Read somewhere recently that grounded grid (cathode drive) has a rather high Zout. Is this effect greater than the increase in zout due to unbypassed cathode resistor (which is of course necessary for GG)? Can anyone point me to a good write up on the derivation of impedances in the GG amplifier stage?
No. It's the same, though it can be influenced by the source impedance. Accounts for why cascoded triodes (GK stage DC coupled to GG stage) have such high effective plate resistance. Typically makes broadband response a good deal harder to come by when using VTs than it does with transistors.
See: Vacuum Tube Amplifier Circuits and Equations
Last edited:
ABSOLUTELY! An untuned I/P connected to the antenna. That's exactly how the 6С45П (6s45p) and 5842 were intended to be used. Examine the data sheets.
An interesting "spud" FM receiver can be found here. Isolating the antenna from the oscillating detector is a least as important as the gain. An interesting thing about the design is the triode working as a GG RF amp is also working as a reflex common cathode audio amp. Slick!
I grew up in ham radio "back in the day" -- most of the high power amps I used were grounded grid -- for SSB and CW.
Yep. I *love* how the tube radios and early transistor radios used one device for multiple circuit functions. That's real electrical engineering.
~Tom
If you guys don't mind going a little further with me on this one... 
The first most obvious way to drive the cathode was (at least to me) to use a stout cathode follower coupled by a large film capacitor such as are used in speaker crossovers. The value needed would seem to be about the same as one would use for a cathode bypass.
In order to avoid the cap it occurred to me that one could direct couple a CF as long as one took into account the bias current of the CF in the selection of the shared cathode resistor. However in the particular project I have in mind I would want to be able to drive the GG stage both solidly into cut off and well into grid current. It seems that direct coupling would not allow that kind of swing. Possibly if one added a negative rail only for the CF driver one could make a driver that could swing well below ground...
In searching around for alternatives I came across Crowhurst's short write up on the GG stage and it seems that he favors transformer coupling.
Input to the Grounded-Grid Amplifier
As usual I find it hard to follow his articles (not sure why) but this would seem to allow for the kind of swing I envision and removes the need for the cap (at the cost of an expensive tranny). The bias is set via. the DCR of the secondary and seems a nice elegant solution.
Toward the end he claims an improvement in performance (he doesn't say in what respect the performance is increased) by using a series resistor for part of the bias and bypassing it with a cap. At first I though that the cap would short out some of the input signal but I realized that the actual signal is all "above" the cap/resistor combination. But it still looks like a tank circuit to me. Is there a danger of ending up with a tuned circuit here if the R and C are not chosen carefully? Or is the value of C so high that it is incapable of leading to instability?
OK, so that was a lot further in reality but I appreciate your patient tutelage.
The first most obvious way to drive the cathode was (at least to me) to use a stout cathode follower coupled by a large film capacitor such as are used in speaker crossovers. The value needed would seem to be about the same as one would use for a cathode bypass.
In order to avoid the cap it occurred to me that one could direct couple a CF as long as one took into account the bias current of the CF in the selection of the shared cathode resistor. However in the particular project I have in mind I would want to be able to drive the GG stage both solidly into cut off and well into grid current. It seems that direct coupling would not allow that kind of swing. Possibly if one added a negative rail only for the CF driver one could make a driver that could swing well below ground...
In searching around for alternatives I came across Crowhurst's short write up on the GG stage and it seems that he favors transformer coupling.
Input to the Grounded-Grid Amplifier
As usual I find it hard to follow his articles (not sure why) but this would seem to allow for the kind of swing I envision and removes the need for the cap (at the cost of an expensive tranny). The bias is set via. the DCR of the secondary and seems a nice elegant solution.
Toward the end he claims an improvement in performance (he doesn't say in what respect the performance is increased) by using a series resistor for part of the bias and bypassing it with a cap. At first I though that the cap would short out some of the input signal but I realized that the actual signal is all "above" the cap/resistor combination. But it still looks like a tank circuit to me. Is there a danger of ending up with a tuned circuit here if the R and C are not chosen carefully? Or is the value of C so high that it is incapable of leading to instability?
OK, so that was a lot further in reality but I appreciate your patient tutelage.
I normally find Crowhurst easy to read, but this time he is confusing. He is not actually adding a resistance to the cathode circuit, he is just making it more obvious. At first he has the resistance hidden inside the transformer secondary. This reduces gain as it can't be bypassed. So he comes up with the brilliant idea of winding the transformer with ordinary wire (instead of resistance wire) and putting a resistor in, which he then bypasses. I am baffled as to why he makes such heavy weather of what should be a fairly simple circuit.
Little risk of a tank circuit formed by the transformer leakage inductance and the bypass C because the resistor and cathode impedance will damp it.
Little risk of a tank circuit formed by the transformer leakage inductance and the bypass C because the resistor and cathode impedance will damp it.
A great way to drive a cathode is using the common cathode configuration. Two identical sections, with cathodes tied together. Left (input) grid using a normal grid leak resistor; right side grid grounded. Two identical plate load resistors, but the left (cathode follower side) has a big cap from plate to ground so no gain is developed there. The cathodes tie to a current source. That current source is connected to a negative voltage supply.
Stuart
Stuart
DF96,
You're probably right about nomenclature. I've picked up the common cathode term from Broskie, who insists this is the correct term. It was redefinition of that term in the SS era that has become the norm and redefined the meaning of the term.
Kind of like watching an episode of "The Flintstones". The phrase in the theme "You'll have a gay old time" has a very different meaning now.
Stuart
"Common X" has always meant that X is the electrode common to the input and output circuits for a three-electrode device. As far as I know this applies equally to SS and valves. I am not aware of any other norm. If Broskie is trying to redefine accepted terms then he is likely to create confusion.
As far back as here
Common Cathode Line Stage (Page 1): June's Circuit of the Month
At some point he makes reference to the misuse of the term by others. FWIW
Common Cathode Line Stage (Page 1): June's Circuit of the Month
At some point he makes reference to the misuse of the term by others. FWIW
I could not find where he says this. Maybe in another article?
I was surprised by his opening sentence. He claims that this circuit has slightly higher gain than an equivalent grounded cathode circuit. I'm not sure what he means by equivalent, as I would expect about half the gain rather than slightly more gain.
I was surprised by his opening sentence. He claims that this circuit has slightly higher gain than an equivalent grounded cathode circuit. I'm not sure what he means by equivalent, as I would expect about half the gain rather than slightly more gain.
Yeah,
I can't remember which of his articles I read that in, but I'm sure it's in his pile somewhere. If you don't ground the first plate, then yes, gain is about 50% of mu. But forcing a high gm tube into CF operation does give (as I'm sure you're aware) gain of >0.9. Since the 2nd tube is running grounded grid, gain is mu + 1. I don't see this circuit having greater gain than simple grounded cathode in any significant way.
Running the numbers in TubeCad (where Broskie is at least consistant and calls the circuit CK), using a 6DJ8 as an example, 10mA and 10k plate loads, it predicts less gain from the CK circuit than GK and a CF gain of 0.96.
mu/mu+1 = 33/34 = 0.97, so the grounded grid section lacks enough gain to make up for the loss in the CF side.
Hate to say it, but he may be wrong on that, unless I'm messing up.
Stuart
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.