I tried Kt-88 in triode with lot of voltage swing and the result was disappointing sound.
I never tried in my amps to have the EL-34 in triode, makes no sense at all, why break the sound of the smoothest of the tubes? I don't know what people try to achieve by going triode with such a tube.
I never tried in my amps to have the EL-34 in triode, makes no sense at all, why break the sound of the smoothest of the tubes? I don't know what people try to achieve by going triode with such a tube.
The beam type tubes are more efficient than a true pentode because of lower G2 current resulting from aligned grids. Maybe it came about by an attempt to avoid Philips' patent, but it resulted in a better tube with lower losses. It may well be a logical advancement after studying a true Pentode. A beam power tube equivalent to an EL34 would be the 6CA7. Bear in mind that the EL34 can be marked as that and really be a 6CA7. The structure of the 6CA7 is fatter than an EL34. So if it is a fat bottle, you are holding a beam power tube.
In my own amplifiers, I feel the 6CA7 sounds a lot better. It also doesn't burn the base to bits over time like EL34's tend to do.
-Chris
In my own amplifiers, I feel the 6CA7 sounds a lot better. It also doesn't burn the base to bits over time like EL34's tend to do.
-Chris
Blitz,
All very interesting. I am trying to wrap my brain round what sort of equivalent anode one would get by connecting a pentode in the various ways. Going to the Va-Ia graphs, I detect nothing but another triode, with which one can do a design etc. and get triode preformance. I.e. should one be able to experiment with various anode constructions however funny, all I see is an equivalent triode with typically triode characteristics.
Similarly then with the various triode 'models' one can make from a pentode as triode. So to speak: The design does not know what the physical construction of a tube is, electrons simply see potentials/fields on their way and act accordingly, i.e. as triodes. This, of course, while g2 and g3 stay at the same potential or potential difference as the anode.
With due respect for your hearing ability, I cannot translate those experiences into meaningful technical detail for me.
I would be interested in equivalent Va-Ia plots, but am not the proud possessor of a U-tracer.
All very interesting. I am trying to wrap my brain round what sort of equivalent anode one would get by connecting a pentode in the various ways. Going to the Va-Ia graphs, I detect nothing but another triode, with which one can do a design etc. and get triode preformance. I.e. should one be able to experiment with various anode constructions however funny, all I see is an equivalent triode with typically triode characteristics.
Similarly then with the various triode 'models' one can make from a pentode as triode. So to speak: The design does not know what the physical construction of a tube is, electrons simply see potentials/fields on their way and act accordingly, i.e. as triodes. This, of course, while g2 and g3 stay at the same potential or potential difference as the anode.
With due respect for your hearing ability, I cannot translate those experiences into meaningful technical detail for me.
I would be interested in equivalent Va-Ia plots, but am not the proud possessor of a U-tracer.
Chris,
I have handled both 6CA7 beam tubes and true pentodes. Others seem to have had the same experience but state that beam 6CA7s are rare! I cannot recall the brand of tubes, but in equipment find more 6CA7 pentodes down here than beam models. I suppose it depends on what has been imported. (There was a thread about the subject somewhere.)
Johan
I have handled both 6CA7 beam tubes and true pentodes. Others seem to have had the same experience but state that beam 6CA7s are rare! I cannot recall the brand of tubes, but in equipment find more 6CA7 pentodes down here than beam models. I suppose it depends on what has been imported. (There was a thread about the subject somewhere.)
Johan
Same impression here, my ear opted for the UL on KT-88, the triode sounded dull and with higher distortion. With the UL I then upgraded to 6.5K from 3.5K p-p, made it even better.
Well, I canunderstand that different people have different speakers, drivers and, and, and...so, in no means I would statethat what I hear and like is what you hear and like.
I tried originally to put in a simple switch, so that my power amp is switchable between UL and triode. Tried that and it was very disappointing.
Nevertheless with all these guys talk so high about triodes, I simply wanted to know what is possible. Utracer shows significantly changing tube parameters when connecting G3 differently, some of them like I posted above. The problem with utracer is that you need to manually export and import the data into excel to get meaningful curves.
As soon as my new toy ( RoeTest - Röhrenprüfgerät Röhrenmessgerät ) is finally tested (it is already build), I will test the EL34 in all the modes you want and publish the curves up to 600V.
When I give the EL34 bigger drive/gain amd corrected the bias, so it has 60-65mA at 350-360V in triode...and doubled the gain of my DAC (easy to do when you use passive I/V)..with the right speaker: For me there is no looking back. UL is history.
I tried originally to put in a simple switch, so that my power amp is switchable between UL and triode. Tried that and it was very disappointing.
