Is it necessary to install a RCL circuit in the speaker crossover when driving them with a tube-amp
To make a impedancecorrection?
Because of the tube-amp impedance rise at 1000 Hz.?
To make a impedancecorrection?
Because of the tube-amp impedance rise at 1000 Hz.?
1. What is the tube amp that has a rise of impedance at 1kHz?
Please list the model number.
Perhaps you meant the impedance rise of the Loudspeaker at the crossover frequency.
2. I have many loudspeakers, about 10 or more pairs of different models.
They all sound good on my vacuum tube: Balanced PP amplifiers, SE amplifiers; and PP amplifiers.
I have many models that do not have a match to make a pair, they all sound good on my vacuum tube amplifiers too.
As far as I know, my loudspeakers do not have a Series LC, Series RC, nor Series RLC impedance correcting networks.
I measure my loudspeaker impedance variations. Most of them have widely varying impedance rises and dips.
Most of my vacuum tube amplifiers have damping factors of from 2.5 to 5.
Just my opinions and experience.
Your Mileage May Vary.
Have Fun!
Please list the model number.
Perhaps you meant the impedance rise of the Loudspeaker at the crossover frequency.
2. I have many loudspeakers, about 10 or more pairs of different models.
They all sound good on my vacuum tube: Balanced PP amplifiers, SE amplifiers; and PP amplifiers.
I have many models that do not have a match to make a pair, they all sound good on my vacuum tube amplifiers too.
As far as I know, my loudspeakers do not have a Series LC, Series RC, nor Series RLC impedance correcting networks.
I measure my loudspeaker impedance variations. Most of them have widely varying impedance rises and dips.
Most of my vacuum tube amplifiers have damping factors of from 2.5 to 5.
Just my opinions and experience.
Your Mileage May Vary.
Have Fun!
All things are recommended.
Some recommendations only work in certain situations.
Many that work, have tradeoffs.
All recommendations are equal, but some recommendations are more equal than others.
One of the key words is Series Resonant. Some will resonate more than others.
Some will resonate in your ears.
Qs of 3 or less are considered by many to be non-resonant.
Qs of 4 or more are considered by many to be resonant.
Radio Days (Daze) keep coming back, and so does the old ARRL Radio Amateurs Handbook.
Some recommendations only work in certain situations.
Many that work, have tradeoffs.
All recommendations are equal, but some recommendations are more equal than others.
One of the key words is Series Resonant. Some will resonate more than others.
Some will resonate in your ears.
Qs of 3 or less are considered by many to be non-resonant.
Qs of 4 or more are considered by many to be resonant.
Radio Days (Daze) keep coming back, and so does the old ARRL Radio Amateurs Handbook.
Thanks for the interresting and useful conversation. I think, Audio technology is very complex...
My thought was unsophisticated about this attachment i send. Is it profitable with a RCL in the crossover to compensate possible problems with the match to a valve amplifier. Do it give sense. I suppose its a question that's not easy to answer.
My thought was unsophisticated about this attachment i send. Is it profitable with a RCL in the crossover to compensate possible problems with the match to a valve amplifier. Do it give sense. I suppose its a question that's not easy to answer.
michaeljessen,
Your original, and re-asked question is a good one.
I wish there was a single and simple answer.
I do not know if there is, or is not.
Something to consider is which amplifier, not just which loudspeaker.
Is the amplifier single ended, or Is the amplifier push pull?
Is the amplifier damping factor low, medium, or high?
Does the amplifier use global negative feedback that is taken from the output transformer secondary?
Is there lots of global negative feedback?
Is the amplifier stable with all kinds of loads; resistive, inductive, and capacitive?
Is the loudspeaker primarily designed for solid state amplifiers, or for vacuum tube amplifiers?
Is there a specific amplifier and a specific loudspeaker that you are considering to use together?
Your original, and re-asked question is a good one.
I wish there was a single and simple answer.
I do not know if there is, or is not.
Something to consider is which amplifier, not just which loudspeaker.
Is the amplifier single ended, or Is the amplifier push pull?
Is the amplifier damping factor low, medium, or high?
Does the amplifier use global negative feedback that is taken from the output transformer secondary?
Is there lots of global negative feedback?
Is the amplifier stable with all kinds of loads; resistive, inductive, and capacitive?
