I have been asked to 'include' a tube in a setup. Why? Hmmm, probably purely conjective, but there you have it!
Why Pentode based? After having looked around at a lot of tube designs, starting off with Triode based designs (one of which I built based on an Aikido CCDS but gave way too much gain for what I was after) I 'think' that a Pentode based design is perhaps better for a buffer?
Hell, I could be wrong: Tube design is not my forte by a looong chalk and TBH I have been thrashing around trying to settle on a design for this.
Hence any input on this would be GREATLY appreciated before I sell my soul to the devil and buy a cr@ppy Chinese kit from fleabay.
Andy
Why Pentode based? After having looked around at a lot of tube designs, starting off with Triode based designs (one of which I built based on an Aikido CCDS but gave way too much gain for what I was after) I 'think' that a Pentode based design is perhaps better for a buffer?
Hell, I could be wrong: Tube design is not my forte by a looong chalk and TBH I have been thrashing around trying to settle on a design for this.
Hence any input on this would be GREATLY appreciated before I sell my soul to the devil and buy a cr@ppy Chinese kit from fleabay.
Andy
Problem is:
Buffer is a cathode follower (gain~1)
Cathode follower usually means load resistance at the cathode.
Hence the cathode is floating with the output voltage.
unfortunately the pentode needs a G2 supply between cathode and G2. Wouldn't be a problem for a grounded cathode amplifier.
In your case, you need a separate, galvanic isolated G2 supply which is floating on the cathode voltage.
You have the same problem with Pentode based tube power supplies or pentode based OTL-Futterman amplifiers.
A lot of effort for a simple buffer if you ask me....
You could the standard configuration for the G2 as well and use a very large bypass cap if the screen current is low, but i don't know abou the performance.
BTW: Depletion MOSFET curves are the same as Pentode curves and I bet everything, that you wont hear a difference between source follower and a pentode cathode follower. (As long as you still habe impedance bridging)
Buffer is a cathode follower (gain~1)
Cathode follower usually means load resistance at the cathode.
Hence the cathode is floating with the output voltage.
unfortunately the pentode needs a G2 supply between cathode and G2. Wouldn't be a problem for a grounded cathode amplifier.
In your case, you need a separate, galvanic isolated G2 supply which is floating on the cathode voltage.
You have the same problem with Pentode based tube power supplies or pentode based OTL-Futterman amplifiers.
A lot of effort for a simple buffer if you ask me....
You could the standard configuration for the G2 as well and use a very large bypass cap if the screen current is low, but i don't know abou the performance.
BTW: Depletion MOSFET curves are the same as Pentode curves and I bet everything, that you wont hear a difference between source follower and a pentode cathode follower. (As long as you still habe impedance bridging)
Last edited:
underwurlde said:
I'm still not sure why. Did you just decide to try it, or has someone else requested it?underwurlde said:
Conjective? Not a word I am familiar with. Wordnik says it means "In projective geometry, said of two projective primal figures of the same kind of elements when both are on the same bearer." Possibly not what you meant to say?
Do you actually need a buffer to cope with long interconnects or non-linear port impedances, or are you really after a mild FX box to give 'tube sound'? Pentodes have very high output impedance when used in grounded cathode, so no good for a buffer. When used as a cathode follower (OK for a buffer) they are either used in triode connection or involve extra complication to ensure g2 tracks the cathode signal; might be simpler to use a triode.
Hell, I thought conjective meant something else entirely. I need to swallow a thesaurus. My bad...
The guy wants a tube in-line with his music. Why? Heavens knows, but he wants to see it there hence so be it, I shall try my best to give my friend what he wants. I'm sure a MOSFET will do the same thing, but unfortunately not the purpose of this strange exercise.
I think therefore that a Pentode connected as a Triode is what we're after. From what I see, this makes it easier than just simply using a Triode to get the lower gain required for a buffer style circuit? Could be miles out of course, very steep learning curve for me this, but VERY interesting!
Thanks again for your input / interest
Andy
The guy wants a tube in-line with his music. Why? Heavens knows, but he wants to see it there hence so be it, I shall try my best to give my friend what he wants. I'm sure a MOSFET will do the same thing, but unfortunately not the purpose of this strange exercise.
Well put! After speaking with him, he was glad this was mentioned, so yes he does indeed require the tube sound of old.
