test results
Hi to all, especially tubelab who has done quite a job... thank you as well to rdf for his effort!
Let me say that among you I feel as a puny teoretitian... no, I'm kidding!
I would like to say that when it comes to the effects of g3 potential, as explained in the part of the thread previous to the published results, those effects are applicable to pentode mode connection. Let me remind you that it is the AC path that matters, if it is not the same for a and g2 than it is pentode connection...
Secondly, the results for g3 tied to anode are quite bad. I would say that it has to do with the tube, the testing rig (equipment and parts) or similia. In effect, that is a result I would call "faulty tube". Tying g3 to one potential or the other in triode mode CANNOT make such a difference in curves, especially in the precision of the curves. Furthermore, the graph was made in excel or similia, it is not a Sofia tester... how many points were taken? Is the equipment working correctly, given the state the lab is in? Sorry about the hurricanes... we used to have similar difficulties, but now bombing season is over (no pun intended -- and most seriously felt).
Well, that is about it, for the time being. I can make you a similar plot using a good spice model. Problem is, it will not show the same problems your plot shows -- because the tube model is a perfect tube, yours might be faulty, as well as the equipment or connections. It is inconclusive, to say the best -- but to people who know little theory, if any, and have little practice... they can mean a life (of mistaken beliefs) of difference.
Btw, how many times did you repeat the measurements? Did you try any other tube?
Best regards,
Aleksandar
Hi to all, especially tubelab who has done quite a job... thank you as well to rdf for his effort!
Let me say that among you I feel as a puny teoretitian... no, I'm kidding!
I would like to say that when it comes to the effects of g3 potential, as explained in the part of the thread previous to the published results, those effects are applicable to pentode mode connection. Let me remind you that it is the AC path that matters, if it is not the same for a and g2 than it is pentode connection...
Secondly, the results for g3 tied to anode are quite bad. I would say that it has to do with the tube, the testing rig (equipment and parts) or similia. In effect, that is a result I would call "faulty tube". Tying g3 to one potential or the other in triode mode CANNOT make such a difference in curves, especially in the precision of the curves. Furthermore, the graph was made in excel or similia, it is not a Sofia tester... how many points were taken? Is the equipment working correctly, given the state the lab is in? Sorry about the hurricanes... we used to have similar difficulties, but now bombing season is over (no pun intended -- and most seriously felt).
Well, that is about it, for the time being. I can make you a similar plot using a good spice model. Problem is, it will not show the same problems your plot shows -- because the tube model is a perfect tube, yours might be faulty, as well as the equipment or connections. It is inconclusive, to say the best -- but to people who know little theory, if any, and have little practice... they can mean a life (of mistaken beliefs) of difference.
Btw, how many times did you repeat the measurements? Did you try any other tube?
Best regards,
Aleksandar
Exactly, that's one of the drawbacks of simulation, as has been pointed out to me before on this forum. How does Spice know the quality of your OPT or which EL34 you're using: NOS Mullard, Winged C, Sovtek etc? Most people would agree these tubes sound quite different from one another and it's possible that some examples work better as triodes with G3 connected to plate, whilst others work better with G3 connected to the cathode.
Incidentally, Steve Bench has an example of a PP amp with triode-connected EL34s (actually E34L) at his web site, in which g3 goes to plate via 1.5k and g2 via 1k. He doesn't say much about it, though!
I admit that the tube might be weak. It was the only EL-34 that I could find on short notice. It was one of the original tubes in a Scott amplifier (45 years old), which does sound nice. I took the measurements by hand and entered the data into Excel, since I don't have a curve tracer. The equipment is fine (only meters and power supplies). The experiments were run only once since it took over an hour each. I was graphing the data as it was being taken, and I went back and verified the kink in the curves , and took a few additional points, since I didn't expect the kink. I believe the data (been doing this kind of work for 30 years for a major electronics company), but the tube may be suspect. It is likely that other EL-34's and other types of pentodes will give different results. Clearly, more experiments are in order.
After I took this data, I began to experiment with other connections. This led to another surprise. I tied all 3 grids together through 150 ohm resistors. I ran curves starting at zero bias, and going up (positive) in two volt steps. The curves looked very linear and very PENTODE like. It got too late to finish, so I will finish it tonight.