Is the loudspeaker primarily designed for solid state amplifiers, or for vacuum tube amplifiers?
Is there a specific amplifier and a specific loudspeaker that you are considering to use together?
I have build the TSE ll from the George's tubelab website with application for the type 45 tube and 6K ohm 60ma output transformer. (No catode feedback)
I want to match them with (crossover upgraded) Klipsch RP600M
Then I sow this attachment with the RCL calculator and was wondering...
I think i do not know enough about this issue.
I want to match them with (crossover upgraded) Klipsch RP600M
Then I sow this attachment with the RCL calculator and was wondering...
I think i do not know enough about this issue.
RLC = Resonance
Resonance has a Q.
Resonance with a Q has attack and decay.
High impedance tube amp has low damping ability for above Resonance.
So it will ring to some degree with slow attack and slow decay, nature of the circuit...just a question of how much it will ring and how audible it will be.
In my personal experience, a RLC cure always sounded worse than the disease it was intended to fix.
Series RLC impedance correctors in the loudspeaker reminds me of a similar resonator . . .
Some full range drivers have a broad peak at perhaps 3.5kHz to 5.5kHz as an example . . .
One suggestion was to put a Parallel RLC network in series with the driver voice coil.
True, that reduces the signal to the driver at those frequencies, But now the RLC rings at those same frequencies. Just move the problem from one resonator to another resonator.
A full range driver that I mostly liked, had a whizzer cone with the expected broad resonance. I tried wool, cotton, etc. to dampen the resonance.
I very carefully cut the whizzer cone down to only 1/4 inch. Problem solved.
Generally, a small amplitude reduction of a frequency range is less objectionable than a peak in the same frequency range.
Just my experience.
Some full range drivers have a broad peak at perhaps 3.5kHz to 5.5kHz as an example . . .
One suggestion was to put a Parallel RLC network in series with the driver voice coil.
True, that reduces the signal to the driver at those frequencies, But now the RLC rings at those same frequencies. Just move the problem from one resonator to another resonator.
A full range driver that I mostly liked, had a whizzer cone with the expected broad resonance. I tried wool, cotton, etc. to dampen the resonance.
I very carefully cut the whizzer cone down to only 1/4 inch. Problem solved.
Generally, a small amplitude reduction of a frequency range is less objectionable than a peak in the same frequency range.
Just my experience.
Last edited:
Thanks for the input. It looks like that it is a good pointe to keep it simple, as possible... What about the quality of the components in the RCL ? I have tried to smooth a objectionable peek with a paralel RCL in serial with + on the Fostex103 EN (do it yourself kit from 2003)
The success was ok and "good", but there difference between Por or good components was Audible.
My eksperience was, that A peek especially at 4000 Hz can be very nasty for the human ear.
Conclusion for me may be to experiment and use my ears.
The success was ok and "good", but there difference between Por or good components was Audible.
My eksperience was, that A peek especially at 4000 Hz can be very nasty for the human ear.
Conclusion for me may be to experiment and use my ears.
1. Consider a speaker driver that has a 6dB elevated amplitude from 3.5kHz to 4.5kHz.
It is 1kHz wide, and centered at 4kHz.
Q = 4k / 1k. The Q is 4.
With a Q of 4, that is a real resonance.
An old general rule was that a Q of less than 3 was not resonant, it was well damped.
Putting a 4kHz parallel RLC in series from the amplifier to the voice coil, that does reduce the amplitude at 4 kHz . . .
There is less signal power from 3.5kHz to 4.5kHz going to the driver.
But the driver is still resonant at 4kHz, so it still "rings", but that "ringing" is at a slightly lower amplitude.
My opinion is that the parallel RLC network is an example of trying to make a silk purse from a sows ear.
In some cases diy'ers, and manufacturers will try to dampen the resonance mechanically with a lossy coating, wool, etc.
If they get rid of the ringing without loosing efficiency at other frequencies, they have succeeded.
2. Consider a speaker that has a 6 dB elevated amplitude from 2kHz to 6kHz. Again centered at 4kHz
Q = 4kHz / 4kHz. Q = 1.0
That is not resonant, it is not "ringing".