Nevertheless with all these guys talk so high about triodes, I simply wanted to know what is possible. Utracer shows significantly changing tube parameters when connecting G3 differently, some of them like I posted above. The problem with utracer is that you need to manually export and import the data into excel to get meaningful curves.
As soon as my new toy ( RoeTest - Röhrenprüfgerät Röhrenmessgerät ) is finally tested (it is already build), I will test the EL34 in all the modes you want and publish the curves up to 600V.
When I give the EL34 bigger drive/gain amd corrected the bias, so it has 60-65mA at 350-360V in triode...and doubled the gain of my DAC (easy to do when you use passive I/V)..with the right speaker: For me there is no looking back. UL is history.
Last edited:
Not sure how to interpret the above ..... To me one would use an EL34 (or for that matter any pentode) as triode where a 27W dissipation triode is desired, for whatever reason.
When you talk of "the smoothest of the tubes", you probably mean as pentodes or in UL? But as triodes they simply have different characteristics to UL or pentodes. ("Smoothness" or whatever is more dependant on circuit design than any particular tube. To me it is simply about application, not relative merits of topologies.)
P.S: One must remember that in designs with a triode-UL-pentode switch, any global NFB is different for the relative topologies, thus would also the distortion be.
Last edited:
I agree that the EL34 can sound very, very good, both in UL, but to my surprise even more in triode. And the Mullard data sheet gives it a design max at 30W in triode, so we have won another 5W !!!, So you can run it really in deep Class A, which should be the mode you want to be in anyhow.
utracer is pretty cheap, maybe 300 Euro all together...roetest is about 800-1000, but still cheap if you compare it with the high-end-tester out there...and FAR better...you do not programm and connectors, the program knows your tube and makes the right connections....600V...and, and, and...its like christmas...hey, you GOT TO CELEBRATE your mid-life crisis, this is why nature gave it to you !
By the way: I have still a build-kit for Roetest v7 (I waited to long to build mine and by the time v9 was out which I have built now). So if you are interested, send me a PM.
utracer is pretty cheap, maybe 300 Euro all together...roetest is about 800-1000, but still cheap if you compare it with the high-end-tester out there...and FAR better...you do not programm and connectors, the program knows your tube and makes the right connections....600V...and, and, and...its like christmas...hey, you GOT TO CELEBRATE your mid-life crisis, this is why nature gave it to you !
By the way: I have still a build-kit for Roetest v7 (I waited to long to build mine and by the time v9 was out which I have built now). So if you are interested, send me a PM.
Last edited:
Reference for my point (design with proper feedback, less stages, not too much damping and UL is the best!!!) Triode will 'help' not really a poor UL design.
Triode Connected Pentodes Article By Herb Reichert of Eddy Electronic Inc. Sound Practices Magazine Online
I don't know, triode is sounding slower. Triode has more harmonics at all powers. Triode EL34 is not using the tube potential. Maybe the problem reside in the difficulty of making an excellent ultralinear design? This issue is resolved by a mellowed triode sound to cover the design faults making predominant the tube distortions.
Technically: too much third and odd harmonics in ultralinear, solution: cover it all with overall increase and higher 2nd.
Link to triode connected pentodes:
http://www.audiodesignguide.com/New2A3/ETF06TS.pdf
Triode Connected Pentodes Article By Herb Reichert of Eddy Electronic Inc. Sound Practices Magazine Online
I don't know, triode is sounding slower. Triode has more harmonics at all powers. Triode EL34 is not using the tube potential. Maybe the problem reside in the difficulty of making an excellent ultralinear design? This issue is resolved by a mellowed triode sound to cover the design faults making predominant the tube distortions.
Technically: too much third and odd harmonics in ultralinear, solution: cover it all with overall increase and higher 2nd.
Link to triode connected pentodes:
http://www.audiodesignguide.com/New2A3/ETF06TS.pdf
Last edited:
Hi Johan,
Here in North America real 6CA7 are easy to find. A new manufacturer is New Sensor and the Electroharmonix line. I just bought another 8 of these for a pair of prototype amplifiers. I'll see if I can test a similar circuit using 6L6 tubes with the same transformer in UL configuration.
-Chris
Here in North America real 6CA7 are easy to find. A new manufacturer is New Sensor and the Electroharmonix line. I just bought another 8 of these for a pair of prototype amplifiers. I'll see if I can test a similar circuit using 6L6 tubes with the same transformer in UL configuration.
