I will make the bold assumption that we all have Morgan Jones Valve Amplifiers 4th Edition.🙂 . OK, lets go to page 101, figure 2.27. Any thoughts on converting this into a differential circuit like the Hedge driver on pg.505 ?
Is it as straight forward as it looks or am I missing something? Also, would it be OK if all the triodes where 6SN7's or should the cascode be a 6922? I need a low distortion driver with low output impedance to drive a high capacitance.
Thanks for the help.
Is it as straight forward as it looks or am I missing something? Also, would it be OK if all the triodes where 6SN7's or should the cascode be a 6922? I need a low distortion driver with low output impedance to drive a high capacitance.
Thanks for the help.
A cascode will not have a low output impedance unless you add a couple of cathode or source followers to the output.
6SN7 is not a great choice for a cascode, at least not for the bottom- it will be too heavily loaded by the top tube and sacrifice some linearity and gain. ECC88-oids are much better in that respect.
6SN7 is not a great choice for a cascode, at least not for the bottom- it will be too heavily loaded by the top tube and sacrifice some linearity and gain. ECC88-oids are much better in that respect.
Cacodes are chosen when you want high gain but lower noise than a pentode, high output impedance, or low input capacitance (or you are constructing a VHF receiver). Generally for small signals. Not a driver circuit.
Thanks Sy, the circuit in question is a cascode direct coupled to a cathode follower with it tied to the top tube to provide negative feedback to each other. I guess we don't have our MJ4e to refer to 🙂. Part 2 of your reply is about what I figured. Thank you
Yeah, I need to get a second copy for my office. 😀
I think that if you look at Morgan's driver circuits, you'll notice a distinct lack of cascodes- there's a reason for that. Dave touched on most of those points; a conventional voltage amp will work better in that application (distortion and swing).
I think that if you look at Morgan's driver circuits, you'll notice a distinct lack of cascodes- there's a reason for that. Dave touched on most of those points; a conventional voltage amp will work better in that application (distortion and swing).
Hedge, in the June 1956 Wireless World, described using a cascode as a differential amp driver for a pair of 1625 tubes. MJ shows an uncredited design, of a cascode feeding a 6SN7 cathode follower with feedback from the top tube to the CF to provide a low output impedance for driving whatever. Later in his book, He, MJ, describes using the Hedge CF as a driver for big power tubes stating that it meets the need (with the CF output) of low distortion and high voltage swing into high capacitance.
I am confident a cascode is capable of being a fine gain stage in a power amp if you use the right tubes and have enough current. My reason for starting this thread was related to wondering if converting the circuit on pg. 101 to something like the one on pg.505 was as simple as it seems. Or am I missing something.
I think this circuit is not used much mostly because you need to bias the heaters to avoid noise and failure.
I want to use this circuit because I can get more gain than one stage of triodes and less phase shift with fewer capacitors than cascaded triodes, BTW, I don't want to use a pentode and a CF. (Pentode phobia🙂 )
I am confident a cascode is capable of being a fine gain stage in a power amp if you use the right tubes and have enough current. My reason for starting this thread was related to wondering if converting the circuit on pg. 101 to something like the one on pg.505 was as simple as it seems. Or am I missing something.
I think this circuit is not used much mostly because you need to bias the heaters to avoid noise and failure.
I want to use this circuit because I can get more gain than one stage of triodes and less phase shift with fewer capacitors than cascaded triodes, BTW, I don't want to use a pentode and a CF. (Pentode phobia🙂 )
Bear in mind that a cascode means that the lower valve gets no anode feedback so you don't get 'triode' linearity. That is why I said it is for small signals. A differential cascode will cancel even-order distortion but you still have odd-order.
You have to be careful with some 1950s designs, as back then not everyone realised just how important it was to reduce higher order distortions. They really did believe in THD.
A cascode could, under certain circumstances, make a high gain input stage but not a driver even with a CF on the output.
You have to be careful with some 1950s designs, as back then not everyone realised just how important it was to reduce higher order distortions. They really did believe in THD.
A cascode could, under certain circumstances, make a high gain input stage but not a driver even with a CF on the output.
I wish you would explain why. MJ mentions the low anode load on the bottom tube causing a loss of linearity but feels it isn't much of a problem due to the low gain and small signal swing on it. I see nothing anyplace that describes the cascode as only suitable for small signal use, in fact I see the opposite.
I did:TUBESMAN said:
Triodes are only as linear as they are because in normal operation they get feedback from the anode voltage. A cascode stops this. It is the price you pay for low input capacitance.DF96 said:
You can use a stage with poor linearity, but only for small signals. By definition, a driver stage handles large signals so distortion is worse.
Worse than a pentode? At the same gain and output level? As I said, I see many examples of a cascode being used as a driver, MJ has nothing bad to say about it as a driver when it is used with a CF to lower the output impedance, in fact, he recommends it for driving a high capacitance load with a large voltage swing. What you say is true about the cascode, the question is how big a deal it really is. And I suspect the harmonic structure is more benign than a pentode.
Hey, it sort of has become moot, the circuit was patented in 1987 so I can't use it in my design😡
Possibly. A pentode is designed in the knowledge that no linearising anode feedback will be available, so they do what they can to get a fairly linear grid characteristic. A triode designer will make different assumptions; he will try to get the grid and anode curves to match so that he gets constant mu, while not worrying too much about their individual shapes.
I don't think I have ever seen a cascode used as a driver. SRPP or mu follower is popular in China, but that works in a completely different way.
Member
Joined 2009
Paid Member
Can I view a Cascode as a pentode, the top tube adds the 'plate' so that the plate of the bottom tube can be the 'screen grid' ?
Well, they are used as drivers. Since I guess you do not have the circuit I'm referring to, I've (hopefully) attached the patent.
Allen Wright and Joe Curcio have both commercialized diff cascode drivers. I spent a few years twiddling with them and finally got over my fascination.
If you read the patent carefully you will see that he admits the cascode has problems. The patent makes it less 'cascode-like' and so reduces both its advantages and disadvantages. My question is: why choose an inappropriate circuit and then reduce its problems? Why not choose a better circuit?
I would be surprised if nobody has thought of that idea before. The fact that a patent has been granted does not mean that the patent should have been granted.
I would be surprised if nobody has thought of that idea before. The fact that a patent has been granted does not mean that the patent should have been granted.
Well if no one tried to patent it before then he deserves it. I still would like to see some numbers rather than just opinions. As I read it, he keeps the main advantages of the cascode and reduces the disadvantages of it. But since it is a current patent I cannot use this circuit anyway. And what is a "better circuit" in your opinion?
The patent is long expired so you're free to use it.
Bill Johnson loved complex circuits and never hesitated to use two tubes where one would do. Take the performance claims in the patent spec with large grains of salt.
Bill Johnson loved complex circuits and never hesitated to use two tubes where one would do. Take the performance claims in the patent spec with large grains of salt.
- Status
- Not open for further replies.