I think therefore that a Pentode connected as a Triode is what we're after. From what I see, this makes it easier than just simply using a Triode to get the lower gain required for a buffer style circuit? Could be miles out of course, very steep learning curve for me this, but VERY interesting!
Thanks again for your input / interest
Andy
A vacuum tube isn't magic. Properly implemented, it will have no discernible sound. I think that the psycho-acoustic effect would probably be fulfilled if you just wired up a black box with a tube poking out of it (with the heater connected for the glow) and the line-in connected to the line out.
A triode is simpler than a pentode connected as a triode. Either can give you the just-below-unity gain of a genuine buffer.
If your friend wants genuine 'tube sound' then he may need a valve amplifier. If he just wants some low order distortion, and a little hum and noise, then a cheap chinese 'tube buffer' may satisfy him. Check that the valve is actually used in the circuit, as some have been known to cheat.
Used properly valves don't produce 'tube sound', except in guitar amplifiers but they are a completely different way of using valves.
If your friend wants genuine 'tube sound' then he may need a valve amplifier. If he just wants some low order distortion, and a little hum and noise, then a cheap chinese 'tube buffer' may satisfy him. Check that the valve is actually used in the circuit, as some have been known to cheat.
Used properly valves don't produce 'tube sound', except in guitar amplifiers but they are a completely different way of using valves.
Understood. He does not want a valve amp though - I think he has something else in mind too (he is a music producer)... this is going somewhere I'm sure. I will try and pin him down tonight if he's around.
In the meantime can you point me to a good design for each implementation you mention please? I have searched and searched many designs both here and on the internet and am getting overwhelmed with all the options - but nothing which does what I want it to (as far as I can tell).
For now I shall avoid those ebay kits and will knock something up myself. I have knocked up a power supply already which I can tailor to suit and more valves to hand than I care to mention! I have the technology, just lacking the knowledge (of valve / tube design).
Thanks again,
Andy
In the meantime can you point me to a good design for each implementation you mention please? I have searched and searched many designs both here and on the internet and am getting overwhelmed with all the options - but nothing which does what I want it to (as far as I can tell).
For now I shall avoid those ebay kits and will knock something up myself. I have knocked up a power supply already which I can tailor to suit and more valves to hand than I care to mention! I have the technology, just lacking the knowledge (of valve / tube design).
Thanks again,
Andy
You probably don't want a good design, as a good design won't change the sound hardly at all. You need to look for a poor design, or take a good design and damage it (e.g. wrong bias, too low supply voltage).
You may have gathered by now that I am not a fan of tube buffers, so I am not the right person to advise you. However, perhaps I have helped you clarify what you are looking for.
You may have gathered by now that I am not a fan of tube buffers, so I am not the right person to advise you. However, perhaps I have helped you clarify what you are looking for.
You have indeed my friend.
OK, I will start with a 'good' design then (and I will then stick with it! - pssst, don't tell him)... I know that if he see's the valve in all it's glowing glory, then he will automatically think and assume 'valve sound' (<- I am also getting the feeling that is NOT a good phrase to mention around here!).
So, any GOOD designs to start me off with then peeps?
Andy
p.s. I am truly grateful people are taking the time & effort to respond, as it must be quite painful for some!
OK, I will start with a 'good' design then (and I will then stick with it! - pssst, don't tell him)... I know that if he see's the valve in all it's glowing glory, then he will automatically think and assume 'valve sound' (<- I am also getting the feeling that is NOT a good phrase to mention around here!).
So, any GOOD designs to start me off with then peeps?
Andy
p.s. I am truly grateful people are taking the time & effort to respond, as it must be quite painful for some!
I have to agree with this... The chinese way of doing a "tube CD player" or a "tube DAC" is usually shoving a badly misdesigned cathode follower after a bunch of op-amps. A bunch of bovine maneur if you ask me, but if that's what the market wants, who am I to tell them not to sell those. But tube equipment they are not.
But if that warmed up and fuzzy "tube sound" - slow'n'mellow - is what your friend wants, then a badly misdesigned CF could do the trick. Maybe an ECC82 with a slightly cold biasing and a too small cathode load might do the trick?