But if that elevated response is also accompanied by an increase in speaker impedance from 2kHz to 6kHz, it will adversely affect the sound of an amplifier that has low damping factor, because the amplitude of the sound will be even further increased at those frequencies, the amplifier will put out more voltage there.
Suppose the increase of impedance is from 7 Ohms to 12 Ohms.
In that case, use a parallel RLC network with Q = 1, and impedance of 5 Ohms (12 Ohms - 7 Ohms = 5 Ohms).
That will flatten out the frequency from 2kHz to 6kHz, without "ringing"
Higher damping factor amplifiers will not be affected that way.
3. I hope you can see that in some cases (1.) a series RLC will Not help;
. . . And in other cases (2.) it Will help.
It is 1kHz wide, and centered at 4kHz.
Q = 4k / 1k. The Q is 4.
With a Q of 4, that is a real resonance.
An old general rule was that a Q of less than 3 was not resonant, it was well damped.
Putting a 4kHz parallel RLC in series from the amplifier to the voice coil, that does reduce the amplitude at 4 kHz . . .
There is less signal power from 3.5kHz to 4.5kHz going to the driver.
But the driver is still resonant at 4kHz, so it still "rings", but that "ringing" is at a slightly lower amplitude.
My opinion is that the parallel RLC network is an example of trying to make a silk purse from a sows ear.
In some cases diy'ers, and manufacturers will try to dampen the resonance mechanically with a lossy coating, wool, etc.
If they get rid of the ringing without loosing efficiency at other frequencies, they have succeeded.
2. Consider a speaker that has a 6 dB elevated amplitude from 2kHz to 6kHz. Again centered at 4kHz
Q = 4kHz / 4kHz. Q = 1.0
That is not resonant, it is not "ringing".
But if that elevated response is also accompanied by an increase in speaker impedance from 2kHz to 6kHz, it will adversely affect the sound of an amplifier that has low damping factor, because the amplitude of the sound will be even further increased at those frequencies, the amplifier will put out more voltage there.
Suppose the increase of impedance is from 7 Ohms to 12 Ohms.
In that case, use a parallel RLC network with Q = 1, and impedance of 5 Ohms (12 Ohms - 7 Ohms = 5 Ohms).
That will flatten out the frequency from 2kHz to 6kHz, without "ringing"
Higher damping factor amplifiers will not be affected that way.
3. I hope you can see that in some cases (1.) a series RLC will Not help;
. . . And in other cases (2.) it Will help.
This I would agree with completely.
A speaker will sin little by omission but greatly by addition of the unwanted.
Thank you very much for the good advise. I am learning and benefit from it and hope it can be useful for other people also.
The TO described his amp having no overall global feedback-
which explains the rise of impedance he perceived.
For me the resonant tank looks like a very gross work around.
Better you cure the amp with a global feedback to achieve some resonable output impedance.
which explains the rise of impedance he perceived.
For me the resonant tank looks like a very gross work around.
Better you cure the amp with a global feedback to achieve some resonable output impedance.
There are ways to design a tube amplifier that does Not have global negative feedback, does Not have Schade negative feedback, does Not have Ultra Linear negative feedback, and does Not have output stage cathode negative feedback . . .
And, without any of those negative feedback methods, you still can get a medium damping factor.
However, that does reduce the number of options of output tube types and circuit topologies:
Pentodes and Beam Power tubes, can not be used in their native operating modes.
Triodes, Triode Wired Pentodes, and Triode Wired Beam Power tubes, Can be used with output transformers that have. primary impedance that is several times the plate impedance of those triode and "triode' wired tubes.
Disadvantage: Less output power.
Advantages: low distortion, not having to use negative feedback, simplicity, and medium damping factor.
Well, I do admit that Triode Wired pentodes and Triode Wired beam power tubes is a form of negative feedback . . . from the plate to the screen.
Think of it as "Ultra Linear" with a 100% Ultra Linear Tap; no special output transformer (No UL Tap needed) to get 100% "Ultra Linear".
If an amplifier is not Triode output, and is not Triode Wired Pentode, and is not Triode Wired Beam Power, then . . .
If the wiring is point to point, it is an easy modification to make, Triode Wire it!
If the wiring is PCB, some trace(s) can be cut, and you can Triode Wire it! Cut carefully, and you can go back to the original Non-Triode Wired circuit.
These are easy modifications . . .
I can not understand why everybody does not try Triode Wiring their Pentodes and Beam Power tubes.
Have Fun!
And, without any of those negative feedback methods, you still can get a medium damping factor.
However, that does reduce the number of options of output tube types and circuit topologies:
Pentodes and Beam Power tubes, can not be used in their native operating modes.
Triodes, Triode Wired Pentodes, and Triode Wired Beam Power tubes, Can be used with output transformers that have. primary impedance that is several times the plate impedance of those triode and "triode' wired tubes.
Disadvantage: Less output power.
Advantages: low distortion, not having to use negative feedback, simplicity, and medium damping factor.
Well, I do admit that Triode Wired pentodes and Triode Wired beam power tubes is a form of negative feedback . . . from the plate to the screen.
Think of it as "Ultra Linear" with a 100% Ultra Linear Tap; no special output transformer (No UL Tap needed) to get 100% "Ultra Linear".
If an amplifier is not Triode output, and is not Triode Wired Pentode, and is not Triode Wired Beam Power, then . . .
If the wiring is point to point, it is an easy modification to make, Triode Wire it!
If the wiring is PCB, some trace(s) can be cut, and you can Triode Wire it! Cut carefully, and you can go back to the original Non-Triode Wired circuit.
These are easy modifications . . .
I can not understand why everybody does not try Triode Wiring their Pentodes and Beam Power tubes.
Have Fun!
Last edited:
michaeljessen,
A 45 tube plate impedance is 1700 Ohms.
A 6k primary output transformer that has low insertion loss . . . combined with a 1700 Ohms rp, gives a damping factor of about 3.5.
Pretty good!
Of course you need to use an output transformer that has low primary DCR, and low secondary DCR.
Otherwise, the 2 Watts from the 45 tube will be lost in the insertion loss of the output transformer's DCRs.
The 45 is an expensive high quality tube.
Use a quality output transformer.
Two questions that I should have asked in the beginning . . .
Have you listened to your amplifier on your loudspeakers?
Do you hear any peaks, ringing, or other artifacts when you listen to the music?
(do not listen to your system; listen to the music).
Many of us sometimes get too focused on our system, and hear things that may, or may not, really be there.
A 45 tube plate impedance is 1700 Ohms.
A 6k primary output transformer that has low insertion loss . . . combined with a 1700 Ohms rp, gives a damping factor of about 3.5.
Pretty good!
Of course you need to use an output transformer that has low primary DCR, and low secondary DCR.
Otherwise, the 2 Watts from the 45 tube will be lost in the insertion loss of the output transformer's DCRs.
The 45 is an expensive high quality tube.
Use a quality output transformer.
Two questions that I should have asked in the beginning . . .
Have you listened to your amplifier on your loudspeakers?
Do you hear any peaks, ringing, or other artifacts when you listen to the music?
(do not listen to your system; listen to the music).
Many of us sometimes get too focused on our system, and hear things that may, or may not, really be there.
Last edited:
"Thank you for the music"
No I have not listened to this match yet. I am still in progress with the upgrade of the crossover to better component quality. I want to create the best possible condition for the match (Klipch RP 600M) to the amplifier. Then i sow this attachment with the RCA calculator and the recommendation to use a RCL in the crossover when the speaker is to connect to a valve amplifier. I was wondering about this and realize that i not have the knowledge to take a decision about it... I ask myself will it give sense to install it.
Then I think i have to ask a question about this issue here on the diy audio forum. I am happy i have done it.
Now I Know better than before... Thanks a lot.
My output transformers are the Swedish Lundahl LL 1620. And Yes you are right, we have to listen to the music
No I have not listened to this match yet. I am still in progress with the upgrade of the crossover to better component quality. I want to create the best possible condition for the match (Klipch RP 600M) to the amplifier. Then i sow this attachment with the RCA calculator and the recommendation to use a RCL in the crossover when the speaker is to connect to a valve amplifier. I was wondering about this and realize that i not have the knowledge to take a decision about it... I ask myself will it give sense to install it.
Then I think i have to ask a question about this issue here on the diy audio forum. I am happy i have done it.
Now I Know better than before... Thanks a lot.
My output transformers are the Swedish Lundahl LL 1620. And Yes you are right, we have to listen to the music