All of this theory stuff is good, but the proof is in the sound. Yes I will eventually put all of this to good use. An amp will be built. Probably a SE amp that can take any of the popular tubes. I think that I can use a resistive load, to eliminate the transformer, and measure frequency response. I can measure to 1.2 MHz.
This leads to another choice. I have no new (trustworthy) EL-34's. I haven't done many amps with them (I favor 6550's and KT-88's) so I am not familiar with all the different ones. I will get a few new ones for experiments. I can't go for any of the "boutique" brands, but are there any suggestions for a good low cost EL-34. I am looking for something that is rather neutral and reliable, since I will test it to its limits.
JJ, Valve Art, and Electro Harmonix are all cheap, are any of these worthy of trying.
After I took this data, I began to experiment with other connections. This led to another surprise. I tied all 3 grids together through 150 ohm resistors. I ran curves starting at zero bias, and going up (positive) in two volt steps. The curves looked very linear and very PENTODE like. It got too late to finish, so I will finish it tonight.
All of this theory stuff is good, but the proof is in the sound. Yes I will eventually put all of this to good use. An amp will be built. Probably a SE amp that can take any of the popular tubes. I think that I can use a resistive load, to eliminate the transformer, and measure frequency response. I can measure to 1.2 MHz.
This leads to another choice. I have no new (trustworthy) EL-34's. I haven't done many amps with them (I favor 6550's and KT-88's) so I am not familiar with all the different ones. I will get a few new ones for experiments. I can't go for any of the "boutique" brands, but are there any suggestions for a good low cost EL-34. I am looking for something that is rather neutral and reliable, since I will test it to its limits.
JJ, Valve Art, and Electro Harmonix are all cheap, are any of these worthy of trying.
I think it is time I dusted off that old HP X-Y plotter I have under my work table and do some curves 
the only problem is I don't have a high enough output current HV power supply for an EL34
Anyway, I was recently browsing through some russian data on tubes from the cold war era. There was a 'universal' pentode called 12Z3L, which was used in various equipment, one of them being a portable transciever where it was the only type of tube used except for the power output tube (the reason probably being that even under the influence of a gallon of Vodka you could not get the tubes wrong ). It's data sheet is interesting because it is the only one where I have seen a Vg3 vs Ip plot. This is a true pentode, and the plot shows that it takes about -80V to drive the tube into cut-off via G3. In comparison, it takes about 10 times less to do the same through G1. The existence of the raph indicates thet the tube probably was used in circuits where G3 was used to control the electron flow...
One more thing - G3 is not really a grid but rather a beam former in beam tetrodes, which suggests that it would not be a good idea to tie it to the plate. In general, datasheets for beam tetrodes show a maximum of 30V on the beam former, I suppose if you go higher it starts acting as a plate and attracts electrons, also the beams are not formed properly so secondary emission increases.
the only problem is I don't have a high enough output current HV power supply for an EL34 Anyway, I was recently browsing through some russian data on tubes from the cold war era. There was a 'universal' pentode called 12Z3L, which was used in various equipment, one of them being a portable transciever where it was the only type of tube used except for the power output tube (the reason probably being that even under the influence of a gallon of Vodka you could not get the tubes wrong
). It's data sheet is interesting because it is the only one where I have seen a Vg3 vs Ip plot. This is a true pentode, and the plot shows that it takes about -80V to drive the tube into cut-off via G3. In comparison, it takes about 10 times less to do the same through G1. The existence of the raph indicates thet the tube probably was used in circuits where G3 was used to control the electron flow...One more thing - G3 is not really a grid but rather a beam former in beam tetrodes, which suggests that it would not be a good idea to tie it to the plate. In general, datasheets for beam tetrodes show a maximum of 30V on the beam former, I suppose if you go higher it starts acting as a plate and attracts electrons, also the beams are not formed properly so secondary emission increases.
The EL-34 is a true pentode, and one of the few power pentodes with G3 pinned out seperately. Most beam tetrodes have G3 internally connected to the plate. The obvious exception is the 807 which has G3 available. I have a bunch of these so some experiments are in order. There are other transmitting tubes but I don't have any.
