I did some tests yesterday. At 1khz I got a relatively nice high output impedance, but at 200Hz, 100Hz and eventually 50Hz I had a couple dB loss and higher distortion. It was not the best OPT, but I do not think that was hte only problem. So, learned something new!
check out post #615+616 under
https://www.diyaudio.com/forums/solid-state/250272-current-drive-loudspeakers-62.html#post6707142
https://www.diyaudio.com/forums/solid-state/250272-current-drive-loudspeakers-62.html#post6707142
Local feedbacks
Do you still use local feedback such as partial bypass / cathode degeneration on pentode?
I was thinking that if the audio transformer is good enough then there is no need to take feedback from the secondary / output side. Single-ended transformers with their necessary air gaps are quite linear. The hysteresis loop is narrow, almost a straight line. So, if the signal is good on the primary side, it should be good also on the secondary side.
Do you still use local feedback such as partial bypass / cathode degeneration on pentode?
I was thinking that if the audio transformer is good enough then there is no need to take feedback from the secondary / output side. Single-ended transformers with their necessary air gaps are quite linear. The hysteresis loop is narrow, almost a straight line. So, if the signal is good on the primary side, it should be good also on the secondary side.
I use a bridge in feedback by current, such a way I can change output impedance from negative to positive changing feedback by current from positive to negative.
Feedback is just a useful tool, if you know how it works, and math is your friend.
Edelweiss-3 2021 production prototypes - YouTube
Feedback is just a useful tool, if you know how it works, and math is your friend.
Edelweiss-3 2021 production prototypes - YouTube
Took a quick look, results are based on opinions. Really needs to be tried on several listeners, young & old, male & female & the Family dog. In a double blind test.
The dog can act as arbiter.😀
Must have feedback
In many cases the inductance is much smaller.
It seems that feedback from the secondary side is necessary.
In many cases the inductance is much smaller.
It seems that feedback from the secondary side is necessary.
inductance is more needed in the primary, the secondary merely converts the high voltage low current primary to low voltage high current fed to the speakers, based on the turns ratio between primary and secondary...
TonyTecson,
Loudspeakers have mass, inertia, resonance, and back EMF.
If the loudspeaker EMF "looks back" at the secondary, (even if the primary is not present),
the secondary is a low impedance at bass frequencies.
The loudspeaker back EMF that "sees" a low impedance, will dampen the loudspeaker.
I stand by my illustration in Post # 111 that the amplifier without feedback has an output impedance that is largely influenced by the inductance of the secondary (at bass frequencies).
The example I gave, was with a typical secondary that was only 8 Ohms of inductive reactance at 40 Hz (even if the primary was mossing).
Loudspeakers have mass, inertia, resonance, and back EMF.
If the loudspeaker EMF "looks back" at the secondary, (even if the primary is not present),
the secondary is a low impedance at bass frequencies.
The loudspeaker back EMF that "sees" a low impedance, will dampen the loudspeaker.
I stand by my illustration in Post # 111 that the amplifier without feedback has an output impedance that is largely influenced by the inductance of the secondary (at bass frequencies).
The example I gave, was with a typical secondary that was only 8 Ohms of inductive reactance at 40 Hz (even if the primary was mossing).
Last edited:
opt primary impedance is not much a factor as far as inductance goes, inductance is a function of number of primary turns, core area, frequency and flux density...
flux density is also influenced by induced voltage in the primary as a result if plate voltage and current swings...
i am not saying you are wrong, just my appreciation of things..
flux density is also influenced by induced voltage in the primary as a result if plate voltage and current swings...
i am not saying you are wrong, just my appreciation of things..
Last edited:
Inductive reactance 3 Ω and load impedance 8 Ω at 40 Hz ... go figure
Think about it this way.
You have an ideal controlled current signal source and a load connected. Everything is fine. But then, add an inductance to the circuit so that the current signal source, load and inductance are all parallel-connected. If the inductance is not great enough it will spoil the signal in low frequencies. That is the situation with a pentode and an output audio transformer. Fortunately feedback from the secondary side may cure the signal.
Think about it this way.
You have an ideal controlled current signal source and a load connected. Everything is fine. But then, add an inductance to the circuit so that the current signal source, load and inductance are all parallel-connected. If the inductance is not great enough it will spoil the signal in low frequencies. That is the situation with a pentode and an output audio transformer. Fortunately feedback from the secondary side may cure the signal.
yes, that is how i get it also...a triode is a controlled voltage source, a pentode a controlled current source...
so between the two, the pentode is much more apt for current drive.....
so between the two, the pentode is much more apt for current drive.....
As always, the output transformer low inductive reactance 'gets in the way' at low frequencies, unless you have some form of current sensed feedback that makes the output a real current source, regardless of whether the output tube is a Pentode/Beam Power tube, or a Triode.
Years ago, I tested output transformer secondary impedance versus frequency.
