Tim, I assume you mean this?
If so, here is the current regulation (green=driven, red=strapped):
The GK strapped configuration shows about 26uA. Much worse than the 0.5uA when the grid is driven.
An externally hosted image should be here but it was not working when we last tested it.
If so, here is the current regulation (green=driven, red=strapped):
An externally hosted image should be here but it was not working when we last tested it.
The GK strapped configuration shows about 26uA. Much worse than the 0.5uA when the grid is driven.
runeight said:
Yes. Except, you can get rid of the 10k between stages and adjust the screen supply resistor for operating point.
Not suprising, it's running purely on plate resistance (high though it may be), rather than amplifying the error signal. Likewise, it'll also trash out pretty good when driving anything.
Oh - how about putting a CCS in place of the "current sense" resistor? I bet that would perfect it.
Thanks for satisfying my curiosity,
Tim
Good idea. Kind of an infinitely recursive design. Each CCS with another CCS inside of it. I like it.
One last thing on the pentode option. I thought I would try a 6BM8 because it is a triode/pentode package and because I have the model. The pentode has half the S of the 6JC6. I wondered if it would work. Also, you get your triode and pentode in the same bottle. Only two sockets for a stereo line stage.
Here's the diagram.
An externally hosted image should be here but it was not working when we last tested it.
To get both devices into good operating region (as I judge it) I raised the B+ to 400V. Idle current is about 4.6mA.
Here is the current draw with a .05v 1KHz input signal. Again the fourier transform:
An externally hosted image should be here but it was not working when we last tested it.
Still below 1ua and so, from the perspective of simulation vs. reality, no different from the 6JC6.
The gain is almost twice the previous configuration. Here is the 2nd harmonic for about the same output voltage:
An externally hosted image should be here but it was not working when we last tested it.
About 0.25%.
And, finally, the response curve:
An externally hosted image should be here but it was not working when we last tested it.
My conclusion is that if you really want to use a pentode as the CCS device, except for the PS change, this is a pretty good option.
Frank, I wonder if a 6GV8 would be a better choice?
However, all of this work still tells me that a triode pair will be just as good. You can get them two to a package too, while keeping a reasonable supply voltage.
Good work that you are doing.
Here is a link that you may wish to read and then try in your sim.
http://www.dddac.de/at08.htm
Here is a link that you may wish to read and then try in your sim.
http://www.dddac.de/at08.htm
Hi,
And what if I tell you that some penthodes can be used to replace a choke? ( Tim loves me already)
For, all things considered, this is what we're doing here except we try to keep current constant.
Moreover, I know of at least one tetrode that can mimmick a nice choke... and it ain't all that small a H value either.
There's a lot of clever things you can do with tubes...there really is no end to it.
Cheers,
And what if I tell you that some penthodes can be used to replace a choke? ( Tim loves me already)
For, all things considered, this is what we're doing here except we try to keep current constant.
Moreover, I know of at least one tetrode that can mimmick a nice choke... and it ain't all that small a H value either.
There's a lot of clever things you can do with tubes...there really is no end to it.
Cheers,
> Kimmel says that a triode cannot be considered a true CCS since "because it hasn't enough gain...
It Depends.
The usual trick is about a 470Ω cathode resistor, with the grid tied to the bottom of that. The CCS effective impedance is then essentially Mu*(1/Gm) which is Rp which is about the same as a simple plate resistor.
Runeight is using a 20K cathode resistor. That forces 3 extra parts to keep DC bias. But then the CCS impedance is Mu*(20K+(1/Gm)). For the 6DJ8, that comes out to about 600KΩ, a very high value indeed. You get 90% of the benefit of a "perfect" CCS.
Another point: 1/Gm is 200 to 2K for most tubes, and 200Ω happens only with "best" (most fragile) tubes at high current. But with the 20K resistor, they all work about the same: you can use a rugged 12AU7 or 12AT7 instead of the 6DJ8 or other too-hot tube. The impedance then depends mostly on Mu, not Gm or the ratio of Gm to cathode bias resistance.
When you get to Pentodes: the effective Mu is very high, but you must drive the Screen to track the Cathode. The obvious way is a capacitor from Cathode to Screen. You also need a path to supply Screen current, usually a resistor. When you do all that, the Pentode's very-high Mu is swamped by loading effects. I put far less trust in the available Pentode models than the triode models, but I think Runeight is right when he shows only modest or even no improvement from a ballasted triode. You could bootstrap the screen with yet another tube, but where are we going?
