You need to beware that capacitive and inductive reactances are not expressed in Ohms, so can't be manipulated like resistors. They're each complex numbers, expressed as a sum of real and "imaginary" parts. Properly expressed, you can manipulate them as you're doing. Real plus imaginary.
All good fortune,
Chris
All good fortune,
Chris
zintolo,
I have never measured any Toroid transformer (other than the trifilar single ended to balanced DSL transformer that I designed and hand-wound).
I think if you really want to know the performance of that output transformer, you need to purchase one, and . . .
Either measure the bare bones transformer,
Or
Build a tube output stage using that transformer, and measure it as a unit.
There are too many unknowns, including the major capacitance that is distributed over what part of the primary windings, and what the other capacitances are over the other primary windings are .
Then you need to know about how the leakage inductance is distributed over those various primary windings, including the turns with the major capacitance, and the turns without the major capacitances.
All the calculation in the world will not solve the real performance, not without a more complete and accurate model of that transformer.
And, I am not prepared to spend that much time and effort to come up with the performance of such a transformer, even if I do have all the details.
I think we may be back to Purchase that transformer, or Select another model.
There has been more than once that I considered purchasing a single ended toroid output transformer.
Now, I will only purchase and use push pull transformers in my amplifiers.
Good Luck in your pursuits of your amplifier design.
I have never measured any Toroid transformer (other than the trifilar single ended to balanced DSL transformer that I designed and hand-wound).
I think if you really want to know the performance of that output transformer, you need to purchase one, and . . .
Either measure the bare bones transformer,
Or
Build a tube output stage using that transformer, and measure it as a unit.
There are too many unknowns, including the major capacitance that is distributed over what part of the primary windings, and what the other capacitances are over the other primary windings are .
Then you need to know about how the leakage inductance is distributed over those various primary windings, including the turns with the major capacitance, and the turns without the major capacitances.
All the calculation in the world will not solve the real performance, not without a more complete and accurate model of that transformer.
And, I am not prepared to spend that much time and effort to come up with the performance of such a transformer, even if I do have all the details.
I think we may be back to Purchase that transformer, or Select another model.
There has been more than once that I considered purchasing a single ended toroid output transformer.
Now, I will only purchase and use push pull transformers in my amplifiers.
Good Luck in your pursuits of your amplifier design.
gabdx,
Every working 300B has a point on a certain grid voltage line, at a specific plate voltage, and a specific plate current, where the plate impedance, rp, is exactly 700 Ohms.
But for more plate current and higher plate voltage on that same grid voltage line: the plate impedance is lower than 700 Ohms.
And, for less plate current and less plate voltage on that same grid voltage line: the plate impedance is higher than 700 Ohms.
For Newbies:
The slope of each and every grid line is not constant. Slope of Volts/Current = Plate Impedance, rp.
When you apply a straight (resistive) load line on several grid lines, you get a varying plate impedance, rp, along that load line.
That sounds like 2nd harmonic distortion to me; the dominant harmonic distortion of a single ended 300B (or other triode too).
Just my opinions
Every working 300B has a point on a certain grid voltage line, at a specific plate voltage, and a specific plate current, where the plate impedance, rp, is exactly 700 Ohms.
But for more plate current and higher plate voltage on that same grid voltage line: the plate impedance is lower than 700 Ohms.
And, for less plate current and less plate voltage on that same grid voltage line: the plate impedance is higher than 700 Ohms.
For Newbies:
The slope of each and every grid line is not constant. Slope of Volts/Current = Plate Impedance, rp.
When you apply a straight (resistive) load line on several grid lines, you get a varying plate impedance, rp, along that load line.
That sounds like 2nd harmonic distortion to me; the dominant harmonic distortion of a single ended 300B (or other triode too).
Just my opinions
Last edited:
It does clip harder if you use GNF. Period. To avoid hard clipping one needs to stay away from it. The only way to do so is to have more power than necessary which is NOT the case of most SE amplifiers. That is the main reason why most people prefer zero GNF SE amplifiers. High efficiency speakers help...
In order to have low distortion at low frequency, possibly not far from that at 1 KHz, the inductive reactance XL of the output transformer needs to be much higher (i.e. 8-10 times at least) than Req. Req is the parallel of the primary load resistance and the resistance of the tube. So this has nothing to do with increasing the load. In practice low distortion at low frequency can only be achieved if the plate resistance of the tube is low, like the 300B or similar, regardless if the transformer is 3.5K or 5K or whatever. An output transformer that saturates in the audio band is inadequate or limited, depending on the opinions. If the transformer is inadequate, GNF will only make it worse. If the transformer is adequate for the job, GNF does NOT restore anything because it's the tube that runs out of current (or the driver is not capable of driving the output tube in positive grid field). GNF will NEVER work if it attempts to drive the transformer into saturation or if a tube (driver or output) runs out of current. It will be just hard clipping. Not to mention that at low-to-medium power, where GNF is normally unnecessary, it sounds worse. It just does, according to the opinion of the vast majority of people. And that's what matters in the end, not the numbers if these are below the threshold of audibility. Such threshold is rather high. Something like 3% THD (with typical triode spectrum) at 90 dB SPL.....
Last edited:
@6A3sUMMER ,
I already have a pair of those, just wanted to know how to calculate everything that can be seen on scope or heard.
It will be used with GU50 or KT88 and a bootstrapped driver.
I already have a pair of those, just wanted to know how to calculate everything that can be seen on scope or heard.
It will be used with GU50 or KT88 and a bootstrapped driver.
To 6A3
I agree with you
The main problem is that too many calculation and simulation but , on field, the approach is scarce
If you haven’t a minimum of test set how you understand the real life?
In the simplest case you can use a good tester and a scope
You can arrange a simply circuit with a fixed bias and a trafo so you can vary the Ibias
Because, in other hand, all the calculations are made without considering the bias current who in many case determine the low frequency response and ( of course) Thd
Walter
I agree with you
The main problem is that too many calculation and simulation but , on field, the approach is scarce
If you haven’t a minimum of test set how you understand the real life?
In the simplest case you can use a good tester and a scope
You can arrange a simply circuit with a fixed bias and a trafo so you can vary the Ibias
Because, in other hand, all the calculations are made without considering the bias current who in many case determine the low frequency response and ( of course) Thd
Walter
I’m late to this thread but thought I’d pass along an anecdote about SE 300b transformers. Tomas of Ultrasonic Studios, a boutique amp manufacturer, went from Hammond to Soweter and finally found Monolithic Magnetics to be the best performers in his amp. Alas, he passed away before delivering that build to any customers.
https://ultrasonicstudios.org/telemachus/
He built my push pull amp for me and it is wonderful. Such a loss, and only 40 years old!
https://ultrasonicstudios.org/telemachus/
He built my push pull amp for me and it is wonderful. Such a loss, and only 40 years old!
grataku,
You asked: "Why the switch [from se] to pp?"
This thread is about an SE 300B (Which by definition is low to medium power)
I ask: "Have you tried, and have you 'lived with' a good low to medium power pp amplifier in your home system?"
Just asking . . .
soulmerchant,
My 300B single ended amplifiers used 2 tubes.
My push pull amplifiers use 3 tubes.
You asked: "Why the switch [from se] to pp?"
This thread is about an SE 300B (Which by definition is low to medium power)
I ask: "Have you tried, and have you 'lived with' a good low to medium power pp amplifier in your home system?"
Just asking . . .
soulmerchant,
My 300B single ended amplifiers used 2 tubes.
My push pull amplifiers use 3 tubes.
I should try to be more clear. Single ended can indeed use more than one output tubes/valves in parallel. But the OPT will need to be bigger than for the same number of output tubes/valves in a comparable PP amplifier.
For the same power output, PP will by nature still have less harmonic distortion than SE. But it is still a pleasure to build SE amplifiers.
6A3sUMMER - Are the above numbers per channel? I tend also to build SE amplifiers with a single gain/driver stage so for a stereo amplifier I end up with 4 valves/tubes. Then I add two rectifying TV damper diodes so that makes 6 valves/tubes.
However, there are some tubes which can be used for dual-duty (pentode and triode in the same envelope). If you did SS rectification then I suppose you could get away with literally only 2 tubes/valves for a 2-channel amplifier, and 3 tubes/valves for PP*.
Ian
* I use a MOSFET as a concertina phase splitter for PP and its performance is superb.
Last edited:
Ian,
My single ended amplifiers used a single triode (or single glass envelope paralleled dual triode), and one output tube.
The one exception to that tube count was for the special case SE amplifier I built with parallel output tubes.
It was used both for the Glass Audio "Paralleling Tube Effects" article;
and for the VSAC 2003 theory and listening test to compare parallel 300B SE, with a Push Pull 300B amplifier).
They both used a dual triode in one glass envelope for the input/driver stage for SE, or input/phase inverter/driver stage for the Push Pull.
With one exception, a stereo push pull 45 amplifier; all my amplifier designs have been mono-blocks.
Long ago I gave up using vacuum tube rectifiers; I now use solid state diodes.
And long ago, I gave up capacitor input filters; now I use choke input filters.
Everybody has their preferences, and sometimes even those change over many years designing, building, and listening.
My push pull amplifiers either use a true LTP phase splitter input stage, or have a self inverting output stage, or are fully balanced from dual triode input tube to the output tubes. 3 tubes, 3 tubes, and 3 tubes; all are mono-blocks.
One of my self inverting output stage push pull still uses real triodes (6CK4).
All my other push pull amplifiers use Beam Power tubes, either triode wired or UL.
I tried lots of variations of SE, and lots of variations of push pull.
Early on, some of my amplifier designs used battery bias, interstage transformers (SE; and with LTP triodes driving the push pull interstage).
None of those topologies for a long time now.
My single ended amplifiers used a single triode (or single glass envelope paralleled dual triode), and one output tube.
The one exception to that tube count was for the special case SE amplifier I built with parallel output tubes.
It was used both for the Glass Audio "Paralleling Tube Effects" article;
and for the VSAC 2003 theory and listening test to compare parallel 300B SE, with a Push Pull 300B amplifier).
They both used a dual triode in one glass envelope for the input/driver stage for SE, or input/phase inverter/driver stage for the Push Pull.
With one exception, a stereo push pull 45 amplifier; all my amplifier designs have been mono-blocks.
Long ago I gave up using vacuum tube rectifiers; I now use solid state diodes.
And long ago, I gave up capacitor input filters; now I use choke input filters.
Everybody has their preferences, and sometimes even those change over many years designing, building, and listening.
My push pull amplifiers either use a true LTP phase splitter input stage, or have a self inverting output stage, or are fully balanced from dual triode input tube to the output tubes. 3 tubes, 3 tubes, and 3 tubes; all are mono-blocks.
One of my self inverting output stage push pull still uses real triodes (6CK4).
All my other push pull amplifiers use Beam Power tubes, either triode wired or UL.
I tried lots of variations of SE, and lots of variations of push pull.
Early on, some of my amplifier designs used battery bias, interstage transformers (SE; and with LTP triodes driving the push pull interstage).
None of those topologies for a long time now.
Last edited:
How would a pair of LL1688/85mA do as output transformers in a 300B SET? I have a pair that from a transmitter tube project 15-20 years ago that never happened, it would be nice to use them in a slightly less deadly project. They can be wired as 5,5k/8R with 0,5dB insertion loss, I guess the 85mA airgap is a bit larger than necessary for 300B but they should still have a rather healthy primary inductance, 50 Hy or so.