Pentodes as cathode followers in pentode mode is something I intend to investigate in the future. The future being anything from this moment to 60 years from now. Theoretically the higher transconductance a pentode gives might be a nice thing, since the output impedance of a CF is roughly 1/gm. Then again, the need for a g2 supply swinging along with the cathode does make the design more complicated compared to a triode design, as has been mentioned here.
I love cathode followers and I believe that pentode, properly configured would make an excellent very low-distortion stage. You need a pentode with nice flat curves, good gm. The gm ensures low output resistance, the very high gain reduces distortion. I have built quite a few. The one I posted here nine years ago was included in a power amplifier; I used a very high-gm device - D3A with an active g2 feature. The circuit worked very well. Here's a link
@Fenris, as mentioned above in my post (#3) I have built a CCDA, here is a pic:
Based on:
I shall investigate adding some feedback to this then to see what happens! Would I just connect a resistor from Lout to Pin 2 of V1?
@7N7, thank you for your encouragement. After looking at your design, I think I will struggle to implement the preamp stage into something that I could use... Too clever for me!
With the CCDA it was very close to what because I can use just one single valve per channel and uses a single, (and relatively) low supply voltage. BUT the damn thing INVERTS the signal ("obviously", I hear many of you shout...). Would this matter??? <- apologies for the stupid question
Fun, fun, fun!
Andy
Based on:
I shall investigate adding some feedback to this then to see what happens! Would I just connect a resistor from Lout to Pin 2 of V1?
@7N7, thank you for your encouragement. After looking at your design, I think I will struggle to implement the preamp stage into something that I could use... Too clever for me!
With the CCDA it was very close to what because I can use just one single valve per channel and uses a single, (and relatively) low supply voltage. BUT the damn thing INVERTS the signal ("obviously", I hear many of you shout...). Would this matter??? <- apologies for the stupid question
Fun, fun, fun!
Andy
You might rethink that heater wiring! A good read through Morgan Jones's "Valve Amplifiers" is highly recommended.
If you run a resistor back to the grid, the input impedance will drop to the value of the grid stopper (since the grid will now be a virtual ground), so that value will have to be greatly increased. That also means an increase in the feedback resistor. You'll probably need to have a compensation cap to keep things stable.
Polarity inversion is unlikely to have any noticeable effect. If you're worried about it, swap + and - on both of your speakers.
If you run a resistor back to the grid, the input impedance will drop to the value of the grid stopper (since the grid will now be a virtual ground), so that value will have to be greatly increased. That also means an increase in the feedback resistor. You'll probably need to have a compensation cap to keep things stable.
Polarity inversion is unlikely to have any noticeable effect. If you're worried about it, swap + and - on both of your speakers.
Circuit is the Aikido CCDA in which it is recommended that the heater (-) be referenced to B+/4, i.e. 50V. Hopefully that is what I have there.
So up R10 to something like 10K and play with that feedback resistor, cool. For that feedback resistor also use a capacitor in parallel with it for the 'compensation' cap you mention?
I really must get that book, but kids / ex-wife / new girlfriend sapping major funds from my shallow pockets! For the record, the aforementioned 'new girlfriend' is worth it!
Thanks again everyone! Once I've finished annoying everyone with my inane valve-based questions, rest assured I will go away back to my day job of designing MOSFET based high current, high volume motor controllers.
Andy
You have one side of the heater grounded, which will throw off all the DC voltages. Float the heater supply, then connect the bias to the centertap of the heater. Also, the heaters should be raised above the cathodes, not above ground. Since the cathode will be at roughly the same DC voltage as the plate of the previous stage, the biasing voltage needs to be above that, though not necessarily 50V higher.
+12V6 supply for the heaters is generated using a separate transformer / regulator setup from the rest of the circuit and does indeed float. The potential divider then brings the heater (-) up to 50V in reference to the main circuit GND connection. I have net labelled the heater ground supply rail 'GNDHtr', the valve circuit ground uses net label 'GND' and are therefore not connected together.
However, what you are suggesting is that the center tap should be the reference point held at B+/4? In other words bias one end of the series connected heaters to 50V-6.3V, i.e. heater (-) to 43.7V thus making the center tap potential 50V?
Andy
However, what you are suggesting is that the center tap should be the reference point held at B+/4? In other words bias one end of the series connected heaters to 50V-6.3V, i.e. heater (-) to 43.7V thus making the center tap potential 50V?
Andy
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.