-Chris
OK, when I measure the EL34 without g3 connected at all, I have 1mA less emission than with g3 connected at anode. Pretty similar and expected result as the grid is not pulling any electrons through, but as well is not pushing any back. So an emission less than g3 at anode and more than g3 at cathode was the expected result. check-mark.
I am using in my preamp another pentode: the 4p1l. I looked into its internal structure and there the g3 is not woven like in the EL34 very wide spaced and far away: It is woven very tight and looks pretty near between g2 and anode.
So, I measured it, all g2 at anode, so triode mode at 163V/Vg -9.5V:
g3
- at cathode: 39mA, Ra=1.58kohm, gm=5.92, mu=9
- not connected at all: 37mA, Ra=1.52kohm, gm=6.06, mu=9
- at anode: 50mA, Ra=1.21kohm, gm=6.81, mu=8
I have to build a new setup with new sockets to do the listening test...a project for next weekend...but it is a bit surprising that emission goes down when g3 is not connected compared to g3 connected to cathode.
The only reason I could imagine is: The g3 is massive and tight in a 4p1l. Very different than the wide and loose g3 in the EL34. I mean: A lot of metal it is. It seems to be able to absorbt electrons rather than push them along. But inly my 2 ct...we will see what sounds bests, but obviously a general rule what to do with g3 might be wrong.
I am using in my preamp another pentode: the 4p1l. I looked into its internal structure and there the g3 is not woven like in the EL34 very wide spaced and far away: It is woven very tight and looks pretty near between g2 and anode.
So, I measured it, all g2 at anode, so triode mode at 163V/Vg -9.5V:
g3
- at cathode: 39mA, Ra=1.58kohm, gm=5.92, mu=9
- not connected at all: 37mA, Ra=1.52kohm, gm=6.06, mu=9
- at anode: 50mA, Ra=1.21kohm, gm=6.81, mu=8
I have to build a new setup with new sockets to do the listening test...a project for next weekend...but it is a bit surprising that emission goes down when g3 is not connected compared to g3 connected to cathode.
The only reason I could imagine is: The g3 is massive and tight in a 4p1l. Very different than the wide and loose g3 in the EL34. I mean: A lot of metal it is. It seems to be able to absorbt electrons rather than push them along. But inly my 2 ct...we will see what sounds bests, but obviously a general rule what to do with g3 might be wrong.
Last edited:
G3 is to remove the 'kink', it should always connect to K. It has almost no influence on the tube, if you don't connect it the electrons will bounce off it and I doubt it will sound 'fast'.
The reason for the G3 is to remove electrons which are not accelerated enough and lack speed. Proper speeded-up emission will pass through G3 (connected or not) and 99.9% reach the A.
When we say triode sound slow it is relative to the thinner 'faster' sound of pentodes, it is a way to describe it without pejorative meaning.
The reason for the G3 is to remove electrons which are not accelerated enough and lack speed. Proper speeded-up emission will pass through G3 (connected or not) and 99.9% reach the A.
When we say triode sound slow it is relative to the thinner 'faster' sound of pentodes, it is a way to describe it without pejorative meaning.
Last edited:
The outcome from a floating electrode in the electron stream is undefined and so no result can be expected. This is because a floating electrode could collect electrons and go negative, or it could emit lots of secondary electrons and go very positive. Which it does depends on many things, including the precise details of its surface. Hence no electrode should ever be left floating, unless you want strange results which may vary with valve sample and from day to day. You were just lucky.Blitz said:
There are no surprises when electrodes float. In this case it seems that g3 went negative.
Not sure I see this. Comparatively, triodes have mostly even order harmonics, which could be dealt with in a proper p.p. arrangment. Pentodes have more odd-order harmonics, which, even after proper design choices, can only be dealt with by NFB.
The purpose is surely to get rid of any harmonics, at least to below audibility, and this can be achieved and was done in several ckassic designs e.g. the Williamson amplifier et al. (The reduction/elimination of higher order harmonic distortion is mainly a semiconductor problem, not a tube based one.)
Though it has less to do with the topic, perhaps another look at characteristics of a typical power tube in UL vs the screen tap from 0% (pentode) to 100% (triode) might be informative. (There is a well-known graph by G.E. of this for the KT88. I am unable to look for that at present, perhaps another member can oblige please?)
It will be observed that distortion decreases more rapidly than maximum output as the screen tap increases through the 25% - 40% region. The precise choice is mostly a balance of various parameters. Pentodes may have lower distortion than triodes, but that is pretty load depandant - not healthy with a loudspeaker load!
An excellent paper for the uninitiated by Herr Schlangen, colourful illustrations (a picture is worth etc.)
Last edited:
Exactly. A bit better triode when it is connected to anode. But is it audible at all?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.