I seem to think in that little single tube version I used small emitter resistors (100ohm) and no grid drive on the 6AS7 so it is a little harsher than my later versions. I feel just a little more oomph is needed for better tone since there is something about the loudness and feel when it gets slightly louder than shouting. So yeah the bigger ones tend to sound better. Most of my builds are with 3 tubes. Never really needed any more. They all get miked up anyways.
Kinda funny how I claim a very simple ckt in the video. I guess for an otl maybe, but for a 3watt amp a 6V6 SE job is lots simpler.
https://www.dropbox.com/s/d6rj9xa53d90fbk/IMG-20161211-WA0000.jpeg?dl=0
Well I want to start by congratulating this great thread and talk, it is the best I have read and the most focused on an interesting topic for many years.
Well where to start ...?
It's been a few years around the OTL for guitar. I have experience in it. But above all I tried to start with the simple thing, that there is a certain thing
(DNA of yeti, or to put together simple ideas to be able to find another idea simple or simple and to be able to see a solution along the way,) But all this must be seen from the position of the guitarist and his needs and preferences.
A guitar amp has another pedigree other than an audio or hifi. If we look from that particular point of the guitar everything is different. I want to clarify that I am and I consider myself a guitarist, I have built point to point many of the classic models of fender and Marshall, to name a few. And I have in my original salon and my reissues of all The Fender Champ 5f1 and 5e1, of the fender tweed deluxe, of the fender deluxe blackface of 1965 !!!! as well as replicas adapted from this in 4-5w to be able to have the same sound with the volume at 10 in the living room or in small spaces, marshall's, Vox, faustone, etc. We must keep in mind the best classical sounds of guitar to work on that basis, to be able to compare at any time and to arrive at reasonable conclusions and with luck and work (and research) to be able to give a good end and the site they deserve The best OTL amplifiers, yes and the guitar otl as well.
To encourage the conversation to say that in my opinion the power and more in guitar (except for applications for direct of course), is not important, the end is always sound quality with the simplest and the least cost, that is sleep Of any engineer who is that the product is good / nice and cheap (Good / nice and cheap = BBB in Spanish) In guitar also other factors come into play in addition to the quality / response or bandwidth, such as portability and weight.
Musicians are always transporting and carrying the equipment behind their backs, the weight and saving of an output transformer are things to keep in mind. And not only the weight of the output OT, but the weight of the speaker itself. We must see everything together.
Returning to the OTL theme, my idea of an ideal OTL amplifier is one that has a clean guitar sound as the best Fender sound, and that is handled perfectly with saturated sounds and / or effects pedals and saturation as the best Or more classic marshall when it comes to saturation. Not forgetting that you should get along with single guitar pickups or humbuckers, all of which make the subject very complex.
My development has gone these years in the sense of low and medium power, without further separating from the idea of the classical amp guitar, ie 0.5 watts to 5 watts. But always taking into account (and in comparison) the quality and simplicity.
It is not necessary a lot of saturation in a guitar amplifier although it is of quality and audible stability denoted or linear as welcome.
But neither is a problem well addressed to the instrument. I am glad to see so many contributions of high level and proven reputation, it is clear that here is debated on the alchemy itself of the amplifiers at the highest level.
Although for me as an old guitarist would say the ear configures what the idea and the circuit look for. A guitar amp if you have a clean huge and good, with punching and sensitive to the execution of the guitarist and you will have everything.
My OTLs are for guitar and mostly use high impedance speakers and real direct output, say about 600/800 ohms The surprise is that they do not need a tone control or ToneStack. They have a generous bandwidth and only sound off when adding an output transformer OT with standard speaker. I if I have done the test of ON / OFF with trafo output and without trafo of output in my amplifiers and with several marks of ot like hammond, melodial, etc, By simply pressing a switch it is switched to without trafo (otl) My conclusions are that a simple circuit, of whatever kind that includes output teafo is usually quite dark in equalization and tonal response. It would appear as if the tone stack would have been born to correct the pitch losses or frequency or response cuts of the output transforms.
