Ok, I just tried and I don't hear any difference. I must be stone deaf.
Happy now?
I do not have to be happy or sad. I can not make anyone do or think anything in life. I have no control over anyone's thoughts or actions.
Hello from Shanghai
I see you folks are having fun, sorry it will take a while to catch up. Most of Joachim's circuits adapt fairly easily to a self biasing JFET quad or complimentary LTP as input to get high impedance with little or no noise penalty. They would then work with external input termination for MM or MC. It might be fun to put together a canonical collection of all permutations (I have a few that have not been mentioned, I think).
BTW Joachim, Wurzer was my original family name (with umlaut). My grandparents Americanized it.
I see you folks are having fun, sorry it will take a while to catch up. Most of Joachim's circuits adapt fairly easily to a self biasing JFET quad or complimentary LTP as input to get high impedance with little or no noise penalty. They would then work with external input termination for MM or MC. It might be fun to put together a canonical collection of all permutations (I have a few that have not been mentioned, I think).
BTW Joachim, Wurzer was my original family name (with umlaut). My grandparents Americanized it.
Syn08, I was trying out the Hitachi low noise bipolar parts that you like so much, in 1978. Yes, 100 of each were given to me by Hitachi. You, sir, are stuck in the 70's as well. Do you know about the 'weakness' in these parts that precluded me from using them in the design that I made for Michaelson and Austin and ultimately SOTA in 1979? Let's test your real knowledge.
It is you, sir, who is 30 years behind, as I now use jfet input stages almost exclusively. That started in 1981, but I was using Toshiba jfets in designs in 1979 as well.
If someone would make something even better, I would use it today.
My designs, manufactured in the hundreds, if not 1000's, today, incorporate the Toshiba 2SK389, 2SJ109, 2SJ74, 2SK170. However, a limited production run of phono stages will incorporate 2SK146, 2SK147, 2SK240, 2SJ72, 2SJ73, and finally 2SJ75, as I have hundreds of these, on hand, and would like to use them up.
No, Syn08, I don't live in the past, it is just that I have a past to remember, when I did designs similar to what you are doing at the moment. Unfortunately, I have to continually remind you of it, if 'today's' designs are going to be put in the proper perspective.
It is you, sir, who is 30 years behind, as I now use jfet input stages almost exclusively. That started in 1981, but I was using Toshiba jfets in designs in 1979 as well.
If someone would make something even better, I would use it today.
My designs, manufactured in the hundreds, if not 1000's, today, incorporate the Toshiba 2SK389, 2SJ109, 2SJ74, 2SK170. However, a limited production run of phono stages will incorporate 2SK146, 2SK147, 2SK240, 2SJ72, 2SJ73, and finally 2SJ75, as I have hundreds of these, on hand, and would like to use them up.
No, Syn08, I don't live in the past, it is just that I have a past to remember, when I did designs similar to what you are doing at the moment. Unfortunately, I have to continually remind you of it, if 'today's' designs are going to be put in the proper perspective.
Curley, I use tantalum metal pots for variable input loading in my Vendetta designs. Where can you get tantalum resistors?
Shinkoh, Audio Note, hmmm. I can not recall others right off the top of my head right now.
Last edited:
one thing we did in the 80s was paralleling devices like Faulkner suggested (you do that too)
we used 4 BC550 and 4 BC560 in a parallel symmetric stage (not balanced)
the noise was quite quiet so to say. look at the new accuphase phono stage held in high regard by some. i could not beleave my eyes how manny they parallel. i think they use long tail pairs and those have a noise penalty too so they had to do this way. of cause the transistors that you choose are much better with current noise and i understand noise sources quite well.
i agree that many cartridges would not be happy with the MPP.
that is the reason i will start to investigate cartridge loading in my MPP thread and eventually coming up with a high impedance version too.
the Sans Parail (VFB version of MPP)should have a higher input impedance due to feedback or am i wrong?
we used 4 BC550 and 4 BC560 in a parallel symmetric stage (not balanced)
the noise was quite quiet so to say. look at the new accuphase phono stage held in high regard by some. i could not beleave my eyes how manny they parallel. i think they use long tail pairs and those have a noise penalty too so they had to do this way. of cause the transistors that you choose are much better with current noise and i understand noise sources quite well.
i agree that many cartridges would not be happy with the MPP.
that is the reason i will start to investigate cartridge loading in my MPP thread and eventually coming up with a high impedance version too.
the Sans Parail (VFB version of MPP)should have a higher input impedance due to feedback or am i wrong?
To Scott:
actually i published 2 alternative high impedance designs in my MPP thread.
see the attachment do you still speak some german ? by the way i like and use your "old" AD797 in some of my designs (albeit in a special circuit) a swiss guy just made a survey with docents of op amps and the AD797 came out best
i think Audio Precision are using it their latest gear too but you know that i asume
amasing how this thread develops : here we have them all : spam bots, subjectivists, old scool heroes from the golden age and the new wave all fighting for the best possible phono stage should be put into a musical on broadway
actually i published 2 alternative high impedance designs in my MPP thread.
see the attachment do you still speak some german ? by the way i like and use your "old" AD797 in some of my designs (albeit in a special circuit) a swiss guy just made a survey with docents of op amps and the AD797 came out best
i think Audio Precision are using it their latest gear too but you know that i asume
amasing how this thread develops : here we have them all : spam bots, subjectivists, old scool heroes from the golden age and the new wave all fighting for the best possible phono stage should be put into a musical on broadway
You raised an awful amount of good points and legitimate questions. It is refreshing to have the opportunity to exchange ideas with people that have a good understanding of the phenomena, having at the same time the guts to explore and question the state of the art.
