No examples available at the moment, but I've built a few prototypes with this concept and it works quite well. Drive levels can be a problem, a well designed driver stage is needed.
Finding suitable output transformers can be a challenge too, I've had some luck making my own but it's not easy.
Taking CFB from the secondary winding of a standard OPT can be a small step in the right direction but putting 10-20% of the primary winding between cathode and ground has a larger effect.
Finding suitable output transformers can be a challenge too, I've had some luck making my own but it's not easy.
Taking CFB from the secondary winding of a standard OPT can be a small step in the right direction but putting 10-20% of the primary winding between cathode and ground has a larger effect.
I have used single ended amplifiers with damping factors less than 1 (less than unity), and as high as 5.
Just a simple 8 Ohm rated 2 way loudspeaker, may have impedance versus frequency as low as 4 Ohms, and as high as 50 Ohms.
With such large impedance variations, some of those low damping factor amplifiers may not sound very good with a loudspeaker like that.
. . . But, you never know when you may find a combination that is a synergistic success.
Just a simple 8 Ohm rated 2 way loudspeaker, may have impedance versus frequency as low as 4 Ohms, and as high as 50 Ohms.
With such large impedance variations, some of those low damping factor amplifiers may not sound very good with a loudspeaker like that.
. . . But, you never know when you may find a combination that is a synergistic success.
It is possible to make a speaker with flat impedance for a SET or a no-feedback pentode amplifier.
1, Sub with its separate amplifier for frequencies below, say, 120 Hz.
2. Main SET amplifier reproducing only 120 Hz and up avoids the problem of speaker damping, which is only relevant for bass. Output transformer design is greatly simplified because it doesn't have to reproduce bass.
3. Mid driver with Fs of 50-60 Hz reproduces 120 - 1,000 Hz. Within this range it has flat largely resistive impedance.
4. Highs driver planar like BG Neo-8s for 1,000 Hz and up, with flat purely resistive impedance.
1, Sub with its separate amplifier for frequencies below, say, 120 Hz.
2. Main SET amplifier reproducing only 120 Hz and up avoids the problem of speaker damping, which is only relevant for bass. Output transformer design is greatly simplified because it doesn't have to reproduce bass.
3. Mid driver with Fs of 50-60 Hz reproduces 120 - 1,000 Hz. Within this range it has flat largely resistive impedance.
4. Highs driver planar like BG Neo-8s for 1,000 Hz and up, with flat purely resistive impedance.
Hi Fuling,
In my main speakers I use 12 planars, with circa 7 ohms each, from 600Hz upwards. All in series they are about 80R, a very nice load for OTLs. Recently I realized that this load is also very nice for CFB at about 10%, If I take a 115V+115V (in series, CT to B+) toroidal with 12V+12V secondaries, connect the secondaries in series with CT to ground, they are about 13% cathode feedback windings. The 80R load across the 12V+12V present about 8k to the primary, which is a nice load for a lot of tubes (I do not need much power, so more class A is welcome). The prototype worked, have not tested it on real speakers though.
sorry for the off topic!
Erik
To Zintolo,
a Zobel, Morgan Jones shows it in this article https://www.diyaudio.com/archive/ar...hly-modern-tube-phono-preamp-prev-thread.html (also in post 13)
Weird that the name in the url is SY's his masters noise, but the link goes to the Arpeggio from MJ.
Best regards, Erik
a Zobel, Morgan Jones shows it in this article https://www.diyaudio.com/archive/ar...hly-modern-tube-phono-preamp-prev-thread.html (also in post 13)
Weird that the name in the url is SY's his masters noise, but the link goes to the Arpeggio from MJ.
Best regards, Erik
Yes they can.
How? Someone who knows passive Xos needs stepping.
