Member
Joined 2009
Paid Member
I have a few Russian 6C4C which is an octal base version of 6B4G, which is the brother of the 6A3, which is the 6V heater version of 2A3. With a friend I have listened to an experimental mono-block, the 6C4C driven by a 6C45Pi. The performance was superb, bass was deep and solid, the high's clean and pleasant. With these tubes, the driver stage is the key (assuming you have a good OT already).
If you think this part will meet my needs for idle current, and they have the low capacitance you recommend - then I'd like to try them. How do I get some ?
The circuit I plan to use them in will be tube driven - I can apply some nfb from the output to the cathode, i.e. current feedback, as you suggest providing that I pick a tube with enough gain to 'throw away' in this fashion. I have two tubes waiting to be used, one has a gain of 33, the other of 10 so it seems all options are available although I need to finish a couple of other projects first
If you think this part will meet my needs for idle current, and they have the low capacitance you recommend - then I'd like to try them. How do I get some ?
The circuit I plan to use them in will be tube driven - I can apply some nfb from the output to the cathode, i.e. current feedback, as you suggest providing that I pick a tube with enough gain to 'throw away' in this fashion. I have two tubes waiting to be used, one has a gain of 33, the other of 10 so it seems all options are available although I need to finish a couple of other projects first
I could send few Mitsubisi parts to you, but afraid shipment costs are around 50USD - seems too much "appendage" to transistors price.
In my vision, your hybrid approach will require more interstage caps, and not forget about the need of flipping the signal phase two times, for using current NFB to cathod of the input tube. My empirical findings are, that small Ciss (few pF) input jFETs equipped with shunt PS are not worse or even better than tube front stage without shunt PS.
Member
Joined 2009
Paid Member
How about simple common cathode front end with shunt PS ? - and which directly dc coupled for FET source follower current buffer.
This is just like in my previous No NFB headphone amp. However, for using Ciuffolli kind follower at the output, one needs p-channel jFET at the input stage, for avoiding interstage cap.
With tube at the input and no interstage cap, one should use p-ch MOSFET at the output.
I strongly not recommend uising ordinary source follower at the output, nothing will remain from genuinely vivid and detailed sound.
If you insist in tube input and n-ch output follower, better to put interstage cap and to use output follower in Ciuffolli configuration.
Member
Joined 2009
Paid Member
I am taking your thread off topic - assume you are OK with this ...
yes, you are correct. So my options are:
a) triode + n-FET and no nfb
b) triode + n-FET with nfb to the grid (then fdbk is depending on source impedance - yuk)
c) tetrode + n-FET with nfb to 2nd grid (Ultra Linear)
d) triode + p-FET with or without nfb to cathode
no inter-stage caps (cap goes in input to the amp).
yes, you are correct. So my options are:
a) triode + n-FET and no nfb
b) triode + n-FET with nfb to the grid (then fdbk is depending on source impedance - yuk)
c) tetrode + n-FET with nfb to 2nd grid (Ultra Linear)
d) triode + p-FET with or without nfb to cathode
no inter-stage caps (cap goes in input to the amp).
I would use option d), tube input stage with shunt PS, and p-FET (possibly IRF9610, I do not see better options) output follower loaded by current source, speaker is in parallel to the p-FET via good caps + shunts. No NFB.
But, I believe, this thread's schematics is better even with IRF510 at the output (instead of KP907A).
Member
Joined 2009
Paid Member
maybe, maybe not - but you built it already so for me it's no longer a new avenue to explore
maybe, maybe not - but you built it already so for me it's no longer a new avenue to explore
Without careful listening and comparisons of various amps one can not select among design solutions, and sound perfectioning is not possible.
Best of all would be, if you build both variants, and report the comparison.
In my situation, I have several cases and power supplies, and when some design becomes definitely a looser among others, it's case go to a new design etc. I have to say, that Zen4 and Zen9, that I loved a lot few years ago, do not exist anymore in my "gear-park". As Mr. Einstein told, everything is relative in this world (a joke).
Hello, guys
Reading about possibly the best SS amp nowdays (the Soulution)
6moons industry features: A visit to Soulution
I have found the following statement:
"Soulution's resident circuit engineer Christoph Schürmann is a very soft-spoken and modest chap. His design ethos revolves around speed (which requires high bandwidth which challenges stability) and concomitant precision. With his unique fixed-gain voltage control amp performing up to 80MHz [left - the upright circuit boards connect via silver pins and the module is cast in synthetic resin] and 3-stage current buffer, the amplifiers' feedback loop has been shortened by a factor of 10 to optimize operational speed and minimize errors. His current sources operate with bandwidth from 10 to 100MHz and the impedances for his power supplies are held very low. "The signal passes through my voltage gain stage in approximately 10 nanoseconds with a max amplitude deviation of 0.1dB."
