As title, can someone explain what is the difference between Self Inverting Push Pull and Anti-triode configuration? Are they single ended or push pull?
Example of SIPP
Example of Antitriode
Example of SIPP
Example of Antitriode
Please note that in my schematic the LM317's were not robust enough and needed replacing with a higher voltage compliance version.
The difference between the two linked circuits is that in the anti-triode the none driven pentode is a more perfect follower than in the SRPP driven one. This means that it will contribute less to the overall sound of the circuit. It is there simply to follow the driven triode in the best way possible.
Shoog
The difference between the two linked circuits is that in the anti-triode the none driven pentode is a more perfect follower than in the SRPP driven one. This means that it will contribute less to the overall sound of the circuit. It is there simply to follow the driven triode in the best way possible.
Shoog
Last edited:
So they are the same principle despite the performance difference?
Would the SIPP one be better as a follower if the CCS is replaced with something "stiffer" such as cascoded CCS?
I should add that based on my simulation to your Antitriode, i got Single Ended harmonic profile. Would the SIPP yield the same harmonics?
Would the SIPP one be better as a follower if the CCS is replaced with something "stiffer" such as cascoded CCS?
I should add that based on my simulation to your Antitriode, i got Single Ended harmonic profile. Would the SIPP yield the same harmonics?
Last edited:
They are the same principle of operation but mine adds shunt feedback and the more perfect antitriode (pentode).
Stiffening the CCS is not the issue here. The pentode is getting its signal from its cathode and it is a better follower of that signal than a triode in the same position. This effectively means that it is contributing less of its own character to the output signal and so the driven triode is the dominant distortion source.
I would guess that the SIPP would have a more PP signature, though for the trouble it would take I would suggest modelling it to see.
I would recommend a CCS in mine with at least the voltage capability of the full supply voltage. The switch on transients in this design are extreme and it will take down a lesser device in short order. I would recommend going straight to the IXCP10M45 or IXCP10M90.
Shoog
Last edited:
In the SIPP amplifier the second pentode is wired in exactly the same way as the first, ie both are in UL configuration. This is more like a triode PP pair amplifier. A UL triode will not follow the signal at its cathode as well as a pentode in the same position because it has an effectively lower gm than the pentode.
In the Antitriode design yes the shunt feedback is R12. This makes the driven pentode V1 behave more like a triode than a pentode. As such V1 is in control of the distortion components and the pentode V2 is slavishly following it because it is behaving as a pentode and has a higher gm.
There are two main components to my design - the shunt feedback which turns the driven pentode into a quasi-triode and the second pentode which follows it. For this to work well - you absolutely need the driver valve to be a pentode such as the 6AU6.
Shoog
In the Antitriode design yes the shunt feedback is R12. This makes the driven pentode V1 behave more like a triode than a pentode. As such V1 is in control of the distortion components and the pentode V2 is slavishly following it because it is behaving as a pentode and has a higher gm.
There are two main components to my design - the shunt feedback which turns the driven pentode into a quasi-triode and the second pentode which follows it. For this to work well - you absolutely need the driver valve to be a pentode such as the 6AU6.
Shoog
Last edited:
In essence there is no difference; they are just new names for an old idea: the differential amplifier, aka long tailed pair, aka cathode-couple amplifier. They are class-A push pull. Both types have CCS 'tails' so are very well balanced for AC. Shoog's version has individual CCSs for both valves so they are balanced for DC too, but both designs are long tailed pairs.
The main difference in my design is the application of SHUNT FEEDBACK and its effects on making the two PP elements behave differently to each other. The driven side is effectively a triode and the follower side is effectively a pentode.
Shoog
Last edited:
Opposing the triode."AntiTriode" ??? a triode with a negative anode and positive cathode !!!
Mona
Br Cornelius
In theory, but you may have issues getting the DC to balance if you attempt that.
The twin CCS arrangement in mine is specifically there to address this issue of uneven DC bias since the CCS will individually bias up to the correct DC potential for the mismatched halves of the PP pair. I wouldn't count on that with the SIPP unless you change the cathode arrangement to something like mine.
You may ultimately consider all of this not worth the extra components and trouble and that is where modelling both will help you decide.
Shoog
"In essence there is no difference; they are just new names for an old idea: the differential amplifier, aka long tailed pair, aka cathode-couple amplifier."
