Hi Tom,
Given my lack of technical background I can't argue with you on basis of maths; however there are different independent people that came up with similar circuits (see for example the thread "triode trick" in this forum)... if there isn't a problem, why would people see one and design a solution for it = defies all logic. The link I included IS to a commercial site but take a look in the forum where the technically-oriented sceptics can find more info on this topic. More importantly, the circuit has a more than subtle impact on sonics says my ears.... A pseude-triode probably behaves very much like a true triode, but I'm sure there's still a difference, and circuits ala bill Perkins / vaughn audio minimize the difference, so I believe.
Regards, Simon
Given my lack of technical background I can't argue with you on basis of maths; however there are different independent people that came up with similar circuits (see for example the thread "triode trick" in this forum)... if there isn't a problem, why would people see one and design a solution for it = defies all logic. The link I included IS to a commercial site but take a look in the forum where the technically-oriented sceptics can find more info on this topic. More importantly, the circuit has a more than subtle impact on sonics says my ears.... A pseude-triode probably behaves very much like a true triode, but I'm sure there's still a difference, and circuits ala bill Perkins / vaughn audio minimize the difference, so I believe.
Regards, Simon
It appears to me that much of this argument is a case of 'it does not follow', namely, if a directly heated triode is linear, it does NOT follow that if you want a linear triode, it must be directly heated.
Ditto WRT triode strapping - not to be confused with 'pseudo-triode' which is often what ultralinear and similar explicit means of providing plate feedback are called (remember that for a pentode, G2 is the equivalent of the triode plate). Triode strapped pentode/tetrode is a triode, but it does NOT follow that it is linear or, for that matter non-linear. It's just a different triode - one that was most probably NOT designed for best triode operation, or it would not have been a pentode to begin with. But, 'on occasion' the gods of electron flow have been kind and you get a pretty linear triode.
Also, comparing on 300B and 2A3 grounds involves some assumptions - in order to be efficient with practical loads, a triode has to have a fairly high gm and relatively low mu, i.e. low Rp. If one wants to compare 2A3, 300B etc with triode strapped tubes, then at least compare with most if not all the relevant parameters in mind.
Ditto WRT triode strapping - not to be confused with 'pseudo-triode' which is often what ultralinear and similar explicit means of providing plate feedback are called (remember that for a pentode, G2 is the equivalent of the triode plate). Triode strapped pentode/tetrode is a triode, but it does NOT follow that it is linear or, for that matter non-linear. It's just a different triode - one that was most probably NOT designed for best triode operation, or it would not have been a pentode to begin with. But, 'on occasion' the gods of electron flow have been kind and you get a pretty linear triode.
Also, comparing on 300B and 2A3 grounds involves some assumptions - in order to be efficient with practical loads, a triode has to have a fairly high gm and relatively low mu, i.e. low Rp. If one wants to compare 2A3, 300B etc with triode strapped tubes, then at least compare with most if not all the relevant parameters in mind.
True. One reason DH power triodes are good is the combination of relatively low Rp and high gm. High mu DH triodes are mediocre at best. That's the reason novals tend to be the preferred gain/driver tubes for DH power triodes.
In the heyday of DHTs they couldn't make a tube with low Rp, high mu and gm. At least nobody did. The E55L/8233, when triode strapped, has a mu of 50 and Rp of 600 Ohms. And it's as linear as any tube. But the latter can be said of lots of noval pentodes.
In the heyday of DHTs they couldn't make a tube with low Rp, high mu and gm. At least nobody did. The E55L/8233, when triode strapped, has a mu of 50 and Rp of 600 Ohms. And it's as linear as any tube. But the latter can be said of lots of noval pentodes.
phn said:
looks like a triode, smells like a triode, distorts like a triode must be a triode?
I wonder if the confusion here arises from the idea that a pentode strapped as a triode doesnt sound like a mental 'reference triode'. this is not arguable. A triode strapped EL34 for instance wouldn't sound like 300B but then It wouldnt sound like a 6AS7 or trioded KT88 either
Klimon, I know youve tried the triode trick but bear with me here:
The circuits that try to minimise any lingering effects of the screen in triode operation, are making the pseudo-triode more like a true triode. But with the internal grid, cathode and anode spacings of the base pentode. Probably not the same as a true triode, thus the difference! Fundamentally its IS a different tube after all. Using a resistor won't suddenly turn it back into a pentode.
