
Carbon pots, conductive plastic pots and cermet pots all have different distortions and other issues. I never liked the sound of the Penny & Giles pots but others swore by them. I preferred the larger Alps pots but those are ancient history. The Taiwan and Chinese copies of the Alps and Noble pots are nowhere near as good with at least 20 dB more distortion.
Wirewound pots, even the precision ones have significant frequency response issues limiting them to below 1 KHz for precision.
Even passive components can cause complex problems in simple applications.
I have researched pots and measured the distortion of the wiper circuit on a number of them. And compared them to resistors, switched resistor pots and relays. The best sound I have found (and really unmeasurable distortion (less than -170 dBC) is the classic silver contact switch like Daven/Shallcross uses. And the most expensive imaginable. then a remote control for this is almost impossible to achieve. The relay comes right behind IF you use frame type relays, reeds are a compromise since the conductor through the reed MUST be magnetic to work. If you can get the miniature Teledyne type relays in TO5 cans Teledyne you are at the highest level of the relay art and paying at least $25 ea last I looked. For the truly obsessive here is an article on IM distortion in passive attenuators using relays IM in Attenuators
Carbon pots, conductive plastic pots and cermet pots all have different distortions and other issues. I never liked the sound of the Penny & Giles pots but others swore by them. I preferred the larger Alps pots but those are ancient history. The Taiwan and Chinese copies of the Alps and Noble pots are nowhere near as good with at least 20 dB more distortion.
Wirewound pots, even the precision ones have significant frequency response issues limiting them to below 1 KHz for precision.
Even passive components can cause complex problems in simple applications.
Demian,
I have sonically never liked the ALPS 'black' or 'blue' that many amps use.
Do you have any measurements on those?
Sigurd
I have sonically never liked the ALPS 'black' or 'blue' that many amps use.
Do you have any measurements on those?
Sigurd
- Status
- Not open for further replies.