My point is that whatever you do by altering the plate resistance of the triode, primary impedance etc...it will never be as efficient as the pentode and so for a reasonable 6dB local feedback the driver will have to work harder and it's not worth the trouble, IMHO. Instead, for the same amount of fbk in pentode mode, requirement is a lot less and despite the modest amount of local fbk both distortion and Zout will drop dramatically behaving like a triode. Actually from my last EL34 class AB amplifier with 7 dB cathode fbk, I can assure you that distortion is lower than triode at ALL levels and output power is 60W (with just 420V anode voltage and 400V on g2) instead of 18W. The EL34 is one of the best pentodes in terms of both distortion and efficiency when triode-wired. The EL34s I use are selected but regular and affordable JJ's.
I would only go for triode if I wanted a simpler amplifier without any feedback.
Last edited:
pentode mode without global negative feedback is possible, it's been done before, see RCA TS RC 19.http://www.myav.com.tw/bbs/printthread.php?s=a493a8f4406acb498aa13e584ca2a659&threadid=20479539
RCA Tube Manual RC-19, http://www.tubebooks.org/tubedata/RC19.pdf
RCA Tube Manual RC-19, http://www.tubebooks.org/tubedata/RC19.pdf
Attachments
Last edited:
45,
Yes, agreed.
I have tried Schade negative feedback circuits. I liked the idea, and the performance was reasonable (it could have been better with more adjustments on my part).
But if I ever use Schade NF again, and if it is push pull, then the driver stage will either be 2 pentodes, or 2 triodes.
It will Not be one pentode and one triode (mixed technology for a balanced circuit is not my preference for a phase inverter).
But having tried pentode operation, beam power operation, ultra linear operation, and triode wired operation, I generally prefer triode wired.
I am not looking for lots of gain, and I am not looking for lots of power.
I do not want to use global negative feedback; and I do not like the intrinsic 2 low frequency poles that are inside of the Schade NF loop.
(driver to output stage RC coupling, and output transformer inductance are inside of the Schade loop; and I do not want to DC couple the driver to the output stage).
Just for fun, I did a highly modified "Schade negative feedback-"like" circuit, for a single ended amplifier.
A resistor and series capacitor were connected from the output tube plate, to the driver tube cathode.
I did not really like that circuit either. Of course the gain of the amplifier increased at low frequencies (yes, I know, a large enough capacitor will delay the gain from increasing until the frequencies are subsonic.
Yes, agreed.
I have tried Schade negative feedback circuits. I liked the idea, and the performance was reasonable (it could have been better with more adjustments on my part).
But if I ever use Schade NF again, and if it is push pull, then the driver stage will either be 2 pentodes, or 2 triodes.
It will Not be one pentode and one triode (mixed technology for a balanced circuit is not my preference for a phase inverter).
But having tried pentode operation, beam power operation, ultra linear operation, and triode wired operation, I generally prefer triode wired.
I am not looking for lots of gain, and I am not looking for lots of power.
I do not want to use global negative feedback; and I do not like the intrinsic 2 low frequency poles that are inside of the Schade NF loop.
(driver to output stage RC coupling, and output transformer inductance are inside of the Schade loop; and I do not want to DC couple the driver to the output stage).
Just for fun, I did a highly modified "Schade negative feedback-"like" circuit, for a single ended amplifier.
A resistor and series capacitor were connected from the output tube plate, to the driver tube cathode.
I did not really like that circuit either. Of course the gain of the amplifier increased at low frequencies (yes, I know, a large enough capacitor will delay the gain from increasing until the frequencies are subsonic.
Last edited:
Tonytecson,
You said:
"Triodes are like Thevenin voltage sources, Pentodes are like Norton's current sources..."
The work 'like' makes the generalization correct.
But triodes are not perfect voltage sources, they have plate impedance, rp.
And pentodes are not perfect current sources, they have plate impedance, rp, also.
You said:
"Triodes are like Thevenin voltage sources, Pentodes are like Norton's current sources..."
The work 'like' makes the generalization correct.
But triodes are not perfect voltage sources, they have plate impedance, rp.
And pentodes are not perfect current sources, they have plate impedance, rp, also.
Last edited:
Certainly, quality EL34 type tubes - including NOS variants - seem to be a lot cheaper and easier to obtain than 6550 / KT88 types.
I don't need any more than 15 good watts of TRIODE output, so I may look at moving the BEWITCH to EL34 operation further down the track. I don't think it would be too difficult.
I also wonder if using EL34s with lower output power, perhaps the BEWITCH OPT will be less of a bottleneck?
45,
I too have used a pair of JJ EL34 in triode wired push pull operation with no negative feedback.
And then I used a pair of JJ KT77 in triode wired push pull operation with no negative feedback, in the same amplifier.
I really liked both of those.
I too have used a pair of JJ EL34 in triode wired push pull operation with no negative feedback.
And then I used a pair of JJ KT77 in triode wired push pull operation with no negative feedback, in the same amplifier.