This does not happen with a simple concept OTL such as a circuit with two pentodes where these do not have to be paired or identical, or type bi-ampli., Where with a simple but incredibly effective prior fender type of a single tube and Few components take advantage of a less pure OTL, with direct output to 800 ohm speaker and regardless of any "ToneStack" because it does not need it
Summary: *OTL sound more than balanced and pleasing to the ear, almost three-dimensional with the 600-ohm speaker. With a clean grandiose and very dynamic at the touch of pua. *We switched to with OT and 8ohms with a simple switch and various high quality and well known trams and the bland sound without grace and off. The magic of simplicity with the OT is lost. As many pointed out I contrast that the OTL for guitar can reach a level as high or more in quality and performance than the OTL of hifi, without oscillations or audible roughness, that can be carried well with guitars type single or humbuckees, with a Clean that fascinates and swallows pedals effects very well. I have assembled with NOS valves of the best quality, with Hammond transformers of quality and with classic components for amplifiers of valves, in point-to-point circuits with ptp made and riveted by hand in cardboard type fender. (I.e. I have a 1954 metal cap siemmens of 1955 metal hopper waiting for the tests of a somewhat higher power range ;-) I have also done satisfactory tests with standard guitar cones of 16 ohms in series in 2x4 and 4x4, but for experimentation I like the direct output and without stacking tubes ... my amplifiers are only 1 or 2 valves in the stage Power and 800 ohm high power. And I am free of commercial inclination since I do not sell commercially (for now )
https://www.dropbox.com/s/rz2igxx5envcoa1/IMG_20170109_224517.jpg?dl=0
https://www.dropbox.com/s/jod6c2yflnodqxn/IMG_20170112_133112.JPG?dl=0
https://www.dropbox.com/s/ics89mk9a3vo5kd/IMG_20170126_074258.jpg?dl=0
FMK-madid-spain
Well I want to start by congratulating this great thread and talk, it is the best I have read and the most focused on an interesting topic for many years.
Well where to start ...?
It's been a few years around the OTL for guitar. I have experience in it. But above all I tried to start with the simple thing, that there is a certain thing
(DNA of yeti, or to put together simple ideas to be able to find another idea simple or simple and to be able to see a solution along the way,) But all this must be seen from the position of the guitarist and his needs and preferences.A guitar amp has another pedigree other than an audio or hifi. If we look from that particular point of the guitar everything is different. I want to clarify that I am and I consider myself a guitarist, I have built point to point many of the classic models of fender and Marshall, to name a few. And I have in my original salon and my reissues of all The Fender Champ 5f1 and 5e1, of the fender tweed deluxe, of the fender deluxe blackface of 1965 !!!!
as well as replicas adapted from this in 4-5w to be able to have the same sound with the volume at 10 in the living room or in small spaces, marshall's, Vox, faustone, etc. We must keep in mind the best classical sounds of guitar to work on that basis, to be able to compare at any time and to arrive at reasonable conclusions and with luck and work (and research) to be able to give a good end and the site they deserve The best OTL amplifiers, yes and the guitar otl as well. To encourage the conversation to say that in my opinion the power and more in guitar (except for applications for direct of course), is not important, the end is always sound quality with the simplest and the least cost, that is sleep Of any engineer who is that the product is good / nice and cheap (Good / nice and cheap = BBB in Spanish
) In guitar also other factors come into play in addition to the quality / response or bandwidth, such as portability and weight.Musicians are always transporting and carrying the equipment behind their backs, the weight and saving of an output transformer are things to keep in mind. And not only the weight of the output OT, but the weight of the speaker itself. We must see everything together.
Returning to the OTL theme, my idea of an ideal OTL amplifier is one that has a clean guitar sound as the best Fender sound, and that is handled perfectly with saturated sounds and / or effects pedals and saturation as the best Or more classic marshall when it comes to saturation. Not forgetting that you should get along with single guitar pickups or humbuckers, all of which make the subject very complex.
My development has gone these years in the sense of low and medium power, without further separating from the idea of the classical amp guitar, ie 0.5 watts to 5 watts. But always taking into account (and in comparison) the quality and simplicity.
It is not necessary a lot of saturation in a guitar amplifier although it is of quality and audible stability denoted or linear as welcome.