There are some sweep tubes that have G3 pinned out. This was done to carefully tweak the tubes linearity, and eliminate parasitic oscillations. The max spec for G3 on most of them was 30 volts.
I think that I am going to have to rig up some type of automated curve tracer, since I have a lot of tubes to test.
There are some sweep tubes that have G3 pinned out. This was done to carefully tweak the tubes linearity, and eliminate parasitic oscillations. The max spec for G3 on most of them was 30 volts.
I think that I am going to have to rig up some type of automated curve tracer, since I have a lot of tubes to test.
If you can get a hold of an X-Y plotter, it's perfect for this kind of thing. Most have differential X and Y inputs and variable sensitivity.
Essentially it uses a servo system to move a pen on a sheet of paper to coordinates set by voltage inputs for X and Y. Sensitivity is set in steps usually in V/cm, and you also have pots to set the X and Y origin.
When you draw the curves, you basically just set grid voltage(s), and do a manual 'sweep' of the plate voltage on your power supply, all the way up and then back down. Set new grid voltage(s) and repeat.
Since I mostly deal with silicon rather than vacuum, I used mine for things like bias current variation and thermal runaway checking because the model I have has a built in time base, that goes from some 1cm/s down to 1cm/30 minutes.
It should be possible to get one of these cheap on eBay...
Essentially it uses a servo system to move a pen on a sheet of paper to coordinates set by voltage inputs for X and Y. Sensitivity is set in steps usually in V/cm, and you also have pots to set the X and Y origin.
When you draw the curves, you basically just set grid voltage(s), and do a manual 'sweep' of the plate voltage on your power supply, all the way up and then back down. Set new grid voltage(s) and repeat.
Since I mostly deal with silicon rather than vacuum, I used mine for things like bias current variation and thermal runaway checking because the model I have has a built in time base, that goes from some 1cm/s down to 1cm/30 minutes.
It should be possible to get one of these cheap on eBay...
I had one and I sold it several years ago. Unfortunately, my work area is small, and larger pieces of equipment that don't get used often has to go. I used to connect it to the X-Y outputs on my HP8903A audio analyzer, push the button, and you get an instant frequency response plot. Now I use a sound card and some software to do the same thing. I will put together something with a PIC chip that dumps data into Excel.
I asked about which EL-34's to buy and I get no response? I thought that question would start a storm of response. I guess we would rather debate Physics?
OK, I am going to plot the curves again, along with the curves for some unusual connections. Which tube do you want to see. Electro Harmonix, JJ, Valve Art (all $10 USD each) or some other low cost EL-34.
I have Orion (old Russian tubes, NIB) and Penta Labs (unknown origin, NIB) I consider both of these to be junk.
I asked about which EL-34's to buy and I get no response? I thought that question would start a storm of response. I guess we would rather debate Physics?
OK, I am going to plot the curves again, along with the curves for some unusual connections. Which tube do you want to see. Electro Harmonix, JJ, Valve Art (all $10 USD each) or some other low cost EL-34.
I have Orion (old Russian tubes, NIB) and Penta Labs (unknown origin, NIB) I consider both of these to be junk.
Modern EL34 experience
Hi Tubelab,
I'm using JJ EL34s in
1) Chinese "Music Angel" EL34 - modified
2) Menno Vandervenne PPP EL34 (PAT4006 trannies) - modified
In both cases I run them triode strapped with G3 to Cathode and G2 to anode via 1K.
I tried some Chinese EL34, some "Winged C" Svetllana and the JJ's - the JJ's were the clear winner on pure listening tests, that is the standard Gingertube Ear was the only test equipment used.
I've bought 4 factory matched quads of JJ EL34 in the last 12 weeks (that is 16 tubes in total) - 1 tube discarded due to unstable anode current, one tube failed with an grid1 to cathode short after 3 days use. Have 3 good matched quads, 1 matched pair (slightly weak) and 2 duds from my 16 tube purchase.
So my recommendation is JJs.