As Frank says "There's a lot of clever things you can do with tubes...there really is no end to it." Recall the words of the preacher in the Bible: "Of the making of many things there is no end... there is no new thing under the sun."
I think the simulation should have a load: say the typical grid resistance and capacitance, or the IHF standard Line Out load. If you cascode and bootstrap and CCS a tube to death, you make an ideal infinite object that can't interface with any real world.
For anybody who really wants to dig: just after WWII the MIT Radiation Labs (RADAR) released a 10-volume summary of their research. One volume is general amplifiers, and talks about such things. In general, as you reduce current while keeping supply voltage high, gain rises (but bandwidth decreases). They did try near-CCS topologies. In real life, the improvement from using another tube as a CCS is not as good as using an extra tube some other way, such as another amplifier stage or a buffer. You never see such frills on commercial gear from the heyday of tubes: they knew better.
There is no remembrance of former things... Ecclesiastes 1:11
(because runeight and I smashed them with hammers....)
It Depends.
The usual trick is about a 470Ω cathode resistor, with the grid tied to the bottom of that. The CCS effective impedance is then essentially Mu*(1/Gm) which is Rp which is about the same as a simple plate resistor.
Runeight is using a 20K cathode resistor. That forces 3 extra parts to keep DC bias. But then the CCS impedance is Mu*(20K+(1/Gm)). For the 6DJ8, that comes out to about 600KΩ, a very high value indeed. You get 90% of the benefit of a "perfect" CCS.
Another point: 1/Gm is 200 to 2K for most tubes, and 200Ω happens only with "best" (most fragile) tubes at high current. But with the 20K resistor, they all work about the same: you can use a rugged 12AU7 or 12AT7 instead of the 6DJ8 or other too-hot tube. The impedance then depends mostly on Mu, not Gm or the ratio of Gm to cathode bias resistance.
When you get to Pentodes: the effective Mu is very high, but you must drive the Screen to track the Cathode. The obvious way is a capacitor from Cathode to Screen. You also need a path to supply Screen current, usually a resistor. When you do all that, the Pentode's very-high Mu is swamped by loading effects. I put far less trust in the available Pentode models than the triode models, but I think Runeight is right when he shows only modest or even no improvement from a ballasted triode. You could bootstrap the screen with yet another tube, but where are we going?
As Frank says "There's a lot of clever things you can do with tubes...there really is no end to it." Recall the words of the preacher in the Bible: "Of the making of many things there is no end... there is no new thing under the sun."
I think the simulation should have a load: say the typical grid resistance and capacitance, or the IHF standard Line Out load. If you cascode and bootstrap and CCS a tube to death, you make an ideal infinite object that can't interface with any real world.
For anybody who really wants to dig: just after WWII the MIT Radiation Labs (RADAR) released a 10-volume summary of their research. One volume is general amplifiers, and talks about such things. In general, as you reduce current while keeping supply voltage high, gain rises (but bandwidth decreases). They did try near-CCS topologies. In real life, the improvement from using another tube as a CCS is not as good as using an extra tube some other way, such as another amplifier stage or a buffer. You never see such frills on commercial gear from the heyday of tubes: they knew better.
There is no remembrance of former things... Ecclesiastes 1:11
(because runeight and I smashed them with hammers....)
Frank, you knew I would have to ask how? So I could simulate it.
PRR, thanks for the additional illumination. As always you managed to do in a few paragraphs what took me a couple of pages and several thousand drawings.
Et tu, Brute?
. . . for the former things are passed away. . . . Behold I make all things new. . . . Revelation 21:4,5
I trust we can be forgiven.
Hi,
Sure....
Vb = 400V
V out = 335V
Va = 60V
Vg2 = 100V
Vg1 = -1.0V
Ik = 110mA
Rg2 = 60K
Rg1 = 470K
Rk = 50R
C in = 8.0µF
C out = 20µF
Cg2 = 8.0µF
Cg1 = 50µF
The tube used here is a 6216 as made by CBS-Hytron.
They claim this circuit emulates a 12H choke at 150mA..
Other than this application note, I fail to see anything extraordinary about this tube tube so I think it's safe to assume other penthodes can be made to work in a similar way.
Cheers,
Sure....