But neither is a problem well addressed to the instrument. I am glad to see so many contributions of high level and proven reputation, it is clear that here is debated on the alchemy itself of the amplifiers at the highest level.
Although for me as an old guitarist would say the ear configures what the idea and the circuit look for. A guitar amp if you have a clean huge and good, with punching and sensitive to the execution of the guitarist and you will have everything.
My OTLs are for guitar and mostly use high impedance speakers and real direct output, say about 600/800 ohms The surprise is that they do not need a tone control or ToneStack. They have a generous bandwidth and only sound off when adding an output transformer OT with standard speaker. I if I have done the test of ON / OFF with trafo output and without trafo of output in my amplifiers and with several marks of ot like hammond, melodial, etc, By simply pressing a switch it is switched to without trafo (otl) My conclusions are that a simple circuit, of whatever kind that includes output teafo is usually quite dark in equalization and tonal response. It would appear as if the tone stack would have been born to correct the pitch losses or frequency or response cuts of the output transforms.
This does not happen with a simple concept OTL such as a circuit with two pentodes where these do not have to be paired or identical, or type bi-ampli., Where with a simple but incredibly effective prior fender type of a single tube and Few components take advantage of a less pure OTL, with direct output to 800 ohm speaker and regardless of any "ToneStack" because it does not need it
Summary: *OTL sound more than balanced and pleasing to the ear, almost three-dimensional with the 600-ohm speaker. With a clean grandiose and very dynamic at the touch of pua. *We switched to with OT and 8ohms with a simple switch and various high quality and well known trams and the bland sound without grace and off. The magic of simplicity with the OT is lost. As many pointed out I contrast that the OTL for guitar can reach a level as high or more in quality and performance than the OTL of hifi, without oscillations or audible roughness, that can be carried well with guitars type single or humbuckees, with a Clean that fascinates and swallows pedals effects very well. I have assembled with NOS valves of the best quality, with Hammond transformers of quality and with classic components for amplifiers of valves, in point-to-point circuits with ptp made and riveted by hand in cardboard type fender. (I.e. I have a 1954 metal cap siemmens of 1955 metal hopper waiting for the tests of a somewhat higher power range ;-) I have also done satisfactory tests with standard guitar cones of 16 ohms in series in 2x4 and 4x4, but for experimentation I like the direct output and without stacking tubes ... my amplifiers are only 1 or 2 valves in the stage Power and 800 ohm high power. And I am free of commercial inclination since I do not sell commercially (for now
)https://www.dropbox.com/s/rz2igxx5envcoa1/IMG_20170109_224517.jpg?dl=0
https://www.dropbox.com/s/jod6c2yflnodqxn/IMG_20170112_133112.JPG?dl=0
https://www.dropbox.com/s/ics89mk9a3vo5kd/IMG_20170126_074258.jpg?dl=0
FMK-madid-spain
Last edited:
Interesting thread *multi-volti* --
Its kinda funny -- I used to post alot on this site (lurk mostly now) but my interests are guitar amps. In particular the exact tone that *atmasphere* describes: heavy 2nd order dis, mild clipping distorion, clean but fat, etc directed toward a hollowbody type guitar. I have been wondering awhile if a push-pull or self-split 6asy might deliver it.
I am nowhere near qualified on the whole OTL thing, but to be honest I don't care about topology that much. To me its all about tone.
And there is another point to make. Something along the lines of the old mountain climber chantey. Hey I got iron I think might work (need some help there ergo, here I am hat in hand), got the tubes ... why build it? Because I CAN (with help -- the rub) because of simple boyish interest in hearing how it *might* sound. Heck 99 94/100% of the guitar amps I acquire get cannabilized or sold anyhow so failure is to be*expected* in this tone quest stuff.
The other amusing part is the response of the local experts to the whole idea of a 6AS7 guitar amp (spoken with awe and true respect for folks who actually *know* what they are doing - no kidding) on a largely hifi forum (here) vs. a largely guitar amp forum. The guitar amp folks can't get past the current consumed by the 6AS7. It not a feasible tube for audio even though Antek 1Txxx PTs offer 6 amps 6.3v (2 taps) for about $35 and that, incidentally, get routinely used for guitar amps (the second 6.3 volt tap ignored). And besides - triodes are not for guitar amps - although AX84 certainly has some triode guitar amps that people seem to like.
