Hello,
I'm still quite new to tubes, and circuits in general to be fair. I've built a KT88 Williamson kit (Ella) and am very happy with it, but after reading through Morgan Jones's book a few times and reading lots on forums like this, I'm hungry to make improvements and start from scratch! I've seen that SRPP stages are very very commonplace in PP KT88 circuits (what I want to stick to for now) and wondered if a beta-follower would be an improvement in this situation. In Morgan Jones book he suggests that beta-followers carry the best traits from mu-followers and srpp stages in general, but then never mentions it again. Is this because it's just more complicated and he wanted to keep his designs for the book simpler? Or is it just because it's no good in this situation. Also there's very little information on beta-followers on the internet, so I'm guessing it's not widely used. Any suggestions from all you helpful experts out there?
Thanks for your help,
Steve
I'm still quite new to tubes, and circuits in general to be fair. I've built a KT88 Williamson kit (Ella) and am very happy with it, but after reading through Morgan Jones's book a few times and reading lots on forums like this, I'm hungry to make improvements and start from scratch! I've seen that SRPP stages are very very commonplace in PP KT88 circuits (what I want to stick to for now) and wondered if a beta-follower would be an improvement in this situation. In Morgan Jones book he suggests that beta-followers carry the best traits from mu-followers and srpp stages in general, but then never mentions it again. Is this because it's just more complicated and he wanted to keep his designs for the book simpler? Or is it just because it's no good in this situation. Also there's very little information on beta-followers on the internet, so I'm guessing it's not widely used. Any suggestions from all you helpful experts out there?
Thanks for your help,
Steve
Aussie DIYers have documented SUCCESS choke loading a SRPP. That idea follows the same general trend as a beta follower, but is simpler to implement.
A plate loading choke whose inductive reactance at 20 Hz. is at least 3X Rp is used in place of the upper triode's cathode resistor. A variable resistor is used in series with the choke and adjusted to make the total resistance of the LC combination = to the lower triode's cathode resistor.
More gain and improved linearity are observed when a SRPP is choke loaded.
A plate loading choke whose inductive reactance at 20 Hz. is at least 3X Rp is used in place of the upper triode's cathode resistor. A variable resistor is used in series with the choke and adjusted to make the total resistance of the LC combination = to the lower triode's cathode resistor.
More gain and improved linearity are observed when a SRPP is choke loaded.
Hi,
Thanks, sounds like a nice simple idea. I take it you mean the var resistor is used to match the dc resistance of LR combination to the bottom triodes cathode resistor? Or am I getting my letters mixed up or just misunderstanding...
Cheers,
Steve
Thanks, sounds like a nice simple idea. I take it you mean the var resistor is used to match the dc resistance of LR combination to the bottom triodes cathode resistor? Or am I getting my letters mixed up or just misunderstanding...
Cheers,
Steve
Cheers Frank,
As it happened that was the only implementation I found earlier too, hehe. Apart from a few theoretical articals on tubecad. Do you think it's a worthwhile thing to try in your opinion then?
Thanks again,
Steve
As it happened that was the only implementation I found earlier too, hehe. Apart from a few theoretical articals on tubecad. Do you think it's a worthwhile thing to try in your opinion then?
Thanks again,
Steve
Hi,
Well....IMO it mostly depends on what you need it for.
Like yourself, I haven't tried it either but it seems attractive enough to give it a go.
Cheers,😉
Well....IMO it mostly depends on what you need it for.
Like yourself, I haven't tried it either but it seems attractive enough to give it a go.
Cheers,😉
Correct. Sorry about the typo; my bad.
BTW, this is 1 time that the inductance of an "ordinary" wire wound pot. is beneficial.
I was thinking a pair of 6j5 or a 6sn7 in beta-follower driving a KT88...not sure what before that though, if anything (I need enough gain not to need an active pre with a cdp as source) all run off a phase splitting transformer and doubled up to make a PP amp. That's my thoughts at the moment
Thanks,
Steve
Thanks,
Steve
Ah cheers. Wouldn't the inductance change depending on where the wipers were in the pot? or would it be too negligable to worry about. Ahh just scanned down the page and saw it says at least 3*Rp so that makes sense now.