Its just that Bill Perkins' circuit adresses some of the compromises inherent in it not being a triode in the first place. OTOH, What about the compromises inherent in having a DHT heater supply floating around in the signal path!
Andy
Exactly my thoughts, although the words are different
@ Phn:
If you will I'ld say a pseudo-triode IS (electrically) a triode. If you permit some nuancation I would propose that although a pseudo-triode is very near a pure triode, there are downsides to the standard resistor strap method (downsides lessened by the extra Bill Perkins blablabla circuit) that should be acknowledged, and form a difference with pure triodes. I perceived this difference (as exemplified by my one amp, hardly pure science but I didn't invent the triode trick so it's not a one-man show) as being negative sound-wise, when compared with the extra circuit which minimises the difference.... There's no point in a "A or B" discussion, my stance is "A with a taste of lemon"; bless Aristotle.
Simon
salas said:
The Western Electric 417A, 437A, EC8010, EC8020, triode-strapped E55L, D3a etc. are all high mu (40 to 70), high gm, low Rp and ultra-low distortion.
The 2A3, 300B, 205D, AD1, RE604, etc. are all great tubes. But they are only great because they stopped making and improving DH triodes.
Klimon said:
Exactly my thoughts, although the words are different
@ Phn:
If you will I'ld say a pseudo-triode IS (electrically) a triode. If you permit some nuancation I would propose that although a pseudo-triode is very near a pure triode, there are downsides to the standard resistor strap method (downsides lessened by the extra Bill Perkins blablabla circuit) that should be acknowledged, and form a difference with pure triodes. I perceived this difference (as exemplified by my one amp, hardly pure science but I didn't invent the triode trick so it's not a one-man show) as being negative sound-wise, when compared with the extra circuit which minimises the difference.... That's my point I think
Simon
So what is it that makes one chip a D/A converter and another chip an op-amp if the chips aren't defined by their characteristics?
Purest sound, and/or phase shifts that you cannot feedback around on top of the OPT. And the more gain needed caveat still applies, cmiller may be an issue with stepup.
The possibly sacrilegious answer is a (either FET or sweep tube) follower driven by a Western Electric 417A, 437A, EC8010, EC8020, triode-strapped E55L, D3a etc.
The possibly sacrilegious answer is a (either FET or sweep tube) follower driven by a Western Electric 417A, 437A, EC8010, EC8020, triode-strapped E55L, D3a etc.
Tweeker said:
It is exactly what I've found. 50L6GC triode strapped drives solid state voltage follower loaded by a voltage to current converter, like this one:
For Hi-fi idle current equals to a half of maximal output current (maximal efficiency), for high end it equals to the maximal current (a voltage follower always sees the same emitter current: minimal possible distortions).
If the 300Bs in question are really as good as Western Electric claims, then do do have some advantages over EL34s. I keep thinking about a pair in push pull at 100ma with an 11K anode to anode load.
Curves are at 350V anode, with fixed bias.
Curve 1: 30ma
Curve 2: 40ma
Curve 3: 60ma
Curve 4: 80ma
Curves are at 350V anode, with fixed bias.
Curve 1: 30ma
Curve 2: 40ma
Curve 3: 60ma
Curve 4: 80ma
Attachments
Hi Tweeker,
Interesting half-sentence in front. Two weeks ago I was given some boutique 300B (considered to be somehow "improved" 300B) to trace, and after sending the plots back to the distributor, he very angrily called me by phone, forbade me to publish those and mention the brand by threatening for lawsuits because it would harm the business. "With your plots you have done enough harm on business already anyway", he shouted. Oh well, some folks just can´t stand the truth, especially when it will have impact on their wallet.
But usually a good real (non-boutique) 300B, WE or not, is considerable more linear than a triode strapped plain EL34, hands down, indeed. Usually.
For those who don´t believe, they should stop repeating legends and pointing at stone-age plots, but should apply my tag-line and see for themselves.
Regards,
Tom
EDIT: A similar effect can be found with modern KT88/6550. Patrick Turner (Australia) did some interesting measurements with these, showing how off modern copies are compared to old original datasheet plots. Funny enough, everbody makes SPICE models after those old plots and draws load lines on those old plots and then, using current production tubes, wonders what is wrong .....