Vb = 400V
V out = 335V
Va = 60V
Vg2 = 100V
Vg1 = -1.0V
Ik = 110mA
Rg2 = 60K
Rg1 = 470K
Rk = 50R
C in = 8.0µF
C out = 20µF
Cg2 = 8.0µF
Cg1 = 50µF
The tube used here is a 6216 as made by CBS-Hytron.
They claim this circuit emulates a 12H choke at 150mA..
Other than this application note, I fail to see anything extraordinary about this tube tube so I think it's safe to assume other penthodes can be made to work in a similar way.
Cheers,
Attachments
> do in a few paragraphs what took me a couple of pages and several thousand drawings
I'm just slapping words. Broskie has a much better article. - SRPP Deconstructed - PDF
He starts with "the 1940 patent (US 2,310,342) for the circuit", moves on to Millman-Taub, and its use in video line drivers. Aside from driving some of the early computer "core" memory, that was the end of the line until its rediscovery by audio tube-heads.
"So what insights can we take away from these circuits? First that the load counts. The load to be driven defines the starting point. - If the load is a punishingly low 32 to 10kohms, then abandon any notions of constant current sources and single-ended operation. - On the other hand, if the load is 1-meg to infinity, then high-current push-pull operation make no sense and single-ended operation is closer to the actual functioning of the circuit. But as the load is so high why not move the connection.... - If the load impedance is not specified, as in the case in a line amplifier, do not use any type of SRPP circuit, as the SRPP should be designed around one specific load impedance. - In sum, the SRPP circuit has its uses, but those uses are limited. Where money and space allow, better more complicated circuits can be used to better effect than the SRPP; but where dirty and cheap are needed and where the load impedance is fixed and purely resistive, the SRPP is a good choice."
The end of that refers back to an earlier article SRPP Decoded. On page 3 is a big box-quote "Unlike the White Cathode Follower, the limitation to the SRPP is that it can only be optimized for one specific load impedance."
(FWIW, my studies say the WCF gets best performance if you DO optimize for the specific load to be driven; but the difference in component values and performance is VERY small from short to infinity. One set of values does work good for any sane load. With the SRPP, you have to change values every time you change the load {to get "best" performance}.)
The fixed-impedance restriction on an optimum SRPP is a killer. You usually have both resistance and capacitiance. Cable capacitance, grid capacitance, whatever. Therefore the load impedance changes over the audio range (including supersonics which may be intermodulated into the audible band). Now, maybe the SRPP can handle "some" change of load: obviously it does not "stop working" when the load is changed a little. But does it handle varying load even as well as a grounded cathode followed by a cathode follower? I suspect the GC-CF is generally better: the GC stage is well isolated from load variation and the CF stage output can drive most loads with ease. With the GC stage essentially unloaded, it can be working in a highly linear range, without trying to "tuck it under" another tube. The biggest advantage of SRPP then is the fact that two tubes take B+ current in series. Over a small range of loads, the SRPP may be more power-thrifty than a GC-CF design. Over a larger range of loads, either the SRPP fails to drive low-Z as well as a CF, or it is eating more power current than a GC-CF amp could be designed for.
I'm just slapping words. Broskie has a much better article. - SRPP Deconstructed - PDF
He starts with "the 1940 patent (US 2,310,342) for the circuit", moves on to Millman-Taub, and its use in video line drivers. Aside from driving some of the early computer "core" memory, that was the end of the line until its rediscovery by audio tube-heads.
- "Today, the prevailing view (a new view) is that the SRPP is an SE circuit made up of a grounded-cathode amplifier that is loaded by a cathode follower that also functions as a constant current source; (a nice trick: providing a low and an infinite impedance at once);...."
- "The alternate view (the old view) sees the SRPP as an ingenious way to do push-pull on the cheap, as the SRPP is simply a two-tube push-pull amplifier that comprises its own phase splitter in the form of a single resistor."
"So what insights can we take away from these circuits? First that the load counts. The load to be driven defines the starting point. - If the load is a punishingly low 32 to 10kohms, then abandon any notions of constant current sources and single-ended operation. - On the other hand, if the load is 1-meg to infinity, then high-current push-pull operation make no sense and single-ended operation is closer to the actual functioning of the circuit. But as the load is so high why not move the connection.... - If the load impedance is not specified, as in the case in a line amplifier, do not use any type of SRPP circuit, as the SRPP should be designed around one specific load impedance. - In sum, the SRPP circuit has its uses, but those uses are limited. Where money and space allow, better more complicated circuits can be used to better effect than the SRPP; but where dirty and cheap are needed and where the load impedance is fixed and purely resistive, the SRPP is a good choice."
