I don't know. Can you explain more? What could dampen a positive current from the speaker.
What I can say, is that my speakers (Pi 4) didn't sound really good until I tried them with my 6S41S cathode followers. I tried them with various single-ended amplifiers, a couple of solid-state amplifiers and a pair of Quad IIs.
Thanks 🙂
In winter i will heat my room with the soldering iron and build this amp.
Tubes from Russia take a lot of time for traveling, so they will arrive in autumn 😊
Oh, it is by no means proof it will sound bad;-)
It just shows that the theoretically non ideal high impedance drive of speakers can in your case produce better subjective sound, even when that output impedance isn`t linear along the phase of the output signal.
Very nice! I have two questions: Do you have triode curves for the PL519? And how did you decide your operating point (Vgk, Vak, ia)?
I am curious about what happened with GU50's as I am building a SE test amp with them however it is not a cathode follower.
Thank you
What?
Cathode follower has lower output impedance than a traditional output stage, meaning the whole amplifier has a lower output impedance, which therefore improves speaker damping.
OK. Single Ended Amplifier output impedance (Consider those words as they relate to Damping Factor).
Cathode Follower output impedance.
Let us first talk about a single ended Plate output stage.
(More readers understand a plate output stage than they do a cathode output stage).
Please follow along all the way to the end of the post, I will get you to a satisfactory conclusion.
The numbers I use here are simply to illustrate the concept of output impedance.
The actual numbers may be different, but you get the idea.
1. Start with a 300B Plate that drives an output transformer with a 3500 Ohm primary.
For ease of calculation, this ideal 3500 Ohm transformer has 0 Ohms DCR of Both primary and Secondary, so essentially has No insertion loss at 1kHz.
At 60mA quiescent plate current, the 300B rp = 700 Ohms
At 120mA plate current, the 300B rp = 600 Ohms
At 10mA plate current, the 300B rp = 1400 Ohms
Damping Factor is Not a static number, Not a fixed number.
With 3500 Ohms / rp = Damping Factor.
We have 3 conditions:
3500 Ohms / 700 Ohms = 5
3500 Ohms / 600 Ohms = 5.83
3500 Ohms / 1400 Ohma = 2.5
8 Ohm Tap Output impedance (8 Ohms / Damping Factor)
8 / 5 = 1.6 Ohms
8 / 5.83 = 1.37 Ohms
8 / 2.5 = 3.2 Ohms
The instantaneous Damping Factor varies from a High of 5.83, to a Low of 2.5.
And the instantaneous output impedance varies from a High of 3.2 Ohms, to a Low of 1.37 Ohms
Well, that looks like a cause of a Non-Symmetrical Variable Gain, which results in 2nd Harmonic Distortion.
Are you with me so far?
Any Woofer Cone and Voice Coil that is in motion is both a motor, and a generator.
All the back EMF (voltage) and Back Current is what the Woofer reflects back to the 8 Ohm Secondary winding of the output transformer.
The speaker "sees" an impedance that Varies.
Single Ended amplifiers have a Non-Symmetrical damping factor.
2. Now, lets talk about the Cathode Follower.
Transconductance is Gm, the value is usually expressed in microMhos.
1 / Gm = Cathode impedance, Ohms
Just for ease of calculation, consider that u (mu) is relatively constant (it usually is)
At 60mA quiescent cathode current, the transconductance is 5500 microMhos.
At 120mA cathode current, the transconductance is perhaps 6417 microMhos.
At 10mA cathode current, the transconductance is perhaps 2750 microMhos.
Cathode Impedance:
1 / 5500 microMhos = 182 Ohms
1 / 6417 microMhos = 156 Ohms
1 / 2750 microMhos = 364 Ohms
As you can see, yes, the speaker "sees" a varying cathode follower output impedance . . .
Just like the varying plate output impedance.
3. A Class A Push Pull output stage has a Symmetrical Instantaneous Damping Factor. Wow!
That cancels the 2nd Harmonic nature that Single Ended Amplifiers have.
4. Note:
I built and loved many single ended amplifiers, and I loved their "sound".
I built and loved many push pull amplifiers, and I loved their "sound".
5. For all of you who read all the way to the end . . .
Thanks!
Congratulations!
Have Fun Designing, Building, and Listening!
Cathode Follower output impedance.
Let us first talk about a single ended Plate output stage.
(More readers understand a plate output stage than they do a cathode output stage).
Please follow along all the way to the end of the post, I will get you to a satisfactory conclusion.
The numbers I use here are simply to illustrate the concept of output impedance.
The actual numbers may be different, but you get the idea.
1. Start with a 300B Plate that drives an output transformer with a 3500 Ohm primary.
For ease of calculation, this ideal 3500 Ohm transformer has 0 Ohms DCR of Both primary and Secondary, so essentially has No insertion loss at 1kHz.
At 60mA quiescent plate current, the 300B rp = 700 Ohms
At 120mA plate current, the 300B rp = 600 Ohms
At 10mA plate current, the 300B rp = 1400 Ohms
Damping Factor is Not a static number, Not a fixed number.
With 3500 Ohms / rp = Damping Factor.