Interesting half-sentence in front. Two weeks ago I was given some boutique 300B (considered to be somehow "improved" 300B) to trace, and after sending the plots back to the distributor, he very angrily called me by phone, forbade me to publish those and mention the brand by threatening for lawsuits because it would harm the business. "With your plots you have done enough harm on business already anyway", he shouted. Oh well, some folks just can´t stand the truth, especially when it will have impact on their wallet.
But usually a good real (non-boutique) 300B, WE or not, is considerable more linear than a triode strapped plain EL34, hands down, indeed. Usually.
For those who don´t believe, they should stop repeating legends and pointing at stone-age plots, but should apply my tag-line and see for themselves.
Regards,
Tom
EDIT: A similar effect can be found with modern KT88/6550. Patrick Turner (Australia) did some interesting measurements with these, showing how off modern copies are compared to old original datasheet plots. Funny enough, everbody makes SPICE models after those old plots and draws load lines on those old plots and then, using current production tubes, wonders what is wrong .....
A similar effect . . . . occured some 30 years ago when tubes disapeared from new consumers products like TV set.
Large respected manufacturers stopped their tube factories and it was almost impossible to find decent replacement for maintenance.
The most sollicited tubes like horizontal sweep and VHF/UHF front end was **** and failed at incredible rate.
This forced the customer to replace its old TV set for a new one !
All (somewhat) aged maintenance technician (like me) could confirm that.
Business is business
Yves.
Large respected manufacturers stopped their tube factories and it was almost impossible to find decent replacement for maintenance.
The most sollicited tubes like horizontal sweep and VHF/UHF front end was **** and failed at incredible rate.
This forced the customer to replace its old TV set for a new one !
All (somewhat) aged maintenance technician (like me) could confirm that.
Business is business
Yves.
The PX4 has similarities with a triode-strapped KT66.
Both are incredibly linear.
IMHO a triode-strapped KT66 could certainly be reagrded as a poor-man's PX4.
I believe this was the intention of MOV, to produce a valve that could replace the PX4.
And sonically, a PX4 wipes the floor with a 300b.
Both are incredibly linear.
IMHO a triode-strapped KT66 could certainly be reagrded as a poor-man's PX4.
I believe this was the intention of MOV, to produce a valve that could replace the PX4.
And sonically, a PX4 wipes the floor with a 300b.
I haven't heard them myself. But is the KT66 a poor man's PX4 because of any sonic difference or because of glamour, in this case lack thereof?
Any low-power DH triode "wipes the floor with a 300b." Usually that's because the 300B is harder to drive and needs a superior driver tube. "Nobody" can fail with a 45. (That doesn't sound right. But I guess it's true for the .45 as well.)
Any low-power DH triode "wipes the floor with a 300b." Usually that's because the 300B is harder to drive and needs a superior driver tube. "Nobody" can fail with a 45. (That doesn't sound right. But I guess it's true for the .45 as well.)
jarthel said:I thought I was clear when I wrote the thread.
hopefully discussions can be restricted to power output tubes like el34, kt88, 6550, 300b, 845, 2a3 and etc.
like why use a 300b at all when el34 are much cheaper. does a 300b has a technical/sonic advantage?
Well it (the 300b or 845) should have a sonic advantage. All other things being equal, the el34 is closest, so far as I can tell others rave over the el84, (this assumes that you can approximately believe the tube curves) and your power out put reqiorements are not fierce
BUT the "problem" in general with DHT triodes is that the heater is in the signal path. So whatever the curves suggest thats the first practical problem to overcome. I'm doing a 300b amp now and this is no small problem. An indirectly heated EL34 looks beautiful just at the moment... Maybe one day I'll get both the el34 amp and the 300b amp working at the same time and then I'll know if its worth the bother.
phn, the KT66 is just lovely if the new tubes match the old ones for performance. But the KT** series are not the most efficient as triodes I believe? I know that the extra current they use can cause LF performance problems in the transformer though. Thats another reason the 300b (and el34) scores, usually low current with lots of voltage. IMHO not the best way, as it seems to negate a lot of that fantastic linearity.
Cheers
Andy
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.