I promise I'll be back with lots of details, for the moment I'm way to pissed at JC, his protective moderator angels and his cheerleader gang, to have the patience to elaborate. Anyway, short story is that feedback will help in increasing the input impedance but nowhere enough to be of any serious practical use. There are also a few conflicting requirements related to noise, input impedance, headroom and loop gain that all deserve a more in depth analysis.
Once again, I'll be back ASAP.
Last edited:
i think it was in the german magazine elektor and that was in the 70th
maybe it was in the AES journal and wireless world too
i think he is still a professor somewhere in england and he is in my opinion the first that paralleled transistors to lower the current noise at least he was the first that published a scientific paper about it
elector made a famous Pre-Pre the Supra that used that technology
i think Holger Barske made a recent layout for that still interesting circuit
maybe it was in the AES journal and wireless world too
i think he is still a professor somewhere in england and he is in my opinion the first that paralleled transistors to lower the current noise at least he was the first that published a scientific paper about it
elector made a famous Pre-Pre the Supra that used that technology
i think Holger Barske made a recent layout for that still interesting circuit
The subjectivist claim it has a soggy and somewhat ill defined bass not me but some selection of the h parameters will help, or feedback, or emitter degeneration or all of it
Well, good for you guys. E.A. Faulkner was the guy who first suggested paralleling transistors in 'Electronic Letters' in 1967, or earlier. I learned it from him.
syn08,
common base amplifier has low input impedance, low noise factor, wide bandwidth, wide dynamic range, high input overload capability, current gain just below 1 and no phase reversal between input and output. Nearly the same amount of current flows through the input and output and in unbalanced mode, through the signal source. The high isolation between input and output gives stability, while the small base-emitter input capacitance and the absence of Miller effect gives high speed.
The output voltage varies exponentially with the input voltage, which means poor linearity, therefore it`s best use is as a transimpedance amplifier, where the input is current, the output is voltage, or as a current buffer, where the input is current and the output is current. In both cases a low, well defined input resistance is crucial. Impedance matching is about optimal energy transfer. The common base amplifier is suitable for matching low internal source impedances to high impedance circuits. A low input impedance is also highly desirable for noise immunity. The signal impedance is not higher than: R = V / I
For the maximum transfer of power, the output impedance should be the same as the input impedance.
For the maximum transfer of current, the output impedance and the input impedance should be low.
For the maximum transfer of voltage, the output impedance should be low and the input impedance should be high.
common base amplifier has low input impedance, low noise factor, wide bandwidth, wide dynamic range, high input overload capability, current gain just below 1 and no phase reversal between input and output. Nearly the same amount of current flows through the input and output and in unbalanced mode, through the signal source. The high isolation between input and output gives stability, while the small base-emitter input capacitance and the absence of Miller effect gives high speed.
The output voltage varies exponentially with the input voltage, which means poor linearity, therefore it`s best use is as a transimpedance amplifier, where the input is current, the output is voltage, or as a current buffer, where the input is current and the output is current. In both cases a low, well defined input resistance is crucial. Impedance matching is about optimal energy transfer. The common base amplifier is suitable for matching low internal source impedances to high impedance circuits. A low input impedance is also highly desirable for noise immunity. The signal impedance is not higher than: R = V / I
For the maximum transfer of power, the output impedance should be the same as the input impedance.
For the maximum transfer of current, the output impedance and the input impedance should be low.
For the maximum transfer of voltage, the output impedance should be low and the input impedance should be high.
lumba i plan to make an mpp with 5.5 ohm impedance that has the same impedance as my titan should at least need 20mA idle on the input transistors how that sounds and performs i do not know yet
Joachim,
you don´t sleep at night like me, right?
For VT = 25 mV and IE = 10 mA you get Rin = 2.5 Ω VT = 0.025V at room temperature.
As usual, Rin is not our only concern. Without emitter degeneration many bipolar transistor parameters are loosely defined, but using emitter resistors is not advisable here. This is even worse in the common base topology, meaning that linearity (the sound) strongly depends on the bias current. DC offset is another worry.
you don´t sleep at night like me, right?
For VT = 25 mV and IE = 10 mA you get Rin = 2.5 Ω VT = 0.025V at room temperature.
As usual, Rin is not our only concern. Without emitter degeneration many bipolar transistor parameters are loosely defined, but using emitter resistors is not advisable here. This is even worse in the common base topology, meaning that linearity (the sound) strongly depends on the bias current. DC offset is another worry.
Half digested EE trivial knowledge (mostly off topic, at least for audio) + a good dose of misuderstanding of small and large signal models = Spambot
- Status
- Not open for further replies.