For some hifi folks guitar amp discussions summon only KISS-like guitar tones and players. I don't have a need to sound like that -- its no challenge at all -- just buy a second-hand Valve Jr and crank'er up -- note definition? Heck note definition only reveals your poor play (I am NOT a great guitar player by any means but I like to hear what I play). Think about how Eric Clapton chose to sound in his 60's (heavy 2nd order dis, mild clipping distorion, clean but fat). Well -- I'm in my 60s too and want to sound like that ("fat" chance LOL)!! Fortunately for me no one cares if I have evolved past "White room" I guess (which I love too) as a artist (for another thread).
Sorry for the rant and over long post. Back to the shadows for me. However, *multi-volti* I am watching and rooting for you. If you decide to move this discussion (for whatever reason) - PLEASE leave a forwarding address!! Thanks.
Its kinda funny -- I used to post alot on this site (lurk mostly now) but my interests are guitar amps. In particular the exact tone that *atmasphere* describes: heavy 2nd order dis, mild clipping distorion, clean but fat, etc directed toward a hollowbody type guitar. I have been wondering awhile if a push-pull or self-split 6asy might deliver it.
I am nowhere near qualified on the whole OTL thing, but to be honest I don't care about topology that much. To me its all about tone.
And there is another point to make. Something along the lines of the old mountain climber chantey. Hey I got iron I think might work (need some help there ergo, here I am hat in hand), got the tubes ... why build it? Because I CAN (with help -- the rub) because of simple boyish interest in hearing how it *might* sound. Heck 99 94/100% of the guitar amps I acquire get cannabilized or sold anyhow so failure is to be*expected* in this tone quest stuff.
The other amusing part is the response of the local experts to the whole idea of a 6AS7 guitar amp (spoken with awe and true respect for folks who actually *know* what they are doing - no kidding) on a largely hifi forum (here) vs. a largely guitar amp forum. The guitar amp folks can't get past the current consumed by the 6AS7. It not a feasible tube for audio even though Antek 1Txxx PTs offer 6 amps 6.3v (2 taps) for about $35 and that, incidentally, get routinely used for guitar amps (the second 6.3 volt tap ignored). And besides - triodes are not for guitar amps - although AX84 certainly has some triode guitar amps that people seem to like.
For some hifi folks guitar amp discussions summon only KISS-like guitar tones and players. I don't have a need to sound like that -- its no challenge at all -- just buy a second-hand Valve Jr and crank'er up -- note definition? Heck note definition only reveals your poor play (I am NOT a great guitar player by any means but I like to hear what I play). Think about how Eric Clapton chose to sound in his 60's (heavy 2nd order dis, mild clipping distorion, clean but fat). Well -- I'm in my 60s too and want to sound like that ("fat" chance LOL)!! Fortunately for me no one cares if I have evolved past "White room" I guess (which I love too) as a artist (for another thread).
Sorry for the rant and over long post. Back to the shadows for me. However, *multi-volti* I am watching and rooting for you. If you decide to move this discussion (for whatever reason) - PLEASE leave a forwarding address!! Thanks.
Last edited:
In first position and time , see how it play and song:
http://youtu.be/7LsjpRNZl2Y
http://youtu.be/Zz4GG6HkpuQ
I not very good guitar player profesional, but one sample of song's is good.
http://youtu.be/7LsjpRNZl2Y
http://youtu.be/Zz4GG6HkpuQ
I not very good guitar player profesional, but one sample of song's is good.
Ok and thanks Moonbird.
I move in my projects in 3 lines or directions.
1- micro amp wit only 0,5w aprox. And whit low voltage real in power, whit 12volts in all, but not infravoltage in power valves.
, this valves are of double impulse and run ok at 12v pentode. This is the valves N.O.S. of low voltage, range 9 to 30v. My fa
In this is necesary one boster guitar for move some, the speaker is 700 ohms, i love BSIAB for this and impulse one point.
2 - one Amp more power in 4,5-5 watts.
Whit two valves, but not push pull, is more one circuit type bi-amp, and not neccesary match tubes, and can be diferent tubes in one or other positión.