The end of that refers back to an earlier article SRPP Decoded. On page 3 is a big box-quote "Unlike the White Cathode Follower, the limitation to the SRPP is that it can only be optimized for one specific load impedance."
(FWIW, my studies say the WCF gets best performance if you DO optimize for the specific load to be driven; but the difference in component values and performance is VERY small from short to infinity. One set of values does work good for any sane load. With the SRPP, you have to change values every time you change the load {to get "best" performance}.)
The fixed-impedance restriction on an optimum SRPP is a killer. You usually have both resistance and capacitiance. Cable capacitance, grid capacitance, whatever. Therefore the load impedance changes over the audio range (including supersonics which may be intermodulated into the audible band). Now, maybe the SRPP can handle "some" change of load: obviously it does not "stop working" when the load is changed a little. But does it handle varying load even as well as a grounded cathode followed by a cathode follower? I suspect the GC-CF is generally better: the GC stage is well isolated from load variation and the CF stage output can drive most loads with ease. With the GC stage essentially unloaded, it can be working in a highly linear range, without trying to "tuck it under" another tube. The biggest advantage of SRPP then is the fact that two tubes take B+ current in series. Over a small range of loads, the SRPP may be more power-thrifty than a GC-CF design. Over a larger range of loads, either the SRPP fails to drive low-Z as well as a CF, or it is eating more power current than a GC-CF amp could be designed for.
> penthodes can be used to replace a choke?
With gain, you can emulate anything.
The most basic emulated choke, using a perfect voltage follower:
The 1K resistor acts a lot like the copper resistance of a real coil. The 1Meg resistor acts a lot like any shunt resistance in a real coil (load impedance and core loss). The capacitor working against (R1*R2)^0.5 sets the "inductance".
Impedance plot:
Where it says "KV", read "KΩ" (since the driving source is 1Amp).
At DC it is 1K, at high-AC it is 1Meg. With these values, the -3dB point is about 32Hz.
You can use a cathode follower. In fact the cathode impedance can be part of R1. It can "be" R1, but then it is fairly non-linear and as much a "tube" as a "choke". The slight loss of gain in a real CF means a non-trivial shift in the computed inductance. Also a cathode follower needs a DC load, which is an additional shunt on the choke. However that DC load could be the plate of another tube, a very interesting plan. However it then becomes a form of SRPP (the kind with a grid capacitor) and Broskie's essay applies.
With gain, you can emulate anything.
The most basic emulated choke, using a perfect voltage follower:
An externally hosted image should be here but it was not working when we last tested it.
The 1K resistor acts a lot like the copper resistance of a real coil. The 1Meg resistor acts a lot like any shunt resistance in a real coil (load impedance and core loss). The capacitor working against (R1*R2)^0.5 sets the "inductance".
Impedance plot:
An externally hosted image should be here but it was not working when we last tested it.
Where it says "KV", read "KΩ" (since the driving source is 1Amp).
At DC it is 1K, at high-AC it is 1Meg. With these values, the -3dB point is about 32Hz.
You can use a cathode follower. In fact the cathode impedance can be part of R1. It can "be" R1, but then it is fairly non-linear and as much a "tube" as a "choke". The slight loss of gain in a real CF means a non-trivial shift in the computed inductance. Also a cathode follower needs a DC load, which is an additional shunt on the choke. However that DC load could be the plate of another tube, a very interesting plan. However it then becomes a form of SRPP (the kind with a grid capacitor) and Broskie's essay applies.
Hi,
Agreed on most counts.
The SRPP looks like a misnomer indeed, it's hardly a shunt regulator and it's even less PP.
Still, it does have a number of merits and when the load conditions as mentioned by PRR are met it can be be a handy solution to some problems.
As with most things electronic nothing's ever perfect, it's always a delicate balance of making the right decision for the job at hand.
One thing about the SRPP and its derivatives is that it's not well suited to low Rp, low µ tubes.
This should be obvious from John Broskies analysis or if you think for your own for that matter.
I've said it before, the SRPP and other stacked tubes deserve a well regulated PS on their roof to give of their best, if those conditions aren't met they'll just disappoint.
The major drawback of the WCF is its having no gain. As a tubed buffer it's hard to beat and add a stiffly regulated SU and all is bliss bar the silliest load.
So, Runeight if you want more work thrown your way; there you have some.