I did however do a restoration on a vintage Hiwatt SA212 Guitar Amp recently and the guy (a very experienced professional muso) thought the Winged C Svets gave him the closest sound to the original Mullards (in pentode mode). They are nearly twice the cost of the JJs.
Cheers,
Ian
Hi Tubelab,
I'm using JJ EL34s in
1) Chinese "Music Angel" EL34 - modified
2) Menno Vandervenne PPP EL34 (PAT4006 trannies) - modified
In both cases I run them triode strapped with G3 to Cathode and G2 to anode via 1K.
I tried some Chinese EL34, some "Winged C" Svetllana and the JJ's - the JJ's were the clear winner on pure listening tests, that is the standard Gingertube Ear was the only test equipment used.
I've bought 4 factory matched quads of JJ EL34 in the last 12 weeks (that is 16 tubes in total) - 1 tube discarded due to unstable anode current, one tube failed with an grid1 to cathode short after 3 days use. Have 3 good matched quads, 1 matched pair (slightly weak) and 2 duds from my 16 tube purchase.
So my recommendation is JJs.
I did however do a restoration on a vintage Hiwatt SA212 Guitar Amp recently and the guy (a very experienced professional muso) thought the Winged C Svets gave him the closest sound to the original Mullards (in pentode mode). They are nearly twice the cost of the JJs.
Cheers,
Ian
I got two responses today that mention 3 brands of tubes. I am going to be out of town for most of the weekend (4 day holiday in the US) , so I decided to order the tubes tonight.
I thought about it, and thought about the totally rude person that answered the phone when I called New Sensor a few years ago, and decided I couldn't give them any more of my money. Sorry.
I ordered 2 JJ's and 2 Svet winged C's. I have never tried JJ's but a friend has them in his Cary, and likes them. I have used, and abused other Svetlana tubes in the past with no problems.
I thought about it, and thought about the totally rude person that answered the phone when I called New Sensor a few years ago, and decided I couldn't give them any more of my money. Sorry.
I ordered 2 JJ's and 2 Svet winged C's. I have never tried JJ's but a friend has them in his Cary, and likes them. I have used, and abused other Svetlana tubes in the past with no problems.
I hope that their customer service has improved in the years since that incident happened.
I had 3 of their 300B's die all from half of the filament going open within 5 months of purchase. These cost me $75USD each. I communicated with New Sensor by e-mail and phone 3 times with the same result. Tough S***. The last response actually suggested that I BUY their new Electro Harmonix 300B, it has an improved filament.
Unfortunately my story was not unique, I have heard from other people that had their 300B's and 6B4's die young with the same answer from New Sensor. I spent $225 for tubes that died, many of my complete amplifiers cost less than this.
I haven't heard of this happening recently, and I have used other Sovtek tubes (6550's) with good results. Maybe I will try them again if I keep hearing good things.
I had 3 of their 300B's die all from half of the filament going open within 5 months of purchase. These cost me $75USD each. I communicated with New Sensor by e-mail and phone 3 times with the same result. Tough S***. The last response actually suggested that I BUY their new Electro Harmonix 300B, it has an improved filament.
Unfortunately my story was not unique, I have heard from other people that had their 300B's and 6B4's die young with the same answer from New Sensor. I spent $225 for tubes that died, many of my complete amplifiers cost less than this.
I haven't heard of this happening recently, and I have used other Sovtek tubes (6550's) with good results. Maybe I will try them again if I keep hearing good things.
Ok, now that I have vented my frustration, I looked up at my 845SE, which is playing as I write this, and there they are, Sovtek 5AR4's. I forgot about them, why? Because they work good. It's funny that you vividly remember the bad experiences, and forget the good ones.
Why did I buy them? Because Ned at Triode electronics said that they would work good or he would refund my money. Given those circumstances I tried them (they were cheap too) and they are the best cheap 5AR4's that I have tried.
Maybe I overreacted. If I take this EL34 experiment far enough along to become an amplifier, maybe I will try them, they are cheap.
Why did I buy them? Because Ned at Triode electronics said that they would work good or he would refund my money. Given those circumstances I tried them (they were cheap too) and they are the best cheap 5AR4's that I have tried.