For this two tubes the +B is 300v, and preamp +300v.
Z out whit two valves is = at 700/800 ohms, and cone is one philips 700 ohms
Vintage, in the video of youtube the cone not is philips.
One note important: the cone philips and not philips only of 6" songs very good and much more good compared whit my CELESTION greenback 12" G12H(blackback) of the year 1970.........
This circuit of 5w is based in power clasic Otl of other decades, and very hight cuality demostrate. The preAmp is one pream whit one only part of ecc82 or ecc83....1/2 eccXX stile fender absolutely whit imput hight impedance estandar for guitars imput, type fender but more simple whit one only stage of amplificatión.
Whit only the basic components in line of the signal for one pure tone how this is posible.
3 - similar circuit of 2 part, but whit more power, in adaptance standar tube EL34 for one more power amp guitar for direct and scenary. One friend claim for this jiji
The idea is migrate in this at speaker standar in 2x10" or 4x10" or 12".
Ane comparate this whit speaker Higth Z an standars.
Others experiments whit one tube for preamp and power at 2w for home in estudy for higth and low voltage.
I move in my projects in 3 lines or directions.
1- micro amp wit only 0,5w aprox. And whit low voltage real in power, whit 12volts in all, but not infravoltage in power valves.
, this valves are of double impulse and run ok at 12v pentode. This is the valves N.O.S. of low voltage, range 9 to 30v. My fa
In this is necesary one boster guitar for move some, the speaker is 700 ohms, i love BSIAB for this and impulse one point.
2 - one Amp more power in 4,5-5 watts.
Whit two valves, but not push pull, is more one circuit type bi-amp, and not neccesary match tubes, and can be diferent tubes in one or other positión.
For this two tubes the +B is 300v, and preamp +300v.
Z out whit two valves is = at 700/800 ohms, and cone is one philips 700 ohms
Vintage, in the video of youtube the cone not is philips.
One note important: the cone philips and not philips only of 6" songs very good and much more good compared whit my CELESTION greenback 12" G12H(blackback) of the year 1970.........
This circuit of 5w is based in power clasic Otl of other decades, and very hight cuality demostrate. The preAmp is one pream whit one only part of ecc82 or ecc83....1/2 eccXX stile fender absolutely whit imput hight impedance estandar for guitars imput, type fender but more simple whit one only stage of amplificatión.
Whit only the basic components in line of the signal for one pure tone how this is posible.
3 - similar circuit of 2 part, but whit more power, in adaptance standar tube EL34 for one more power amp guitar for direct and scenary. One friend claim for this jiji
The idea is migrate in this at speaker standar in 2x10" or 4x10" or 12".
Ane comparate this whit speaker Higth Z an standars.
Others experiments whit one tube for preamp and power at 2w for home in estudy for higth and low voltage.
Moonbird, yes the sound is clean whit punch, open, more 3D, whit equilibrate tone whit only one pot of volumen, and good when this saturate.
This in clean song and saturate more good for my which other fenders i reference and clasic reputed tone.
Probe some similar and see this, is very easy and single of contruct.
This in clean song and saturate more good for my which other fenders i reference and clasic reputed tone.
Probe some similar and see this, is very easy and single of contruct.
Atty - Hmm thats interesting ... makes it kinda low voltage for the preamp tubes tho, right. What preamp tubes work with that voltage or would this amp require a "backwards" PS (i.e, premap taps 1st). I am thinking a pentode input tube might be a nice addition given the triode power stage ...
Wouldn't matter to an OTL amp as much I guess because there is no leadoff OT quashing ripple anyway. Would adding a PS choke offset some of the OTL sound benefits? Seems like it might tighten the bass a bit - might be good. It would need to be a big one
-- think I recall a 150mA Hammond choke sleeping on a shelf somewhere around the shop ... hmmm. SemperFi, thanks, this combo and circuit are all point to point (ptp style fender ) and hand made, and recicle some componets NOS but in new estate.
yes it'a amps have one special clean sound, and whithout toneStack, only one potentiometer volume, only, but very balanced tones, and all here.
Yes i saturate this whit pedal, it is correct, in next time probe more saturate in circuit.
