The CF has around 6 DB of feedback, and the GNF is another 6 db I never measured it.
On UL it has less feedback, but a lot of UL 50% tap.
In overall this morning I can see a reduction of -12 db of second harmonics, and -8 db of 3rd on the Cathode feedback, this is superior than CF+UL, or UL,
On UL it has less feedback, but a lot of UL 50% tap.
In overall this morning I can see a reduction of -12 db of second harmonics, and -8 db of 3rd on the Cathode feedback, this is superior than CF+UL, or UL,
I tuned the feedback by ear then I adjust the compensation filters with maths, check with scope, and adjust, usually the math works!, this is super stable, no oscillation ever, the LTP has some oscillations vs a cathode follower driver.
Williamson made an article long ago on Triode - UL - CF connections with scope pics and concluded that triode connection is the best, CF next & both better than UL.
Search this :
Wireless-World-1952-09 from world radio history site, article is there, "Amplifiers and superlatives" by Wiiliamson & Walker.
Search this :
Wireless-World-1952-09 from world radio history site, article is there, "Amplifiers and superlatives" by Wiiliamson & Walker.
Thanks Krca, I read the article briefly, and it completely lacks analysis, there are a bunch of formulas at the end that is it.
I am using a OT which has symmetrical windings which are computer programmed, it has very less ringing than a traditional E core and way better high frequency.
I listened to my kt120 amp (LTP) in triode mode and it was such a mess an confused sound with a halo of distortion, the harmonics were very bad and a lot of problems.
I can tell by ear right away something was not right, my ultralinear amplifier is decent sounding without global feedback.
I don't see the point of getting triode measurements , even with the new OT, I highly doubt that I would enjoy the sound.
I am using a OT which has symmetrical windings which are computer programmed, it has very less ringing than a traditional E core and way better high frequency.
I listened to my kt120 amp (LTP) in triode mode and it was such a mess an confused sound with a halo of distortion, the harmonics were very bad and a lot of problems.
I can tell by ear right away something was not right, my ultralinear amplifier is decent sounding without global feedback.
I don't see the point of getting triode measurements , even with the new OT, I highly doubt that I would enjoy the sound.
Instead I will focus on how the sound of the cathode feedback is better, how is it measurable besides the reduced THD.
I converted the whole amplifier to CF, one channel has 5 db more distortion, I will sort this out, it has also more gain,
I don't use the same test resistors so that could be the cause, I will just make sure by first changing the input tube
Then I will change the cathode follower tube which I doubt has any impact.
Then I will swap the output tube pairs in last resort.
I don't use the same test resistors so that could be the cause, I will just make sure by first changing the input tube
Then I will change the cathode follower tube which I doubt has any impact.
Then I will swap the output tube pairs in last resort.
I placed a 12sn7gt smoked RCA, which now sells from 250 to 350 $ !!!
I am just wondering if i didn't reversed the cathode feedback on one channel...
It has ringing interference and more THD when the bias of the output tubes is equal. The ringing disappear when I bias the tubes with 10ma DC... this is strange
I am just wondering if i didn't reversed the cathode feedback on one channel...
It has ringing interference and more THD when the bias of the output tubes is equal. The ringing disappear when I bias the tubes with 10ma DC... this is strange
Mmm I was mixing up the bias, on the THD I have both channels and I was turning the bias while looking at the wrong channel LOL
So yes, in Cathode feedback, one side has to be connected 16 ohm tap to tube output, the other to the 0 Ohm then 10R current resistor, On the other side you have to connect the 0 ohm of the Cathode feedback instead of the 16 ohm to the tube output cathode, then the 16 ohm goes to the 10R then ground.
So yes, in Cathode feedback, one side has to be connected 16 ohm tap to tube output, the other to the 0 Ohm then 10R current resistor, On the other side you have to connect the 0 ohm of the Cathode feedback instead of the 16 ohm to the tube output cathode, then the 16 ohm goes to the 10R then ground.
I not a designer of amplifiers, just had some EL34 pp with standard E/I opt and a switch for triode and pentode mode. Can't tell for which mode it was optimised but triode mode worked much better than pentode in this case and i almost never used it in pentode mode. Topology was ordinary common cathode gain stage for VAS, Concertina phase splitter and output tubes ( no UL taps ).
Trafomatic locally made some amps/toroid opts with CFD and thats what interests me ( OPTs with CFD possibility ).
Good luck with the amp.
Regards, Krca
Bias 52mA , Kt88, load 5K to 8 ohm, 6 db CFB and 6 db GNF ,
1.5 Watt:
0.02% thd , mainly 2nd harmonics. -86db third harmonics
-3db points at 10 watts : around 8hz, to 32.5khz.
The B+ is 440 V
The screen is 380 V
This is just the basis , the only difference in the final amplifier will be more film capacitors in the power supply, a dedicated transformer for the input stage which also feeds the pentode screens and maybe regulators for the pre amplifier sections.
A higher voltage mono power transformer with B+ around 500V, and Kt120 running at 75ma bias.
1.5 Watt:
0.02% thd , mainly 2nd harmonics. -86db third harmonics
-3db points at 10 watts : around 8hz, to 32.5khz.
The B+ is 440 V
The screen is 380 V
This is just the basis , the only difference in the final amplifier will be more film capacitors in the power supply, a dedicated transformer for the input stage which also feeds the pentode screens and maybe regulators for the pre amplifier sections.
A higher voltage mono power transformer with B+ around 500V, and Kt120 running at 75ma bias.
The pentode sound very , 'there' no smoothness but it is a bit harsh without the CFB and it has higher impedance.
With the 6 db of CFB it reduces greatly the harshness, the transformer is very different than toroids but I can see a big weakness,
In UL it has dominant 3rd harmonics and not very good sound, but it is less sensitive to bias problems,
In Pentode and CFB the bias is critical and cannot simply be equalized, it has a very narrow sweet spot where every distortion cancels out perfectly and the ringing is none.
However if the bias drifts more than 1mA from this sweet spot the transformer starts to ring on multiple signals, you cannot see
What makes the big difference is the output transformer... and the bias which I adjusted with a spectrometer to the sweetest spot with also the paraphrase adjusted to lower THD, there is no ringing at all, just pure music.
wow , this amp changed a lot to my sound system, its everything I ever wished, smooth sailing vocals, nothing harsh, but true to source, bass full there all the time and quite dry, just everything holds up together, high are extended without limit.
I know I can change the input cap... and redo the amp with a lower noise floor and a bigger better power supply and tweak a little the feedback and filters etc. but damn it satisfy a lot...
I will compare it to a SS state of the art amp and I really doubt the SS amp will touch this amp...
I know I can change the input cap... and redo the amp with a lower noise floor and a bigger better power supply and tweak a little the feedback and filters etc. but damn it satisfy a lot...
I will compare it to a SS state of the art amp and I really doubt the SS amp will touch this amp...
this is just a redrawn diagram of the AM-2 block 🙂
to tell the truth - with increased voltage to reduce TND