Maybe I overreacted. If I take this EL34 experiment far enough along to become an amplifier, maybe I will try them, they are cheap.
expertise and experience
Hi to all!
I did not necessarily need post this time, as I'm a little late on most topics, but I just had to
First -- about g3 and datasheets. The 307A datasheet shows lots of curves where g3 is varied, etc. Most tubes that have separate g3 connections were ment for applications in which g3 would be use to excercise some control on the electron flow. While most people think of valves as audio device, actually the vast majority of tubes were everything but audio devices -- just think of computers and radar and all that stuff which had to be made with tubes. I.e. there is one German scientist who built (arguably) the first computer that did something useful, at the beginning of WWII -- the man was just very similar to me in being in the wrong place at the wrong time. He died after 1964 but else I would hardly shake off the impression that I was his reincarnation and we were to suffer for some previous sins... or choose yourself.
Secondly, when I say g3, I mean pentode. Tetrode tubes do not have g3, and beam tetrode tubes might have something connected to a pin which might be marked as g3, but that is not a grid, rather a screen using electron optics etc. to the advantage of the tube's performance. Therefore, g3 to anode, that is something reserved to PENTODES that want to be TRIODES!!!
When it comes to tubes EL34, try finding some Ei Nis Yugoslavia -- those are very good and should not be too expensive. Even better the 6CA7 -- those will perform very well in vintage guitar amps etc.
Last but not least -- what was the original need to tie g3 anywhere, or transform pentodes into triodes? If you need to listen to triodes, for some unknown marketing caused reasons, go find yourself some real triodes, DH or IH like 6AS7 6080 5998 etc. On the other hand, if you own real world speakers, go build yourself an RH34 (my own version of SE EL34 with the power of the pentode and the sound of the triode... ).
You know, maybe this is self-gratifying, but if you look at it from the standpoint of one who wants to listen to good sound, it might just make your day.
Regards to all,
Aleksandar
PS
About spice and the rest -- tubes differ in sound, of course... mostly because of different materials used inside them, different geometry... etc. Furthermore, g3 will affect the sound in triode mode -- the discussion is probably trying to proove WHY will that happen. While g3 to anode is the only true way to "imitate" a triode... some might like it better with g3 to cathode or ground.
Hi to all!
I did not necessarily need post this time, as I'm a little late on most topics, but I just had to
First -- about g3 and datasheets. The 307A datasheet shows lots of curves where g3 is varied, etc. Most tubes that have separate g3 connections were ment for applications in which g3 would be use to excercise some control on the electron flow. While most people think of valves as audio device, actually the vast majority of tubes were everything but audio devices -- just think of computers and radar and all that stuff which had to be made with tubes. I.e. there is one German scientist who built (arguably) the first computer that did something useful, at the beginning of WWII -- the man was just very similar to me in being in the wrong place at the wrong time. He died after 1964 but else I would hardly shake off the impression that I was his reincarnation and we were to suffer for some previous sins...
or choose yourself.Secondly, when I say g3, I mean pentode. Tetrode tubes do not have g3, and beam tetrode tubes might have something connected to a pin which might be marked as g3, but that is not a grid, rather a screen using electron optics etc. to the advantage of the tube's performance. Therefore, g3 to anode, that is something reserved to PENTODES that want to be TRIODES!!!
When it comes to tubes EL34, try finding some Ei Nis Yugoslavia -- those are very good and should not be too expensive. Even better the 6CA7 -- those will perform very well in vintage guitar amps etc.
Last but not least -- what was the original need to tie g3 anywhere, or transform pentodes into triodes? If you need to listen to triodes, for some unknown marketing caused reasons, go find yourself some real triodes, DH or IH like 6AS7 6080 5998 etc. On the other hand, if you own real world speakers, go build yourself an RH34 (my own version of SE EL34 with the power of the pentode and the sound of the triode...
).You know, maybe this is self-gratifying, but if you look at it from the standpoint of one who wants to listen to good sound, it might just make your day.
Regards to all,
Aleksandar
PS
About spice and the rest -- tubes differ in sound, of course... mostly because of different materials used inside them, different geometry... etc. Furthermore, g3 will affect the sound in triode mode -- the discussion is probably trying to proove WHY will that happen. While g3 to anode is the only true way to "imitate" a triode... some might like it better with g3 to cathode or ground.