But volume at 100% its is more cool.
Not, the saturate whit pedals not is hard, in general, is very stable, and whithout oscilatións, the note end in good progresión. And good sustain in saturations
The sound in sample saturate i put in this site is record whit 4,5w / 2 valve bi-amp and pedal BSIB2 ("brown sound in a box") of kit GGG page, similar to MXR van hallen.
Its is my preference, but good whit pedals in generic form.
And note whit 600/800 ohms speaker can send to thid more 3D an big sound.
I repeat it how data interesting, the sound of my 6" speakers 700 ohms in this amp is more much big and beautiful.
and in loudnes (bass), in all range or spectre.
In my celestions 12GH Blackbacks of 1970 "not is identical" whit OT an whithout OT is more insulse, or less dinamic?
Its are one good reference in comparation...
Only one stage preamp, sounds more cleans but very good y armonics presents in cleand And saturate sound, easy appears armonics in clean whit touch the guitar.
yes it'a amps have one special clean sound, and whithout toneStack, only one potentiometer volume, only, but very balanced tones, and all here.
Yes i saturate this whit pedal, it is correct, in next time probe more saturate in circuit.
But volume at 100% its is more cool.
Not, the saturate whit pedals not is hard, in general, is very stable, and whithout oscilatións, the note end in good progresión. And good sustain in saturations
The sound in sample saturate i put in this site is record whit 4,5w / 2 valve bi-amp and pedal BSIB2 ("brown sound in a box") of kit GGG page, similar to MXR van hallen.
Its is my preference, but good whit pedals in generic form.
And note whit 600/800 ohms speaker can send to thid more 3D an big sound.
I repeat it how data interesting, the sound of my 6" speakers 700 ohms in this amp is more much big and beautiful.
and in loudnes (bass), in all range or spectre.
In my celestions 12GH Blackbacks of 1970 "not is identical" whit OT an whithout OT is more insulse, or less dinamic?
Its are one good reference in comparation...
Only one stage preamp, sounds more cleans but very good y armonics presents in cleand And saturate sound, easy appears armonics in clean whit touch the guitar.
Last edited:
Atty - Hmm thats interesting ... makes it kinda low voltage for the preamp tubes tho, right. What preamp tubes work with that voltage or would this amp require a "backwards" PS (i.e, premap taps 1st). I am thinking a pentode input tube might be a nice addition given the triode power stage ...
Wouldn't matter to an OTL amp as much I guess because there is no leadoff OT quashing ripple anyway. Would adding a PS choke offset some of the OTL sound benefits? Seems like it might tighten the bass a bit - might be good. It would need to be a big one
The output tubes need some current! I usually build a separate power supply for the driver and preamp tubes.
This thread is very interesting to me, partly because my notebooks have contained sketches of triode valves feeding PNP current mirrors for several years now.
I came to that idea because I realized that, with a common-cathode triode gain stage, the lower the load resistance, the more second harmonic distortion you get. (See figure 2, attached.)
In figure 2, the middle dot on each load-line is the operating point (quiescent current and voltage). The two end dots represent positive and negative peaks; grid voltage swings are equal (look at the grid curve numbers), but the resulting current and voltage swings at the anode are not equal. The more the asymmetry between positive and negative half-cycles, the larger the amount of second harmonic distortion.
I've estimated this distortion from the graph, and the estimated second harmonic distortion numbers are at the bottom of each image.
Okay, low load resistance is good for lots of second harmonic distortion, which we want for a guitar amp. The lowest load resistance you can have is zero ohms (figure 4, constant voltage source load). But zero ohm loads produce no voltage swing across them, so you also have zero voltage gain.
An amplification stage with no (voltage) amplification isn't very useful, so how can we get around that?
The obvious answer was to use a current mirror. It acts like a nearly zero-ohm load to the triode, so it produces lots of that lovely "tubey" second harmonic distortion. But the current in the second arm of the current mirror can be dumped into a normal-value load resistance (R3 in the attached image), producing about the same voltage gain as a "normal" common-cathode gain stage. Best of both worlds! (See attached figure "triode w current mirror".)