The difference is subtle. One side of the pair must have much higher gm than the other side for an "anti-triode" effect. (no, its not an upside down or reversed voltage triode! And it was originally conceived using a Mosfet for the anti device, so some called it the anti-Triode (as in anti-Christ) design... the design from H_ll.)
The point of the high gm device is to accurately mirror (actually counter mirror via the CCS tail effect) the signal + distortion of the driven device. The high gm device holds the tail (cathodes) voltage near constant, so that the usual tail voltage variation cannot cause cancellation of the even harmonics as in typical differential setups. (ie, the driven device is kept in a clean SE environment) So as was mentioned above, the harmonic spectra include even harmonics still, like a SE amp has.
This makes for a convenient way to get SE using a P-P OT and with twice the usual power output of the driven device.
There was also another form of the original anti-triode circuit, which looks like an SRPP
setup, except the top device was a depletion mode Mosfet (- gate bias like a JFet, IXTP01N100D) and the output was taken from a low equal resistance (Isensing) divider between the Source and Plate. (gate drive taken from the plate connection) This was designed so that the sum of the tube and Mosfet currents (Isense divider derived) sum to a constant (= the bias V on the Mosfet).
One should be aware however that the lopsided 2nd harmonic effects (at large signal level) on the OT in the P-P OT case can cause a couple of % DC imbalance in the OT. This can be countered by offseting the initial biases slightly, or by using a slow acting DC bias servo (audio envelope level tracking speed).
The difference is subtle. One side of the pair must have much higher gm than the other side for an "anti-triode" effect. (no, its not an upside down or reversed voltage triode! And it was originally conceived using a Mosfet for the anti device, so some called it the anti-Triode (as in anti-Christ) design... the design from H_ll.)
The point of the high gm device is to accurately mirror (actually counter mirror via the CCS tail effect) the signal + distortion of the driven device. The high gm device holds the tail (cathodes) voltage near constant, so that the usual tail voltage variation cannot cause cancellation of the even harmonics as in typical differential setups. (ie, the driven device is kept in a clean SE environment) So as was mentioned above, the harmonic spectra include even harmonics still, like a SE amp has.
This makes for a convenient way to get SE using a P-P OT and with twice the usual power output of the driven device.
There was also another form of the original anti-triode circuit, which looks like an SRPP
setup, except the top device was a depletion mode Mosfet (- gate bias like a JFet, IXTP01N100D) and the output was taken from a low equal resistance (Isensing) divider between the Source and Plate. (gate drive taken from the plate connection) This was designed so that the sum of the tube and Mosfet currents (Isense divider derived) sum to a constant (= the bias V on the Mosfet).
One should be aware however that the lopsided 2nd harmonic effects (at large signal level) on the OT in the P-P OT case can cause a couple of % DC imbalance in the OT. This can be countered by offseting the initial biases slightly, or by using a slow acting DC bias servo (audio envelope level tracking speed).
Last edited:
I made my amp basically like Shooq's but with EL34s and 200W toroids, It works fine and I have noted no change in sound over a couple of years use. I have it packed up for moving but when I next get the opportunity I will check out the Ia on each half of the toroids. I don't drive it very hard probably never more than a few watts but is much clearer than my EL84 SE amp.
Who cares about names this is the poor man's way to cheap audio power and household heating with minimum sand ( too many people with too much silicon these days) and without those Kilo$$$$ lumps of Fe. What about the cancer implications caused by radiated fields from Lunddahls and Hammonds etc? No anti snurgle skin effect DC to UHF gold plated silver speaker cables for me, just good old Cu.
Had it not been possible to use "Cheap" toroids as O/P transformers with this type of circuit I for one would never have gotten past the SE EL84 or 30W sand amp. This is where the satisfaction starts. As for the 813 tubes, I bought 3, one used and 2 NIB for US$30 on ebay in 2003, The used one has already been pressed into service for something else while the others ( quite well matched) are still waiting.
Had it not been possible to use "Cheap" toroids as O/P transformers with this type of circuit I for one would never have gotten past the SE EL84 or 30W sand amp. This is where the satisfaction starts. As for the 813 tubes, I bought 3, one used and 2 NIB for US$30 on ebay in 2003, The used one has already been pressed into service for something else while the others ( quite well matched) are still waiting.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.