One of the tests of a good discussion is that it makes one think, and I daresay more has come from this than originally imagined. Thus musing .....
I would suggest that original intentions be kept in mind, however wide one's journey into application fields go (and I have done my share of wild applications! - did not work at a research organisation for nothing). The EL34 (as an example) was originally intended for use as a pentode, where G3 is there to control secondary emmision from the anode back to the screen, especially when the anode goes lower in voltage than G2 as in normal pentode practice. In that application it would not appear to have an influence on linearity, otherwise the manufacturers would have done something about it. It gives comparably low distortion as is, so the premise appears to be in order. With that outlook it seems illogical to expect that an improvement would result from taking G3 elsewhere than to common.
As I remember I have not seen too many data sheets for the EL34 (or 6CA7) as a triode, in contrast to beam tubes where they just about always occur. The triode mu is hardly ever given. All this might indicate that the manufacturers did not see the EL34 in use as a triode, although it was used as such. In the hey-day of tubes such applications made far more use of beam tubes, where there is nothing between G2 and anode to possibly upset things.
Perhaps there is a message in that.
Aleksandar, yes. Your argument is indeed logical regarding the use of "real" triodes. But then one must remember that power triode development seems to have stopped with the advent of pentodes and beam tubes. The latter was conveniently available in a range of models while triodes were not. (The ones you mention were developed mainly for regulator use etc.) So for me it was not unexpected that the original hi-fi generation (Williamson, Leak et al) started by triode-connecting types like KT66 and 6L6 rather than using the (limited) triodes of the day, especially as they appeared to do the job admirably. When the advantages of distributed load appeared shortly afterwards that was about it, to my mind. This not to offend triode enthusiasts (I often used then myself in the 50s) - just as a comparison.
This perhaps somewhat off-thread, but a look from this angle does appear to me relevant. One final observation: I am uninspired by the apparent inconsistencies found in tubes of the same number these days. The Russians just seem to group vaguely similar types together, e.g. using the same innards for their KT66s and 6L6s. I have even picked up a 5Y3 (old directly heated rectifier) but with indirectly heated cathodes inside. What is the world coming to?
I would suggest that original intentions be kept in mind, however wide one's journey into application fields go (and I have done my share of wild applications! - did not work at a research organisation for nothing). The EL34 (as an example) was originally intended for use as a pentode, where G3 is there to control secondary emmision from the anode back to the screen, especially when the anode goes lower in voltage than G2 as in normal pentode practice. In that application it would not appear to have an influence on linearity, otherwise the manufacturers would have done something about it. It gives comparably low distortion as is, so the premise appears to be in order. With that outlook it seems illogical to expect that an improvement would result from taking G3 elsewhere than to common.
As I remember I have not seen too many data sheets for the EL34 (or 6CA7) as a triode, in contrast to beam tubes where they just about always occur. The triode mu is hardly ever given. All this might indicate that the manufacturers did not see the EL34 in use as a triode, although it was used as such. In the hey-day of tubes such applications made far more use of beam tubes, where there is nothing between G2 and anode to possibly upset things.
Perhaps there is a message in that.
Aleksandar, yes. Your argument is indeed logical regarding the use of "real" triodes. But then one must remember that power triode development seems to have stopped with the advent of pentodes and beam tubes. The latter was conveniently available in a range of models while triodes were not. (The ones you mention were developed mainly for regulator use etc.) So for me it was not unexpected that the original hi-fi generation (Williamson, Leak et al) started by triode-connecting types like KT66 and 6L6 rather than using the (limited) triodes of the day, especially as they appeared to do the job admirably. When the advantages of distributed load appeared shortly afterwards that was about it, to my mind. This not to offend triode enthusiasts (I often used then myself in the 50s) - just as a comparison.
This perhaps somewhat off-thread, but a look from this angle does appear to me relevant. One final observation: I am uninspired by the apparent inconsistencies found in tubes of the same number these days. The Russians just seem to group vaguely similar types together, e.g. using the same innards for their KT66s and 6L6s. I have even picked up a 5Y3 (old directly heated rectifier) but with indirectly heated cathodes inside. What is the world coming to?