I really liked both of those.
nothing in this world are ever perfect, once you realize this, then you gain the ease of mind and not wrack your brains on impossible dreams...just get as close to the goal post and try your best...
a look at the tube curves on data sheets will make you see all this....
i put a lot of trust on the engineers that designed the RCA tubes and tube circuits, they know what they were doing,
i also keep reminding people, that aside from tubes, the topology, and passive parts, there is the output transformers that really impacted the sound, and the psu itself is 50% of any amplifier....you can not cherry pick, instead, judge the amp on its entirety, not on its piece parts....the amplifier is bigger than the sum of its parts...
oh, and you still have to connect that amp to a pair of speakers, and therein lies the exciting part....
after a while tube amps become boring...
it may surprise you to know, you will not be putting out more than 5 watts of power to your speakers....
but why use more powerful amp? and my answer would be, transients, music are a lot of those,
better have power even though you may not need it..... than need it and not have it....
Having the right conditions to get the best out of it makes a significant difference. I don't use Schade fbk either so I won't make more comments on that, I only use cathode fbk which is even better for the driver. My output transformers are 5K, have bifilar primary and 10% cathode fbk (5K is actually the total primary load with 90% of turns at the plates and 10% at the cathodes). Good quality EI core, nothing fancy. I confess I have never looked at poles, possible compensation etc because I have not seen any instability with any load I have fed it with. Thanks to bifilar primary also cross-over distortion is inexistent even if the output tubes are biased at 30 mA each, however I prefer 50 mA anode current for each tube (same idea as the original McInstosh class B design but here is definitely class AB....). FR only drops beyond 150 KHz. 20KHz near perfect square waves (they would look perfect if the rise time were a bit faster but definitely flat top with no detectable ringing and very symmetrical).
Last edited:
No idea about your output transformers but for sure, less power = less distortion at low frequency.
output transformers have that "funny business" as coined by Patrick Turner formed with the resonance of the leakage inductance and shunt capacitance of the OPT, he mentioned 70khz as well beyond the audio band, and that is where you set your dominant pole, usually at the first voltage gain stage..
45,
For a single pole high frequency roll off, 150kHz at -3dB has a rise time of 2.33usec.
That is very good.
I wonder if your output transformers are one or two pole roll off; given any distribted capacitance, and any leakage inductance from primary to secondary.
All my output transformers do have those parasitics; they are two pole high frequency roll off (most of my amplifiers are -3dB at 50kHz, including not just the output transformer, but also miller effect, etc.).
My amplifiers at -3dB at 50kHz, are not a concern for me.
My ears, and my tweeters do not respond to 50kHz.
And I do not have any music signal sources that put out 50kHz.
For a single pole high frequency roll off, 150kHz at -3dB has a rise time of 2.33usec.
That is very good.
I wonder if your output transformers are one or two pole roll off; given any distribted capacitance, and any leakage inductance from primary to secondary.
All my output transformers do have those parasitics; they are two pole high frequency roll off (most of my amplifiers are -3dB at 50kHz, including not just the output transformer, but also miller effect, etc.).
My amplifiers at -3dB at 50kHz, are not a concern for me.
My ears, and my tweeters do not respond to 50kHz.
And I do not have any music signal sources that put out 50kHz.
TonyTecson,
Having a more powerful amplifier is great.
So is owning an NSX or a Lamborghini.
All of the above are very much alike. In my area, you never get the chance to wind them up to full output.
Currently, I do not have any global negative feedback on my amplifiers.
I have cathode to cathode negative feedback for phase splitters; cathode to cathode negative feedback for balanced stages; plate to screen negative feedback beam power outputs; and plate to plate negative feedback of push pull stages.
Is a Dominant Pole needed? . . . No.
Having a more powerful amplifier is great.
So is owning an NSX or a Lamborghini.
All of the above are very much alike. In my area, you never get the chance to wind them up to full output.
Currently, I do not have any global negative feedback on my amplifiers.
I have cathode to cathode negative feedback for phase splitters; cathode to cathode negative feedback for balanced stages; plate to screen negative feedback beam power outputs; and plate to plate negative feedback of push pull stages.
Is a Dominant Pole needed? . . . No.
Last edited:
Honestly I don't know. I have only set the output at 2.83V rms @1 KHz into 8R and checked on the oscilloscope where it drops at 2V rms and that was somewhere between 150 and 160 KHz.
The Miller capacitance is also another advantage of pentode + cathode fbk. Much lower than triode and can use medium-to-high gain preamp tubes without many troubles.
Last edited:
I think I have found the problem, causing the sound of my amp's sound becoming hard and dry.
Reducing the second 6SN7 cathode resistor R6 from 18K down to 10K means that there is presently more than 21mA of current on the second 6SN7.
This is beyond the recommended maximum of 20mA and is no-doubt shortening the life of my 6SN7 tube!
My next step - today - is to increase R6 to 12K - and measure again - in the hopes that the plate will idle with less than 20mA and that the (new) harsh sound departs from the amp. I feel confident this is the issue.
Reducing the second 6SN7 cathode resistor R6 from 18K down to 10K means that there is presently more than 21mA of current on the second 6SN7.
This is beyond the recommended maximum of 20mA and is no-doubt shortening the life of my 6SN7 tube!
My next step - today - is to increase R6 to 12K - and measure again - in the hopes that the plate will idle with less than 20mA and that the (new) harsh sound departs from the amp. I feel confident this is the issue.
have you tried beefing up the psu, like i like to use 1000ufd/450volts in my tube builds...
beating up the topology is good but beefing up the power supply, very exciting...
beating up the topology is good but beefing up the power supply, very exciting...