Only 10nS are spent by a signal for passing through the voltage amplification stage.
From the pictures at the beginning of this thread, I have got near 69nS (output sine signal shifted from the input sine). Do you think, is the speed factor really have such a top priority? Is it a key factor to Soulution's success?
Reading about possibly the best SS amp nowdays (the Soulution)
6moons industry features: A visit to Soulution
I have found the following statement:
"Soulution's resident circuit engineer Christoph Schürmann is a very soft-spoken and modest chap. His design ethos revolves around speed (which requires high bandwidth which challenges stability) and concomitant precision. With his unique fixed-gain voltage control amp performing up to 80MHz [left - the upright circuit boards connect via silver pins and the module is cast in synthetic resin] and 3-stage current buffer, the amplifiers' feedback loop has been shortened by a factor of 10 to optimize operational speed and minimize errors. His current sources operate with bandwidth from 10 to 100MHz and the impedances for his power supplies are held very low. "The signal passes through my voltage gain stage in approximately 10 nanoseconds with a max amplitude deviation of 0.1dB."
Only 10nS are spent by a signal for passing through the voltage amplification stage.
From the pictures at the beginning of this thread, I have got near 69nS (output sine signal shifted from the input sine). Do you think, is the speed factor really have such a top priority? Is it a key factor to Soulution's success?
I hope you dont take all the mambo jumbo written in so called audiophile magazines seriously. Read between the lines and see the circuit is in reality nothing special, for example :3 stage current buffer sounds exotic but is nothing more than a triple outputstage we all know from ages ago coupled to probably hawksford error correction to get those THD figures. Speed requires high bandwith which challenges stability ?? mmm a bit dodgy.
I wonder if the 10ns measurement is as accurate as he has measured the DF of that amp.
Yes, output impedans figure near 80 mOhms is nothing special at all. But what I find in consonance with my vision, is the big role of power supplies solutions. How to make a PS fast enough, maybe shunt kind PS are used?
They claim that 10nS duration transients at PS output are very important for sound.
They claim that 10nS duration transients at PS output are very important for sound.
Power supplies are I agree, especially with class AB amps.
Shunts, could be .... Use of SIC or at least high speed fast recovery or shotkeys diodes as rectifiers I hope as well, to justify the price of that amp.
Apparently a patent for this amp was turned down, probably the design is already public knowledge.
Your preamp regulators look more than adequate for your circuit.
Shunts, could be .... Use of SIC or at least high speed fast recovery or shotkeys diodes as rectifiers I hope as well, to justify the price of that amp.
Apparently a patent for this amp was turned down, probably the design is already public knowledge.
Your preamp regulators look more than adequate for your circuit.
Hi, Eric
It is around 25W per channel, maximum output power could vary from 1 to 3W depending on the load resistance (maximum output power from this schematics is at 20-30 Ohms load resistance)
Yes, with sensitive speakers this could be enought. But, nothing preclude you to make this schematics more powerful, by replacing the output transistors and by rising up their idle current.
Is there some new update to the circuit on post #1? Is this design still ok to build?
I plan to build a headphone / pre base on the design so i think i can lower voltage and output current to 100ma, it it fine? Do i need to decrease feedback current, too?
Why the psu of input stage go from - 38v then to zenner like circuit to reduce to - 31v? What is the advantages of this method?
I plan to build a headphone / pre base on the design so i think i can lower voltage and output current to 100ma, it it fine? Do i need to decrease feedback current, too?
Why the psu of input stage go from - 38v then to zenner like circuit to reduce to - 31v? What is the advantages of this method?
Last edited:
This circuit needs adjustments of R2, R3, R4 depending on rail voltages and on examples of VT1 and VT2. So, one must understand and "feel" how j-fets operate. The most important feature of this amp is related with your last question - why do we need those complications with excessive PS components.
I found this from listening experience - sound becomes crystal clear, if one does not allow signals to walk along PS chains - all voltages and consumed currents variations under signal must be stopped as close to active parts as possible. Therefore I use solutions based on VT3-VT4, and VT7. At this, simple capacitors shunting does not work well, active parts based solutions are better.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.