I agree that "true triodes" will always be better triodes than pentodes imitating triodes. Most of my recent amplifiers have been triode based, varying in size from the 45 to the 833A.
Triode development may have stopped a long time ago, but there are still plenty of them around to play with. I think that some of the "boutique triodes" may be so popular because the sweet spot is so big that it is hard to build a bad sounding amp with them. Other than those there are plenty of triodes left that are waiting to be discovered. The problem lies when someone chats them up on one of the audio snob forums and the price triples overnight.
My experiences with "regulator triodes" (6336, 6AS7, 6080, 5998)has found them to be somewhat unstable in common cathode configuration. The current will drift considerably if used with fixed bias, and cathode bias wastes a lot of power.
My whole reason for the EL-34 experiments is not to build a good triode amp by using a pentode. I just want to try something that I have never done before. The curves that I posted here were baselines to determine the characteristics of the EL-34 for comparison purposes.
I have been trying "alternative connections" such as screen drive, G1 and G2 tied together, and G1, G2 and G3 tied together. I have seen some unexpected surprizes. Since it has been suggested that my tube may be defective (it is 45 years old), I have put these tests on hold until the new tubes arrive.
Ei tubes have become pretty scarce in the US recently. Fortunately I bought a bunch of 7189A's when they were plentiful.
Triode development may have stopped a long time ago, but there are still plenty of them around to play with. I think that some of the "boutique triodes" may be so popular because the sweet spot is so big that it is hard to build a bad sounding amp with them. Other than those there are plenty of triodes left that are waiting to be discovered. The problem lies when someone chats them up on one of the audio snob forums and the price triples overnight.
My experiences with "regulator triodes" (6336, 6AS7, 6080, 5998)has found them to be somewhat unstable in common cathode configuration. The current will drift considerably if used with fixed bias, and cathode bias wastes a lot of power.
My whole reason for the EL-34 experiments is not to build a good triode amp by using a pentode. I just want to try something that I have never done before. The curves that I posted here were baselines to determine the characteristics of the EL-34 for comparison purposes.
I have been trying "alternative connections" such as screen drive, G1 and G2 tied together, and G1, G2 and G3 tied together. I have seen some unexpected surprizes. Since it has been suggested that my tube may be defective (it is 45 years old), I have put these tests on hold until the new tubes arrive.
Ei tubes have become pretty scarce in the US recently. Fortunately I bought a bunch of 7189A's when they were plentiful.
CATHODE BIAS...
Hi to all, especially the usual suspects who follow what I have to say but stay in the shadows
My experiences with regulator triodes are wonderful -- especially with 6AS7/6080. Those tubes are utmost stable and pack a fenomenal punch you woud not expect from electron tubes!
Of course, I have used them in the power consuming cathode bias configuration, for the following reason:
1. IT SOUNDS WAY BETTER with all tubes and is considered to be prime quality sound when you have Rk below K...
2. It is clearly stated in the datasheets that those tubes should have Rk for at least 15% of the total bias, else instability will arise... with potential diode-type suicide results.
3. I have had some indirect experience with 6C33C (to be read 6S33S) -- and found fixed bias on these tubes to be a pain in the... decide for yourselves what would make you hurt the most.
Regards,
Aleksandar
Hi to all, especially the usual suspects who follow what I have to say but stay in the shadows
My experiences with regulator triodes are wonderful -- especially with 6AS7/6080. Those tubes are utmost stable and pack a fenomenal punch you woud not expect from electron tubes!
Of course, I have used them in the power consuming cathode bias configuration, for the following reason:
1. IT SOUNDS WAY BETTER with all tubes and is considered to be prime quality sound when you have Rk below K...
2. It is clearly stated in the datasheets that those tubes should have Rk for at least 15% of the total bias, else instability will arise... with potential diode-type suicide results.
3. I have had some indirect experience with 6C33C (to be read 6S33S) -- and found fixed bias on these tubes to be a pain in the... decide for yourselves what would make you hurt the most.
Regards,
Aleksandar
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.