PRR,
Thanks for the liitle analysis on the penthode choke. It confirms what I suspected: any penthode with good transconductance could qualify.
This opens up a lot of perspectives; penthodes, at least some of them are dirt cheap and cost way less than an actual choke. Moreover they also only weigh a fraction of a choke and take up much less real estate.
Precisely, nothing that can't be overcome and not really a big deal if you follow me.
So, in a way we can have a penthode serving as replacement in a CCS ( thank you Gary Pimm), it mimmicks our classic anode choke, but also can we now use a penthode to replace a choke as a reactor in a PS (thank you CBS-Hytron).
And we can push those ideas further still, Dr. Evil still has a few aces up his sleeve...
Your call Runeight ,
Agreed on most counts.
The SRPP looks like a misnomer indeed, it's hardly a shunt regulator and it's even less PP.
Still, it does have a number of merits and when the load conditions as mentioned by PRR are met it can be be a handy solution to some problems.
As with most things electronic nothing's ever perfect, it's always a delicate balance of making the right decision for the job at hand.
One thing about the SRPP and its derivatives is that it's not well suited to low Rp, low µ tubes.
This should be obvious from John Broskies analysis or if you think for your own for that matter.
I've said it before, the SRPP and other stacked tubes deserve a well regulated PS on their roof to give of their best, if those conditions aren't met they'll just disappoint.
The major drawback of the WCF is its having no gain. As a tubed buffer it's hard to beat and add a stiffly regulated SU and all is bliss bar the silliest load.
So, Runeight if you want more work thrown your way; there you have some.
PRR,
Thanks for the liitle analysis on the penthode choke. It confirms what I suspected: any penthode with good transconductance could qualify.
This opens up a lot of perspectives; penthodes, at least some of them are dirt cheap and cost way less than an actual choke. Moreover they also only weigh a fraction of a choke and take up much less real estate.
Precisely, nothing that can't be overcome and not really a big deal if you follow me.
So, in a way we can have a penthode serving as replacement in a CCS ( thank you Gary Pimm), it mimmicks our classic anode choke, but also can we now use a penthode to replace a choke as a reactor in a PS (thank you CBS-Hytron).
And we can push those ideas further still, Dr. Evil still has a few aces up his sleeve...
Your call Runeight ,
Thanks PRR for reminding me of Broskie's article. I remember reading it quite a while ago. I should go back and read it again.
I agree with you on this one. If I were planning a line stage to drive an arbitrary set of loads, I would use a constant current draw GC-CF pair. If the operating region and gain are reasonably chosen, I believe the audio quality will be very high.
But, I'm already reeling from rolling!! Yes, there could be some more work to do. Where would you suggest we go? I am becoming more and more convinced that CCS loaded devices are not all that useful or "sonically" different. But, I am willing to listen and try more stuff.
Now you're making me really nervous.
I agree with you on this one. If I were planning a line stage to drive an arbitrary set of loads, I would use a constant current draw GC-CF pair. If the operating region and gain are reasonably chosen, I believe the audio quality will be very high.
But, I'm already reeling from rolling!! Yes, there could be some more work to do. Where would you suggest we go?
I am becoming more and more convinced that CCS loaded devices are not all that useful or "sonically" different. But, I am willing to listen and try more stuff.Now you're making me really nervous.
Hi,
And what if I tell you that most of this exercise is just not necessary for all things audio?
Don't get me wrong, I like high PSR and all that when it comes to audio circuits but it's oh so circuit dependent you see.
When it comes to linear loads, it's hard to beat a good old resistor...
Anything else is horses for courses IMO...
I should catch some sleep...3.16 PM
The seemingly contradiction isn't as contradictary as it seems at first look...
Think about it...
Don't be...back to the future!!!
Cheers,
And what if I tell you that most of this exercise is just not necessary for all things audio?
Don't get me wrong, I like high PSR and all that when it comes to audio circuits but it's oh so circuit dependent you see.
When it comes to linear loads, it's hard to beat a good old resistor...
Anything else is horses for courses IMO...
I should catch some sleep...3.16 PM
The seemingly contradiction isn't as contradictary as it seems at first look...
Think about it...
Don't be...back to the future!!!
Cheers,
Say, I went back and re-read Broskie's SRPP articles. And now I'm asking myself, why did I even crank up this thread? I should have just pointed folks to the TCJ.
Well, not exactly, it was worth the time to look at it from a simulation perspective. And I have noticed that since becoming part of Beige Bag SW, Mr. Broskie has taken a real liking to simulations because they really do work when you know how to use them.