We have 3 conditions:
3500 Ohms / 700 Ohms = 5
3500 Ohms / 600 Ohms = 5.83
3500 Ohms / 1400 Ohma = 2.5
8 Ohm Tap Output impedance (8 Ohms / Damping Factor)
8 / 5 = 1.6 Ohms
8 / 5.83 = 1.37 Ohms
8 / 2.5 = 3.2 Ohms
The instantaneous Damping Factor varies from a High of 5.83, to a Low of 2.5.
And the instantaneous output impedance varies from a High of 3.2 Ohms, to a Low of 1.37 Ohms
Well, that looks like a cause of a Non-Symmetrical Variable Gain, which results in 2nd Harmonic Distortion.
Are you with me so far?
Any Woofer Cone and Voice Coil that is in motion is both a motor, and a generator.
All the back EMF (voltage) and Back Current is what the Woofer reflects back to the 8 Ohm Secondary winding of the output transformer.
The speaker "sees" an impedance that Varies.
Single Ended amplifiers have a Non-Symmetrical damping factor.
2. Now, lets talk about the Cathode Follower.
Transconductance is Gm, the value is usually expressed in microMhos.
1 / Gm = Cathode impedance, Ohms
Just for ease of calculation, consider that u (mu) is relatively constant (it usually is)
At 60mA quiescent cathode current, the transconductance is 5500 microMhos.
At 120mA cathode current, the transconductance is perhaps 6417 microMhos.
At 10mA cathode current, the transconductance is perhaps 2750 microMhos.
Cathode Impedance:
1 / 5500 microMhos = 182 Ohms
1 / 6417 microMhos = 156 Ohms
1 / 2750 microMhos = 364 Ohms
As you can see, yes, the speaker "sees" a varying cathode follower output impedance . . .
Just like the varying plate output impedance.
3. A Class A Push Pull output stage has a Symmetrical Instantaneous Damping Factor. Wow!
That cancels the 2nd Harmonic nature that Single Ended Amplifiers have.
4. Note:
I built and loved many single ended amplifiers, and I loved their "sound".
I built and loved many push pull amplifiers, and I loved their "sound".
5. For all of you who read all the way to the end . . .
Thanks!
Congratulations!
Have Fun Designing, Building, and Listening!
For how long do you propose before use? Did you mean that baking without anode voltage means a heated cathode? I think, the hot cathode gives a getter function, can eliminate the remain gasses.
No, I put them in a toaster oven at 300F for an hour. I slowly ramp up the temperature and let them cool slowly. I only do this for tubes with dummet leads to pins, not tubes with phenolic bases (Octal).
article by Morgan Jones here: Baking Tubes
article by Morgan Jones here: Baking Tubes
Last edited:
Tubes that are baked:
Does it increase "Reliability"?
Does it improve the "Sound"?
If the Getter is still intact, the thing to do is get an RF heater, and heat up the elements, get them to outgas, and hope the getter collects that.
Then, operate the tube at its maximum filament voltage, and maximum plate dissipation.
Caution: many multi-grid tubes with maximum dissipation ratings for Screen plus Plate . . .
Often the total of the combined Screen and Plate maximum dissipation is LESS than the simple sum of the two maximums.
Does it increase "Reliability"?
Does it improve the "Sound"?
If the Getter is still intact, the thing to do is get an RF heater, and heat up the elements, get them to outgas, and hope the getter collects that.
Then, operate the tube at its maximum filament voltage, and maximum plate dissipation.
Caution: many multi-grid tubes with maximum dissipation ratings for Screen plus Plate . . .
Often the total of the combined Screen and Plate maximum dissipation is LESS than the simple sum of the two maximums.
Last edited:
6A3sUMMER, In theory it keeps them from cathode stripping due to poor vacuum by reactivating the Getter.
Yes, it vastly improves the sound over that of a dead tube that experienced cathode stripping. 😆
Yes a induction heater would be nice. However a toaster oven is readily available and easy to operate.
Yes, it vastly improves the sound over that of a dead tube that experienced cathode stripping. 😆
Yes a induction heater would be nice. However a toaster oven is readily available and easy to operate.
TheGimp,
Tektronix built their own CRTs for many products. RF Heaters were a must.
The special cathodes that could do a strong beam allowed the CRTs to be bright enough for the very high light attenuation of a color shutter.
And, the small bright CRTs were used for fighter jet Heads Up Displays.
I once worked right next to the RF Heaters for the $85,000 spectrum analyzer with the color shutters.
Gee, is that what happened to my health? Just kidding.
Tektronix built their own CRTs for many products. RF Heaters were a must.
The special cathodes that could do a strong beam allowed the CRTs to be bright enough for the very high light attenuation of a color shutter.
And, the small bright CRTs were used for fighter jet Heads Up Displays.
I once worked right next to the RF Heaters for the $85,000 spectrum analyzer with the color shutters.
Gee, is that what happened to my health? Just kidding.
I have attached a 1947 article describing a push-pull 6L6 cathode follower power amp that used a power supply transformer as the output transformer. From 1940 to 1950, electronic supplies were hard to find in Australia - and in other countries too! Local magazines published amp designs that used what was available - this article is one example.
I think the cathode resistor is not bypassed, but it also does not produce current series negative feedback, it and the transformer primary together become the load of the tube.