That idea is very, very close to parts of SemperFi's schematic in post #54, the only real difference being that SemperFi is using a long-tailed-pair rather than a single common-cathode triode stage. And, of course, SemperFi actually built the thing and showed that it works, while my ideas stayed on the pages of my notebooks.
I'll continue the discussion in a later post. For now, here's the raw current-mirror idea, along with some of the load-line sketches that led me to it.
Note that a constant current source load is the highest load impedance you can get, and therefore has the opposite effect, producing very little second harmonic distortion from a triode (fig_3_CCS).
Not surprisingly, a CCS load is frequently used in modern valve Hi-Fi preamp stages, where designers are attempting to produce low levels of distortion.
-Gnobuddy
I came to that idea because I realized that, with a common-cathode triode gain stage, the lower the load resistance, the more second harmonic distortion you get. (See figure 2, attached.)
In figure 2, the middle dot on each load-line is the operating point (quiescent current and voltage). The two end dots represent positive and negative peaks; grid voltage swings are equal (look at the grid curve numbers), but the resulting current and voltage swings at the anode are not equal. The more the asymmetry between positive and negative half-cycles, the larger the amount of second harmonic distortion.
I've estimated this distortion from the graph, and the estimated second harmonic distortion numbers are at the bottom of each image.
Okay, low load resistance is good for lots of second harmonic distortion, which we want for a guitar amp. The lowest load resistance you can have is zero ohms (figure 4, constant voltage source load). But zero ohm loads produce no voltage swing across them, so you also have zero voltage gain.
An amplification stage with no (voltage) amplification isn't very useful, so how can we get around that?
The obvious answer was to use a current mirror. It acts like a nearly zero-ohm load to the triode, so it produces lots of that lovely "tubey" second harmonic distortion. But the current in the second arm of the current mirror can be dumped into a normal-value load resistance (R3 in the attached image), producing about the same voltage gain as a "normal" common-cathode gain stage. Best of both worlds! (See attached figure "triode w current mirror".)
That idea is very, very close to parts of SemperFi's schematic in post #54, the only real difference being that SemperFi is using a long-tailed-pair rather than a single common-cathode triode stage. And, of course, SemperFi actually built the thing and showed that it works, while my ideas stayed on the pages of my notebooks.
I'll continue the discussion in a later post. For now, here's the raw current-mirror idea, along with some of the load-line sketches that led me to it.
Note that a constant current source load is the highest load impedance you can get, and therefore has the opposite effect, producing very little second harmonic distortion from a triode (fig_3_CCS).
Not surprisingly, a CCS load is frequently used in modern valve Hi-Fi preamp stages, where designers are attempting to produce low levels of distortion.
-Gnobuddy
Attachments
Okay, now we have some reason to think that a current mirror might actually be a way to get a good clean sound (i.e., lots of second-harmonic distortion) out of a triode stage.
But the primary focus of this thread wasn't about current mirrors, it was about eliminating output transformers.
But those two ideas (current mirrors, and OTL) can actually be connected! At this point, we know valve OTL guitar amps can sound good. But valves really struggle to deliver the huge currents that speakers need. Transistors, on the other hand, are wonderful at delivering large currents to speakers, but transistor-based guitar amps don't sound as good as valve ones.
So: what if we use a current mirror, arranged to force an output transistor to exactly duplicate the current in the output valve? Now the transistor should sound exactly like the valve.
And what if we add the transistor current to the valve current? Now we have twice as much current as one valve can deliver. Four times the audio power to the speaker, without doubling the number of output valves.
And what if we can modify the current mirror so that the second branch not only mirrors, but also multiplies, the current from the output valve? Then we could deliver triple, or quadruple, the current to the speaker, and it would still sound exactly like the output valve, because of the current-mirror action that makes the transistor follow exactly what the valve does.
There are a few ways to make a current mirror that also multiplies current. One way is to use transistors with unequal (semiconductor) junction areas. The bigger junction-area transistor will flow more current than the smaller one.
A similar effect occurs if you parallel two or three (or more) identical transistors. If they're identical, you've now doubled or tripled the total collector current, for the same base-emitter voltage.
Unfortunately, identical power transistors are not easy to find, and also cost money.