There was nothing connected to the transformer other than the $50,000 Vector Network Analyzer that connected to the secondary.
The results were quite revealing.
Just my observations and opinions.
Years ago, I tested output transformer secondary impedance versus frequency.
There was nothing connected to the transformer other than the $50,000 Vector Network Analyzer that connected to the secondary.
The results were quite revealing.
Just my observations and opinions.
Last edited:
imho, the opt primary winding is where it is all at....the secondary winding merely translate what happened in the primary side converting primary power into safe levels for the speaker...
In the beginning of this thread I showed few simple circuit drawings. Later I realized that while the current feedback arrangement would be correct in principle, in practice the feedback would be too weak. So I guess it may be better to bring the feedback to the preceding amplifier stage. Then, I suppose, the current sense resistor may be kept small and feedback sufficient. Playing with resistance values becomes simple.
Start with an output transformer with ratings of:
Primary 100 Henry
5K Ohms to 8 Ohms
That is a turns ratio of 25:1
Inductance is proportional to turns ratio squared
25 squared = 625
100H / 625 = 0.16 Henry secondary
Z = 2x pi x F x L
At 20Hz, 0.16 Henry, Z = 20 Ohm secondary impedance.
A large woofer and enclosure might resonate at 20Hz.
The woofer impedance might be 25 Ohms at resonance.
A “current source” amplifier with a transformer secondary of 20 Ohms.
Is a 20 Ohm current source that drives a 25 Ohm woofer resonant impedance considered to be a current source?
Is a 20 Ohm current source that drives an 8 Ohm woofer considered to be a current source?
Good Luck with that one.
For a current source amplifier, either a pentode or a triode that drives the 5k primary is not the major factor at 20 Hz (or even at 40 Hz).
I think the current source amplifier needs some form of feedback to raise its output impedance that is limited by the output secondary inductive reactance.
Just my opinion.
Primary 100 Henry
5K Ohms to 8 Ohms
That is a turns ratio of 25:1
Inductance is proportional to turns ratio squared
25 squared = 625
100H / 625 = 0.16 Henry secondary
Z = 2x pi x F x L
At 20Hz, 0.16 Henry, Z = 20 Ohm secondary impedance.
A large woofer and enclosure might resonate at 20Hz.
The woofer impedance might be 25 Ohms at resonance.
A “current source” amplifier with a transformer secondary of 20 Ohms.
Is a 20 Ohm current source that drives a 25 Ohm woofer resonant impedance considered to be a current source?
Is a 20 Ohm current source that drives an 8 Ohm woofer considered to be a current source?
Good Luck with that one.
For a current source amplifier, either a pentode or a triode that drives the 5k primary is not the major factor at 20 Hz (or even at 40 Hz).
I think the current source amplifier needs some form of feedback to raise its output impedance that is limited by the output secondary inductive reactance.
Just my opinion.
Last edited:
Doesn't that make a case for OTL ?
Circlotron, no FB, something like an ATMA-Sphere 60, but with Pentodes in pentode mode, screen grids cross connected as in Electro-Voice's A50 ?
Maybe a single pair of EL509s, MOSFET driven into grid current, could do the job of three pairs of 6AS7s but with 300V supply rather than the usual 150V ...
Not many watts though but more than 20ohm output impedance and not limited by OPT reactance ...
Correct, an all pentode/beam power OTL stage, with plate output only might be a good choice for a current source amplifier.
But it can not be a totem pole output, because one of the two tubes in the totem pole is cathode output, which is a relatively low impedance.
However, generally, an OTL can be quite in-efficient. Lots of power in to get a much lower percentage power out.
Like most amplifiers, there are one or more tradeoffs.
A traditional push pull amplifier with an output transformer can be modified to become a current source amplifier, and it will be much more power efficient than an OTL amplifier.
I am talking about driving 4, 8, or even 16 Ohm nominal impedance loudspeakers; with an output transformer, versus with an OTL.
I merely wanted to point out that a push pull pentode/beam power amplifier with an output transformer, by itself does not make a good current source amplifier at all frequencies.
It must have a form of feedback that is designed to overcome the output transformers low impedance secondary.
But it can not be a totem pole output, because one of the two tubes in the totem pole is cathode output, which is a relatively low impedance.
However, generally, an OTL can be quite in-efficient. Lots of power in to get a much lower percentage power out.
Like most amplifiers, there are one or more tradeoffs.
A traditional push pull amplifier with an output transformer can be modified to become a current source amplifier, and it will be much more power efficient than an OTL amplifier.
I am talking about driving 4, 8, or even 16 Ohm nominal impedance loudspeakers; with an output transformer, versus with an OTL.
I merely wanted to point out that a push pull pentode/beam power amplifier with an output transformer, by itself does not make a good current source amplifier at all frequencies.
It must have a form of feedback that is designed to overcome the output transformers low impedance secondary.
Last edited: