I have 2 output transformers from an ARTEC 225 amplifier (EL34 PP). The indication is 5000/30W and the EI-iron measures 92x78x64.
3 individual outputs (I call them A,B,C), a FB? (F) and a CT primary (D-E). I measured the impedance @ 1kHz:
An 8 ohm @ A gives 7 ohm @ B, 13.5 ohm @ C, 32k2 ohm @ D-E and 2k32 ohm @ F
An 8 @ B 4.2 ohm @ A, 12 ohm @ C, 24k5 ohm @ D-E 1k98 ohm @ F
An 8 @ C 2.65. @ A, 3.65 ohm @ B, 13k45 ohm @ D-E 1k24 ohm @ F
An 8 @ A-B in series , 7k64 ohm @ D-E 775ohm @ F
An 8. @ B-C , 4k79 ohm @ D-E 488 ohm @ F
An 8 @ A-C , 2k7 ohm @ D-E 280 ohm @ F
I would like to build an amp with ECC82 as phase shift and EF86 as input stage. I also have 2 power transformers each 100VA/ 700VCT 5V 6V3 an 2 GZ34's (=5AR4).
3 individual outputs (I call them A,B,C), a FB? (F) and a CT primary (D-E). I measured the impedance @ 1kHz:
An 8 ohm @ A gives 7 ohm @ B, 13.5 ohm @ C, 32k2 ohm @ D-E and 2k32 ohm @ F
An 8 @ B 4.2 ohm @ A, 12 ohm @ C, 24k5 ohm @ D-E 1k98 ohm @ F
An 8 @ C 2.65. @ A, 3.65 ohm @ B, 13k45 ohm @ D-E 1k24 ohm @ F
An 8 @ A-B in series , 7k64 ohm @ D-E 775ohm @ F
An 8. @ B-C , 4k79 ohm @ D-E 488 ohm @ F
An 8 @ A-C , 2k7 ohm @ D-E 280 ohm @ F
I would like to build an amp with ECC82 as phase shift and EF86 as input stage. I also have 2 power transformers each 100VA/ 700VCT 5V 6V3 an 2 GZ34's (=5AR4).
Last edited:
Have a look at this Audionote schematic. It might give you a place to start or some ideas. It uses a pentode/triode for gain and phase splitting. Search around on the forum, there are threads about this amp and some have wired the pentode as a triode. It's a SS B+ supply but you could convert it to tube rectification. A 100VA power transformer might not be powerful enough for a mono EL34 amp and 700 volt winding might not have enough B+ with a tube rectifier.
Cheers, S.
Cheers, S.
Attachments
Some possible EL34 push pull combinations:
Pentode operation, global negative feedback, and hardest to compensate.
Highest power.
Williamson circuit needs the low frequency poles adjusted to prevent low frequency oscillation.
All circuit topologies will need careful high frequency compensation to prevent high frequency oscillation.
The output transformer model affects both the low frequency compensation, and the high frequency compensation.
Triode wired, no global negative feedback needed.
Less power
With proper implementation, not chance of oscillation.
Simplest to build, and get up and running.
Ultra Linear operation . . . output transformer has to have the UL taps. I think you do not have any UL taps.
Power is medium
global negative feedback may, or may not, be needed.
Degree of difficulty to make work well, lies between pentode and triode wired modes.
Have Fun!
Pentode operation, global negative feedback, and hardest to compensate.
Highest power.
Williamson circuit needs the low frequency poles adjusted to prevent low frequency oscillation.
All circuit topologies will need careful high frequency compensation to prevent high frequency oscillation.
The output transformer model affects both the low frequency compensation, and the high frequency compensation.
Triode wired, no global negative feedback needed.
Less power
With proper implementation, not chance of oscillation.
Simplest to build, and get up and running.
Ultra Linear operation . . . output transformer has to have the UL taps. I think you do not have any UL taps.
Power is medium
global negative feedback may, or may not, be needed.
Degree of difficulty to make work well, lies between pentode and triode wired modes.
Have Fun!
The Brimar 25P1 amplifier used Brimar tubes, but the 6BR7 is similar to the EF86, and the 5B/255M similar to a 6L6..
Valve amplifier output transformers are such a love hate thing. I love the primary secondary isolation between all the speaker wiring and the amplifier itself but I hate the always suspect linearity (and phase shifts) that transformers can't avoid.
richgwilliams,
You said:
"I hate the always suspect linearity (and phase shifts) that transformers can't avoid".
The major need for using an output transformer is to increase the power efficiency of a tube amplifier.
That is necessary to match the relatively High tube plate impedance, to the relatively Low loudspeaker impedance.
And, it makes it easier for the amplifier to have a reasonable damping factor, often with little or no global negative feedback.
And the loudspeaker can affect the linearity and stability of amplifiers that use lots of global negative feedback.
If you prefer, you can use OTL tube amplifiers, but be ready to accept the horrible power efficiency of the amplifier.
And in many cases, you need 16 Ohm speakers.
How much distortion of an amplifier will you accept?
How much phase shift of an amplifier will you accept?
And related to the amplifier:
How much distortion of a loudspeaker will you accept?
How much phase shift of a loudspeaker will you accept?
How much phase shift at 5kHz between the left channel and right channel will you accept?
If the distance from the left channel loudspeaker to your Left Ear, versus the right channel loudspeaker to the same Left Ear is 1.3 inches different,
The phase shift is 180degrees. We call that Cancellation of the 5kHz tone (2nd harmonic at 5kHz of a 2.5kHz fundamental for example)
Left channel distance 6 feet, right channel distance 6 feet 1.3 inches: All to the left ear.
Or, just turn you head slightly clockwise, left ear 0.65 inch forward, and right ear 0.65 inch backward, and the 5kHz harmonic overtone of the Soprano who is centered between the left and right channels . . .
The 5kHz 2nd harmonic of that opera singer is phase shifted 180 degrees between the left and right ear.
There are lots of other things that happen just like the above examples, when you know the complete system . . .
From singer, stereo microphone, recorder, mixing production house, CD player, preamp, power amp, loudspeakers, and your listening room.
How many good stereo systems with output transformers have you heard?
How many bad stereo systems with output transformers have you heard?
In my experience, output transformers are not the problem (other than cost, weight, and the real estate they take on the amplifier chassis).
You said:
"I hate the always suspect linearity (and phase shifts) that transformers can't avoid".
The major need for using an output transformer is to increase the power efficiency of a tube amplifier.
That is necessary to match the relatively High tube plate impedance, to the relatively Low loudspeaker impedance.
And, it makes it easier for the amplifier to have a reasonable damping factor, often with little or no global negative feedback.
And the loudspeaker can affect the linearity and stability of amplifiers that use lots of global negative feedback.
If you prefer, you can use OTL tube amplifiers, but be ready to accept the horrible power efficiency of the amplifier.
And in many cases, you need 16 Ohm speakers.
How much distortion of an amplifier will you accept?
How much phase shift of an amplifier will you accept?
And related to the amplifier:
How much distortion of a loudspeaker will you accept?
How much phase shift of a loudspeaker will you accept?
How much phase shift at 5kHz between the left channel and right channel will you accept?
If the distance from the left channel loudspeaker to your Left Ear, versus the right channel loudspeaker to the same Left Ear is 1.3 inches different,
The phase shift is 180degrees. We call that Cancellation of the 5kHz tone (2nd harmonic at 5kHz of a 2.5kHz fundamental for example)
Left channel distance 6 feet, right channel distance 6 feet 1.3 inches: All to the left ear.
Or, just turn you head slightly clockwise, left ear 0.65 inch forward, and right ear 0.65 inch backward, and the 5kHz harmonic overtone of the Soprano who is centered between the left and right channels . . .
The 5kHz 2nd harmonic of that opera singer is phase shifted 180 degrees between the left and right ear.
There are lots of other things that happen just like the above examples, when you know the complete system . . .
From singer, stereo microphone, recorder, mixing production house, CD player, preamp, power amp, loudspeakers, and your listening room.
How many good stereo systems with output transformers have you heard?
How many bad stereo systems with output transformers have you heard?
In my experience, output transformers are not the problem (other than cost, weight, and the real estate they take on the amplifier chassis).
@6A3sUMMER
OK but I'm not criticising output transformers far from it, I understand that they are needed to step the high tube impedance down to the low speaker impedance and as I said the decoupling they give between the tube amplifier and the speakers/speaker wiring is great (safety, RF interference removal, fault tolerance etc).
But transformers are never 100% linear, try designing an output transformer that is linear from 20Hz to 20kHz. The phase shift, I mentioned it not because of its effect on the sound coming from the speaker but because of its effect in the feedback loop (as most tube amp circuits I see feed the output transformer secondary signal back to the input tube). Phase shift in feedback loops is unwanted and hard to quantify.
I would go as far as to say that the output transformer primary to secondary isolation/decoupling does a lot to explain why so many people find the sound from a tube amp better than the sound from a solid state amp.
If I was designing an output transformer "all over again" I would build in some of the following features:
So as you see we are on the same page.
OK but I'm not criticising output transformers far from it, I understand that they are needed to step the high tube impedance down to the low speaker impedance and as I said the decoupling they give between the tube amplifier and the speakers/speaker wiring is great (safety, RF interference removal, fault tolerance etc).
But transformers are never 100% linear, try designing an output transformer that is linear from 20Hz to 20kHz. The phase shift, I mentioned it not because of its effect on the sound coming from the speaker but because of its effect in the feedback loop (as most tube amp circuits I see feed the output transformer secondary signal back to the input tube). Phase shift in feedback loops is unwanted and hard to quantify.
I would go as far as to say that the output transformer primary to secondary isolation/decoupling does a lot to explain why so many people find the sound from a tube amp better than the sound from a solid state amp.
If I was designing an output transformer "all over again" I would build in some of the following features:
- Make it a toroidal transformer with carefully chosen core material
- Keep the primary (centre tapped) winding and secondary winding simple, no extra primary taps or secondary taps (for different speaker impedances)
- Wind an extra winding (closely coupled to the primary) to give the needed feedback signal
- Keep the isolation between primary and secondary very good (as toroidal cores can do) for safety reasons
So as you see we are on the same page.
Richgwilliams,
I like your idea that transformers might be one of the differences between the sound of some solid stage amps, versus some tube amps (regardless of whether isolation is the cause, or other things are the cause of the sound difference.
But I was reminded of this:
Over 50 years ago, I was auditioning a pair of KLH 2-way loudspeakers.
The amplifier was a Solid State McIntosh . . . it had a multi-tap transformer with several output impedance selections.
I did purchase the KLH loudspeakers.
My first vacuum tube amplifier was a mono-block Knight Kit, with a 12AX7 phono preamp, and a 12AU7 gain stage and concertina phase splitter. The push pull EL84s drove a small output transformer. There was global negative feedback. It drove an Electro Voice Wolverine 12 inch woofer and Wolverine T35 tweeter in an EV Marquis enclosure, then later a Stephens full range replaced the woofer.
Using that system . . .
I remember at college, listening to orchestral music from LPs, and turning the volume Way Down, late at night, and seeing how much detail my friend's ears, and my ears could pick out when the dorm room was silent.
Early Hi Fi days were some of the most fun and interesting.
Over the years I have tried using the output secondary signal for global negative feedback, and also for negative feedback to the output stage cathode.
I also tried Schade negative feedback, and negative feedback similar to Schade but instead fed to the cathode of the driver tube. Then there was Ultra Linear negative feedback, and Triode Wired pentodes and Triode Wired beam power tubes. I still use Ultra Linear, but I mostly use triode wired beam power tubes.
Because I no longer use any kind of negative feedback other than Ultra Linear, and Triode Wired Pentodes and Triode Wired Beam Power tubes, I have not had trouble with amplifier output stage stability.
My latest amplifiers use balanced signals from a CD player with XLR 2-phase outputs. The input stage is balanced, and the output stage is balanced. The only negative feedback is Triode Wired Beam Power tubes (Look at it as if it is Pseudo Ultra Linear with a 100% Tap).
I like your idea that transformers might be one of the differences between the sound of some solid stage amps, versus some tube amps (regardless of whether isolation is the cause, or other things are the cause of the sound difference.
But I was reminded of this:
Over 50 years ago, I was auditioning a pair of KLH 2-way loudspeakers.
The amplifier was a Solid State McIntosh . . . it had a multi-tap transformer with several output impedance selections.
I did purchase the KLH loudspeakers.
My first vacuum tube amplifier was a mono-block Knight Kit, with a 12AX7 phono preamp, and a 12AU7 gain stage and concertina phase splitter. The push pull EL84s drove a small output transformer. There was global negative feedback. It drove an Electro Voice Wolverine 12 inch woofer and Wolverine T35 tweeter in an EV Marquis enclosure, then later a Stephens full range replaced the woofer.
Using that system . . .
I remember at college, listening to orchestral music from LPs, and turning the volume Way Down, late at night, and seeing how much detail my friend's ears, and my ears could pick out when the dorm room was silent.
Early Hi Fi days were some of the most fun and interesting.
Over the years I have tried using the output secondary signal for global negative feedback, and also for negative feedback to the output stage cathode.
I also tried Schade negative feedback, and negative feedback similar to Schade but instead fed to the cathode of the driver tube. Then there was Ultra Linear negative feedback, and Triode Wired pentodes and Triode Wired beam power tubes. I still use Ultra Linear, but I mostly use triode wired beam power tubes.
Because I no longer use any kind of negative feedback other than Ultra Linear, and Triode Wired Pentodes and Triode Wired Beam Power tubes, I have not had trouble with amplifier output stage stability.
My latest amplifiers use balanced signals from a CD player with XLR 2-phase outputs. The input stage is balanced, and the output stage is balanced. The only negative feedback is Triode Wired Beam Power tubes (Look at it as if it is Pseudo Ultra Linear with a 100% Tap).
Last edited:
1) I like the AN schematic Steve M. mentioned. (Rev3.03 july 2019 has 2 tube rectifiers and 630VCT)
2) 6A3sUMMER and richgwilliams discussed the FB winding.
So the question is still : can I use the 775 ohm winding (or is it better) an output winding ( see #1) for FB. If so what value change will you recommend.
Thank you all for the brainstorm
2) 6A3sUMMER and richgwilliams discussed the FB winding.
So the question is still : can I use the 775 ohm winding (or is it better) an output winding ( see #1) for FB. If so what value change will you recommend.
Thank you all for the brainstorm
To answer your question - I don't understand your measurements, your description does not define what the transformer is. I suggest you do the following tests/measurements to your transformer:
- With a DVM measure continuity between pins, a simple bleeper will do.
- Draw up a diagram of what you think the transformer is.
- You can tell which windings are separate/isolated because there is continuity between all pins of a winding, but no continuity between separate windings.
- You will get a drawing a bit like this but you still don't know how pins of each winding are arranged.
- Now you need to know which is primary winding which is secondary winding etc.
- Also which pin is centre tap etc
- And most importantly the ratio between windings, their interconnectivity and sense (polarity)
- So the best way to proceed is get a low voltage output AC transformer (bell transformer, heater supply etc) giving you typically 6 to 12 Volts AC at line frequency (50 or 60 Hz) and connect it say between pins 6 and 9 on my drawing
- Then measure the voltages between other pins for example 1 and 5 using a DVM set to AC volts, be aware that the turns ratio from 6,9 to 1,5 might be high say 1:30 so the voltage you measure will be high for a 6 VAC bell transformer you might see 6x30 or 180 Volts AC so safety consideration while you are testing.
- if you measure between 1 and 3 you will see half the turns ratio 1:15 and so on.
- Keep going until you understand the transformer, all its windings and turns ratios and sense (polarity)
The resistances you measured previously give you clues but don't tell you the turns ratios, rather when you know turns ratios from above you can work out how much audio power the transformer is likely to be comfortable handling. In some designs 1 and 5 might not be used (might not be there) in other designs they are used to help the Pentode Tubes.
@6A3sUMMER said "Ultra Linear operation . . . output transformer has to have the UL taps. I think you do not have any UL taps" I think he is referring to those pins 1 and 5 on my drawing, I agree and have seen those designs that need five pins on the primary side to connect to typically 2 x EL34 Pentode tubes.
If you can come back with a better description and drawing of your transformer someone can more likely answer your question. Hope that helps.
-
Attachments
Last edited:
"transformers might be one of the differences between the sound of some solid stage amps" Yes I am thinking that 20 or 30 metres of Speaker wiring is going to pick up HF Radio transmissions and Cellular phone noise and feed it straight into a solid state amplifier. Whereas it can't get through a good tube output transformer (differential isolation between primary and secondary).
"Because I no longer use any kind of negative feedback other than Ultra Linear, and Triode Wired Pentodes and" Yes I agree no negative feedback or only negative feedback at the Triode using the Cathode resistor. You might be interested in my thread called Vacuum Tube bias for designers its about a different way to bias Triodes and a headphone preamp design using GE JAN 5670 tubes. In any event I would welcome your opinion about this kind of bias.
At the moment I'm a bit lost when it comes to Tetrodes and Pentodes and wanting to do everything with Triodes only because of their simplicity.
It doesn’t matter whether transformer phase shift is audible or not. If it results in amplifier oscillation, you have a problem regardless. Avoid global NFB, then the only thing YOU have to worry about is whether you can hear coloration from the transformer. And even if you do it’s not the end of the world.
decramer,
and richgwilliams,
I did not see the connections of the transformer in question, until Post # 13.
I did not interpret and draw out the data that was given in post # 1.
It appears that Pins 1 and 5 normally connect to the plates (this is the highest primary impedance).
It also appears that Pins 2 and 4 are Ultra Linear taps (or else the transformer primary is sold as a Dual Primary impedance model).
But if those are 40% UL taps, then the impedance is 0.4 x 0.4 x the plate - plate impedance.
(example: I will use a 6k plate to plate rated primary. 6k p-p primary, would give 0.16 x 6k = 0.96k Ohms from UL - UL taps.
You want to know if the 775 Ohm winding can be used to give the negative feedback signal.
With a 7.64k primary, we have . . .
7,640 / 775 = 9.86 impedance ratio. The voltage ratio is Root (9.86) = 3.14. The 775 winding has 1/3.14, or about 1/3 of the plate to plate voltage.
First question: Can you use a negative feedback signal amplitude that is 1/3 of the plate to plate signal voltage, or will you have to attenuate it?
Second question: How good is the 775 winding coupled to Both the Push primary, and the Pull primary (is the leakage inductance low, and how equal is that leakage inductance to both the push and the pull primary halves)?
Third question, How equal is the 775 Ohm leakage reactance effect, versus the 4, 8, 16 Ohm leakage reactance effects . . . to the push and the pull primary halves?
Fourth question: do you really want to use global negative feedback?
From the 775 Ohm winding?
From the 4, or 8, or 16 Ohm winding that you connect your loudspeaker to?
On a different note . . .
A True Tetrode has the following:
1. Filament (or filament and cathode),
2. Control Grid,
3. Screen Grid,
4. Plate.
Tetra . . . from the Greek referring to 4. Thus, Tetrode
In a Tetrode . . . there is No Suppressor Grid, and there is No Beam Former, that Pentodes and Beam Power tubes have, respectively.
By the same rules, Pentodes and Beam Power tubes have 5 elements.
I once used a 4-65A RF Tetrode. I did not have a high voltage B+ (such as 800V to 2kV) . . .
Instead I grounded the plate, and used the screen as a plate with about 350 to 400V B+ (my 45 tube went gassy, glowed bright, and was good for either the museum or the trash bin.
My amplifier used the 4-65A screen (rated for 10 Watts maximum Screen dissipation, just like the 45 plate dissipation).
and richgwilliams,
I did not see the connections of the transformer in question, until Post # 13.
I did not interpret and draw out the data that was given in post # 1.
It appears that Pins 1 and 5 normally connect to the plates (this is the highest primary impedance).
It also appears that Pins 2 and 4 are Ultra Linear taps (or else the transformer primary is sold as a Dual Primary impedance model).
But if those are 40% UL taps, then the impedance is 0.4 x 0.4 x the plate - plate impedance.
(example: I will use a 6k plate to plate rated primary. 6k p-p primary, would give 0.16 x 6k = 0.96k Ohms from UL - UL taps.
You want to know if the 775 Ohm winding can be used to give the negative feedback signal.
With a 7.64k primary, we have . . .
7,640 / 775 = 9.86 impedance ratio. The voltage ratio is Root (9.86) = 3.14. The 775 winding has 1/3.14, or about 1/3 of the plate to plate voltage.
First question: Can you use a negative feedback signal amplitude that is 1/3 of the plate to plate signal voltage, or will you have to attenuate it?
Second question: How good is the 775 winding coupled to Both the Push primary, and the Pull primary (is the leakage inductance low, and how equal is that leakage inductance to both the push and the pull primary halves)?
Third question, How equal is the 775 Ohm leakage reactance effect, versus the 4, 8, 16 Ohm leakage reactance effects . . . to the push and the pull primary halves?
Fourth question: do you really want to use global negative feedback?
From the 775 Ohm winding?
From the 4, or 8, or 16 Ohm winding that you connect your loudspeaker to?
On a different note . . .
A True Tetrode has the following:
1. Filament (or filament and cathode),
2. Control Grid,
3. Screen Grid,
4. Plate.
Tetra . . . from the Greek referring to 4. Thus, Tetrode
In a Tetrode . . . there is No Suppressor Grid, and there is No Beam Former, that Pentodes and Beam Power tubes have, respectively.
By the same rules, Pentodes and Beam Power tubes have 5 elements.
I once used a 4-65A RF Tetrode. I did not have a high voltage B+ (such as 800V to 2kV) . . .
Instead I grounded the plate, and used the screen as a plate with about 350 to 400V B+ (my 45 tube went gassy, glowed bright, and was good for either the museum or the trash bin.
My amplifier used the 4-65A screen (rated for 10 Watts maximum Screen dissipation, just like the 45 plate dissipation).
Last edited:
Yes pins 1 and 5 should be the Anode 1 and Anode 2 and pins 2 and 4 should be your UL1 and UL2.
Comparing measured winding resistance at 1kHz does not tell the turns ratio, in fact winding resistance can be very low in an over designed transformer.
My transformer drawing is fictional and in no way represents info from post #1. The transformer details have to be discovered in ways I have described (also Googling the type numbers might help).
Comparing measured winding resistance at 1kHz does not tell the turns ratio, in fact winding resistance can be very low in an over designed transformer.
My transformer drawing is fictional and in no way represents info from post #1. The transformer details have to be discovered in ways I have described (also Googling the type numbers might help).
The impedance is measured @ 1kHz with a very close to 8 ohm resistor as explained in #1.
The transformer has a primary as drawn so D-CT-E. 3 single separated output windings that you can combine to have the LS impedance you want.
The output winding combination to have around 7k5 (7k65) was the 8 ohm load at the second and third winding in series.
Then there is 1 single winding of around 775 ohm, so i bet it can only be used as FB (or a 600 ohm output ?).
The transformer has a primary as drawn so D-CT-E. 3 single separated output windings that you can combine to have the LS impedance you want.
The output winding combination to have around 7k5 (7k65) was the 8 ohm load at the second and third winding in series.
Then there is 1 single winding of around 775 ohm, so i bet it can only be used as FB (or a 600 ohm output ?).
Last edited:
Those high values such as 7K65 are probably just capacitive coupling between isolated windings, if you measured at DC they would probably be open circuit. If that single winding at 775 Ohm was intended as a feedback winding then the high impedance is probably because they wound it will very thin wire (no current to carry for feedback). Also yes it could be intended as a high impedance, public address horn speakers historically and currently can use a 100 Volt line system (maybe not 600 Ohm but high impedance) idea was to allow long cable runs between horn speakers.
The impedances you measured don't tell you definitively about the turns ratios because different thicknesses of wire are used for different windings.
As an easy test - try feeding your 1kHz sine wave across D to E and see what 1KHz voltages you get at the speaker windings and the line/feedback winding.
Seems that you know much more about what the transformer is.
The impedances you measured don't tell you definitively about the turns ratios because different thicknesses of wire are used for different windings.
As an easy test - try feeding your 1kHz sine wave across D to E and see what 1KHz voltages you get at the speaker windings and the line/feedback winding.
Seems that you know much more about what the transformer is.