Thanks again
Steve,
I have not tried to work out how much gain the "SRPP" will have to supply. However, the ECC99, 6n1p, 12AV7, and 12AT7 with mus in the 20s, 30s, 40s, and 60s respectively should provide you with a good choice.
BTW, are you going to triode wire the KT88s or run them UL? If you are triode wiring, I suggest you acquire a set of EI KT90s manufactured before the current quality control problems set in.
I have not tried to work out how much gain the "SRPP" will have to supply. However, the ECC99, 6n1p, 12AV7, and 12AT7 with mus in the 20s, 30s, 40s, and 60s respectively should provide you with a good choice.
BTW, are you going to triode wire the KT88s or run them UL? If you are triode wiring, I suggest you acquire a set of EI KT90s manufactured before the current quality control problems set in.
UL with my current speakers, and probably triode I guess when I get round to building my new ones. I thought about KT90's, aren't the new Electro Harmonix ones supposed to be quite good? The availability of the older EI's put me off a bit. As for little driver tubes I have some spare 6922 and 6n1p from my current amp
Steve
Steve
A number of people, including Jim McShane, have reported a flat sound stage when the EH KT90 is used.
Jim McShane likes the EH KT88, but still prefers properly made EI KT90s.
IMO, for driving the 6550/KT88/KT90, the 6n1p is a better choice than the 6922.
Ill keep an eye out then. Whenever I see them advertised it always says theyre the new mk3 versions or something like that though, which i guess is bad.
I tried the 6n1p's (I only have 2) and on one the heater and cathode seems to glow brighter, and seems to suffer quite badly from heater hum (still running ac heaters on the drivers 🙁) so Ive had to go back to the original 6922's for now until I go home and can build stuff
I tried the 6n1p's (I only have 2) and on one the heater and cathode seems to glow brighter, and seems to suffer quite badly from heater hum (still running ac heaters on the drivers 🙁) so Ive had to go back to the original 6922's for now until I go home and can build stuff
It's not fair! I really want to get started with this, plotting out loadlines and stuff but I can't because I have an exam tomorrow 
and my brains just going to be pondering all of this stuff now hehe.
Thanks,
Steve
and my brains just going to be pondering all of this stuff now hehe.By that do you mean that the resistance tends towards infinity? Is that how the top half ends up acting more like a constant current device? Im not sure but I think Im starting to understand all this a bit now. Also I noted that in the pictures in the book it shows LED's providing the cathode bias voltage for the lower valve. This sounds like a good idea to me (for a low current driver stage at least) as the bias voltage remains unchanged irrespective of what the current is, but I wondered what you all thought on that point too.
Thanks,
Steve
Yes.
The top tube senses voltage across the resistor (in SRPP/mu follower; in this case, the CCS) and tries to reduce the voltage towards zero. The net result being it acts as a CCS to the bottom tube, and a CF to the output (if taken from top cathode). If the resistor is zero, it sees no signal and is simply a biased tube, no more than a chunky, nonlinear resistor. (Taking the output from its cathode will still have a small advantageous effect.) With some finite resistance, it senses as mentioned. With an infinite resistance, current cannot change and thus any variation appears immediately on the top tube's grid, where it is canceled, keeping stability.
A very good approximation of this could be done with three triodes, basically you stack another "top tube" on a mu follower.
Tim
The top tube senses voltage across the resistor (in SRPP/mu follower; in this case, the CCS) and tries to reduce the voltage towards zero. The net result being it acts as a CCS to the bottom tube, and a CF to the output (if taken from top cathode). If the resistor is zero, it sees no signal and is simply a biased tube, no more than a chunky, nonlinear resistor. (Taking the output from its cathode will still have a small advantageous effect.) With some finite resistance, it senses as mentioned. With an infinite resistance, current cannot change and thus any variation appears immediately on the top tube's grid, where it is canceled, keeping stability.
A very good approximation of this could be done with three triodes, basically you stack another "top tube" on a mu follower.
Tim
Hi Tim,
Yeh I think that's what this artical in tubecad was on about
triple triode mu follower
Might be worth trying too i guess...
Steve
Yeh I think that's what this artical in tubecad was on about
triple triode mu follower
Might be worth trying too i guess...
Steve
- Status
- Not open for further replies.