There is a simpler (and somewhat less exact) way. Add a small emitter resistor to each transistor in the current mirror. Make one resistor smaller than the other. The corresponding transistor will flow a larger current.
This idea is shown in the attached image. I have not built and tried it out. So I don't know how well it works (yet).
This idea is not as elegantly perfect as the PNP current mirror connected to the anode. In this circuit, the voltages across Q2 & R1 do interfere a bit with the voltage at the cathode of the pentode, so they will, to some degree, alter the current through the pentode.
However, the voltage across Q2 is nearly a constant 0.7 volts or so, quite small compared to a typical output valve bias voltage (often twenty volts or so). Similarly, R1 can be chosen pretty small, so that only a few hundred millivolts are dropped across it.
So I think the circuit can be built with component values that pretty much leave the pentode alone, only minimally altering its transfer characteristic. That means it will still sound like a pentode.
One more thing - while I've shown the pentode's anode connected to the transistor collector, there is no need to actually do this; the transistor could be run off a much lower voltage B+ rail than the pentode. This would reduce problems with second-breakdown of the transistor.
Ironically, my idea would involve removing the output transformer from the valve - but adding an output transformer to the transistor, to get the power actually to the speaker!
But the potential benefit would be easy delivery of large amounts of power, courtesy of the transistors, while maintaining the all-important "tube sound" or "valve sound" that we want in our guitar amps.
-Gnobuddy
But the primary focus of this thread wasn't about current mirrors, it was about eliminating output transformers.
But those two ideas (current mirrors, and OTL) can actually be connected! At this point, we know valve OTL guitar amps can sound good. But valves really struggle to deliver the huge currents that speakers need. Transistors, on the other hand, are wonderful at delivering large currents to speakers, but transistor-based guitar amps don't sound as good as valve ones.
So: what if we use a current mirror, arranged to force an output transistor to exactly duplicate the current in the output valve? Now the transistor should sound exactly like the valve.
And what if we add the transistor current to the valve current? Now we have twice as much current as one valve can deliver. Four times the audio power to the speaker, without doubling the number of output valves.
And what if we can modify the current mirror so that the second branch not only mirrors, but also multiplies, the current from the output valve? Then we could deliver triple, or quadruple, the current to the speaker, and it would still sound exactly like the output valve, because of the current-mirror action that makes the transistor follow exactly what the valve does.
There are a few ways to make a current mirror that also multiplies current. One way is to use transistors with unequal (semiconductor) junction areas. The bigger junction-area transistor will flow more current than the smaller one.
A similar effect occurs if you parallel two or three (or more) identical transistors. If they're identical, you've now doubled or tripled the total collector current, for the same base-emitter voltage.
Unfortunately, identical power transistors are not easy to find, and also cost money.
There is a simpler (and somewhat less exact) way. Add a small emitter resistor to each transistor in the current mirror. Make one resistor smaller than the other. The corresponding transistor will flow a larger current.
This idea is shown in the attached image. I have not built and tried it out. So I don't know how well it works (yet).
This idea is not as elegantly perfect as the PNP current mirror connected to the anode. In this circuit, the voltages across Q2 & R1 do interfere a bit with the voltage at the cathode of the pentode, so they will, to some degree, alter the current through the pentode.
However, the voltage across Q2 is nearly a constant 0.7 volts or so, quite small compared to a typical output valve bias voltage (often twenty volts or so). Similarly, R1 can be chosen pretty small, so that only a few hundred millivolts are dropped across it.
So I think the circuit can be built with component values that pretty much leave the pentode alone, only minimally altering its transfer characteristic. That means it will still sound like a pentode.
One more thing - while I've shown the pentode's anode connected to the transistor collector, there is no need to actually do this; the transistor could be run off a much lower voltage B+ rail than the pentode. This would reduce problems with second-breakdown of the transistor.
Ironically, my idea would involve removing the output transformer from the valve - but adding an output transformer to the transistor, to get the power actually to the speaker!
But the potential benefit would be easy delivery of large amounts of power, courtesy of the transistors, while maintaining the all-important "tube sound" or "valve sound" that we want in our guitar amps.
-Gnobuddy
Attachments
- Status
- Not open for further replies.