An example (which i can no longer find since they reved their site) were/are (some) Crites XOs that dramatically flatten the impedance of (some) Klipsch making them much moire suitable for SETs. https://critesspeakers.com/collections/new-crossovers
The Mk 6 version of the Elsinore has flat impedance, uses a lot of jiggery-pokering. https://www.diyaudio.com/community/threads/the-elsinore-project-thread.97043/
dave
How? Someone who knows passive Xos needs stepping.
An example (which i can no longer find since they reved their site) were/are (some) Crites XOs that dramatically flatten the impedance of (some) Klipsch making them much moire suitable for SETs. https://critesspeakers.com/collections/new-crossovers
The Mk 6 version of the Elsinore has flat impedance, uses a lot of jiggery-pokering. https://www.diyaudio.com/community/threads/the-elsinore-project-thread.97043/
dave
Sounds like a good idea!
I thought about something similar a year or two ago when I was experimenting with a pair of mains toroids with 2*115V primaries and dual 9 and 18V secondaries. I ended up using them as 5,6k/8R OPTs with the 9V windings in // for the speakers and the 18V windings for CFB.
Sounds just fine, but the low number of turns limits the LF performance a bit. Your idea should work very well, especially if the amp never sees anything below 600Hz.
Thanks Erik, but a zobel can only adjust the highest range of the impedance by paralleling a load that drops with frequency.
Klipsch speakers have many impedance peaks and dips on the frequency plot.
Zintolo, I agree! I commented as Planet10 is mostly into designs with full ranges, so first thought was how to tame those. But indeed, as shown, there are more difficult ones.
Mike C,
Your thread title is a good one:
A schematic, and a statement of the specifications; including damping factor, output power, etc. are needed.
Then, there are limits of the cost, size, and weight of the single ended amplifier, not to mention the multiple topologies of a single ended amplifier.
Does the term single ended also include: Parafeed amplifiers; and balanced quiescent current on a center tapped primary of a push pull amplifier output transformer (one CCS, and one driven output tube)?
Single ended amplifiers use DHT, pentodes, beam power tubes, indirect heated triodes, and Real Tetrodes (not pentode, not beam power).
RF tube types can apply too, such as True Tetrode bird cage grids (beam power, DH, 4 element tubes).
Single ended amplifiers use Ultra Linear, cathode negative feedback, global negative feedback, Schade negative feedback, nested negative feedback, and . . . No negative feedback.
I am pretty sure that not all principles of single ended amplifiers are the same;
And that the performance of different amplifiers are not the same, especially when different loudspeakers are used with them.
There are single ended amplifiers that put out as much as 200Watts.
Those have plenty of power for most loudspeakers.
There are loud speakers that have series LR, LC, and LCR networks, that are connected in parallel across drivers, that flatten the impedance curve of the loudspeakers. That takes care of the impedance variations.
These networks reduce the overall efficiency of the loudspeaker.
Without more specifics, the solution is obscure, like a blindfolded Archer trying to hit the target.
Your thread title is a good one:
SET for Real Speakers - Principles
I am not familiar with which model amplifier you are talking about.A schematic, and a statement of the specifications; including damping factor, output power, etc. are needed.
Then, there are limits of the cost, size, and weight of the single ended amplifier, not to mention the multiple topologies of a single ended amplifier.
Does the term single ended also include: Parafeed amplifiers; and balanced quiescent current on a center tapped primary of a push pull amplifier output transformer (one CCS, and one driven output tube)?
Single ended amplifiers use DHT, pentodes, beam power tubes, indirect heated triodes, and Real Tetrodes (not pentode, not beam power).
RF tube types can apply too, such as True Tetrode bird cage grids (beam power, DH, 4 element tubes).
Single ended amplifiers use Ultra Linear, cathode negative feedback, global negative feedback, Schade negative feedback, nested negative feedback, and . . . No negative feedback.
I am pretty sure that not all principles of single ended amplifiers are the same;
And that the performance of different amplifiers are not the same, especially when different loudspeakers are used with them.
There are single ended amplifiers that put out as much as 200Watts.
Those have plenty of power for most loudspeakers.
There are loud speakers that have series LR, LC, and LCR networks, that are connected in parallel across drivers, that flatten the impedance curve of the loudspeakers. That takes care of the impedance variations.
These networks reduce the overall efficiency of the loudspeaker.
Without more specifics, the solution is obscure, like a blindfolded Archer trying to hit the target.
Last edited:
If the target is big enough, even a blindfolded archer can more easily hit it.
If you start with a “big enough“ speaker (in the 100dB neighborhood), a SET can more easily hit its target.
If you start with a “big enough“ speaker (in the 100dB neighborhood), a SET can more easily hit its target.
I would argue that any added circuitry in the speaker's crossover with the purpose of "flattening" the amplifier load can only be an added burden to a triode amplifier, and can only increase distortion. If necessary, like a Zobel to make a driver compatable with the crossover, so be it, but arbitrarily attempting to make a speaker a "better" load for a triode amplifier by adding parasitic elements, I would argue is always wrong.
All good fortune,
Chris
All good fortune,
Chris
1. I forgot to say that after the Archer is blindfolded, you turn him around a few unspecified turns.
497 degrees, anybody? Hit the target!
2. A 104dB / Watt speaker that has impedance minima of 6 Ohms and maxima of 50 Ohms may, or may not, sound good if the amplifier has a damping factor of 1.0.
3. There are any number of "fixes" that make a speaker better, whether they actually do make things better for all amplifiers remains to be Seen . . .
Oh, remains to be Heard.
497 degrees, anybody? Hit the target!
2. A 104dB / Watt speaker that has impedance minima of 6 Ohms and maxima of 50 Ohms may, or may not, sound good if the amplifier has a damping factor of 1.0.
3. There are any number of "fixes" that make a speaker better, whether they actually do make things better for all amplifiers remains to be Seen . . .
Oh, remains to be Heard.
Buy the software by Audioreview in Italy
Lot of configuration possibilities and simulation on crossover and compensation
One of the best in the world. Only Italian unfortunately
One solution, I normally do, is to have an high ratios on OT even if you are using a low power triode
I use secondary fixed at 5-6 ohm and around 25-30 as ratio on 300B or 2A3 or Kt150 triode
With big iron and, if possible, a big wire (= low Rdc both sides)
This give a reasonable DF even you loose some rms power
Hello, taking advantage of this thread, I would like to ask a question about a doubt I have.
I have an amplifier set with 6C4C tubes that gives me 4.5w output. With it I plan to feed the mid and treble part of some speakers that I am setting up. Here the data:
-sensitivity 92 dB/2.8V/1 meter (MT).
-Impedance 8 Ohms for MT section, minimum 6.3 Ohms @ 1250 Hz.
-The set amplifier will push from 240Hz, from there down a class D amplifier will take care of it.
Attached is the impedance curve graph. Basically my question is the impedance peak seen on the graph. I think the set will provide a spl/db level of music listening loud enough for me, about 88db...am I right?
Thanks in advance
I have an amplifier set with 6C4C tubes that gives me 4.5w output. With it I plan to feed the mid and treble part of some speakers that I am setting up. Here the data:
-sensitivity 92 dB/2.8V/1 meter (MT).
-Impedance 8 Ohms for MT section, minimum 6.3 Ohms @ 1250 Hz.
-The set amplifier will push from 240Hz, from there down a class D amplifier will take care of it.
Attached is the impedance curve graph. Basically my question is the impedance peak seen on the graph. I think the set will provide a spl/db level of music listening loud enough for me, about 88db...am I right?
Thanks in advance
Ballpark of a half watt to hit 88 dB.
But (likely) if the SET has a high Rout you will hear that big impedance spike at 1kHz. 70Ω.
dave
But (likely) if the SET has a high Rout you will hear that big impedance spike at 1kHz. 70Ω.
dave