I have had some very good correspondence with him and he has been quite generous in giving me some of his time and expertise.
I should have just pointed folks to the TCJ.Well, not exactly, it was worth the time to look at it from a simulation perspective. And I have noticed that since becoming part of Beige Bag SW, Mr. Broskie has taken a real liking to simulations because they really do work when you know how to use them.
I have had some very good correspondence with him and he has been quite generous in giving me some of his time and expertise.
Another opinion....
Runeight,
I think you have made a very interesting analysis, and I thank you very much for it.
However, I'd like to submit two variants to your analysis.
1) Un-bypassed tube diode in the cathode circuit
2) High impedance plate choke loading.
Specifically, I'd like to ask how much the presence of a high impedance plate choke loading and/or a tube diode in the cathode may improve the performance of your basic grounded cathode amplifier.
My questions spring from a recent schematic proposed by Dennys Boyle (http://home.att.net/~chimeraone/axiom300bschematic.html). In this project he utilizes a tube diode in the cathode (eliminating both resistor and capacitor) and a high impedance plate choke loading, obaining amazing results ( http://home.att.net/~chimeraone/axiom300btestdata.html)!
I'd be very grateful if you could furnish your considerations and/or analysis about this variation.
Runeight,
I think you have made a very interesting analysis, and I thank you very much for it.
However, I'd like to submit two variants to your analysis.
1) Un-bypassed tube diode in the cathode circuit
2) High impedance plate choke loading.
Specifically, I'd like to ask how much the presence of a high impedance plate choke loading and/or a tube diode in the cathode may improve the performance of your basic grounded cathode amplifier.
My questions spring from a recent schematic proposed by Dennys Boyle (http://home.att.net/~chimeraone/axiom300bschematic.html). In this project he utilizes a tube diode in the cathode (eliminating both resistor and capacitor) and a high impedance plate choke loading, obaining amazing results ( http://home.att.net/~chimeraone/axiom300btestdata.html)!
I'd be very grateful if you could furnish your considerations and/or analysis about this variation.
> I'd like to submit two variants to your analysis.
Looks like food for a whole new thread.
I will comment that the "a well regarded traditional design 300B amp" is clearly flawed. A 300B should not be making 1%THD at 1 watt, 2%THD at 2 watts. This is driver-stage distortion. That amp needs to either justify its THD as euphonious, or clean up its act.
Looks like food for a whole new thread.
I will comment that the "a well regarded traditional design 300B amp" is clearly flawed. A 300B should not be making 1%THD at 1 watt, 2%THD at 2 watts. This is driver-stage distortion. That amp needs to either justify its THD as euphonious, or clean up its act.
Antonio, here is a preliminary sim. I tried the drive stage. I don't have a 6y6 or a 6gc3, so I used a 6kg6 and a 6d22s damper diode.
This is the circuit:
//home.swbell.net/acavalli/images/ccs/Choke_Loaded_Diode_Biased.gif
Bias voltages a little different from the axioim schematic and the idle current is around 73mA.
I put a 5Vp 1KHz signal into the pentode. Here is the FFT of the output at the plate:
//home.swbell.net/acavalli/images/ccs/Choke_Loaded_Diode_Biased_Distortion.gif
The peak output is about 19V so the distorion from 2nd harmonic at 160mV is about 0.84%
Here it is with 10Vp at the input:
//home.swbell.net/acavalli/images/ccs/Choke_Loaded_Diode_Biased_Distortion_2.gif
About 39V output, 650mV at 2KHz, or about 1.7%.
Doesn't seem particularly good to me. But, you guys have to tell me.
This is the circuit:
//home.swbell.net/acavalli/images/ccs/Choke_Loaded_Diode_Biased.gif
Bias voltages a little different from the axioim schematic and the idle current is around 73mA.
I put a 5Vp 1KHz signal into the pentode. Here is the FFT of the output at the plate:
//home.swbell.net/acavalli/images/ccs/Choke_Loaded_Diode_Biased_Distortion.gif
The peak output is about 19V so the distorion from 2nd harmonic at 160mV is about 0.84%
Here it is with 10Vp at the input:
//home.swbell.net/acavalli/images/ccs/Choke_Loaded_Diode_Biased_Distortion_2.gif
About 39V output, 650mV at 2KHz, or about 1.7%.
Doesn't seem particularly good to me. But, you guys have to tell me.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.