You can calculate it from:
1/(2 * pi * Cout * ra)
In most cases Cout is around 10pF, and ra varies from say 15k for an EL34 to 50k for a 6V6, so you'd be looking at flat responses up to 1MHz and 318kHz respectively. Multiply by gmra and the gain-bandwidth products would be about 160MHz and 63MHz respectively. Like Miles said, what exactly are you asking?
1/(2 * pi * Cout * ra)
In most cases Cout is around 10pF, and ra varies from say 15k for an EL34 to 50k for a 6V6, so you'd be looking at flat responses up to 1MHz and 318kHz respectively. Multiply by gmra and the gain-bandwidth products would be about 160MHz and 63MHz respectively. Like Miles said, what exactly are you asking?
Last edited:
Apologies for the open ended question, I meant for audio / guitar amp applications.
People often describe each tube in terms of what they hear. I would simply like to see how they reproduce the audio frequency spectrum; much like a speaker 20hz-20khz SPL chart.
Of course these properties get thrown out the window once the output tubes are pushed into overdrive, but my uses are mostly in the clean / non-clipping region.
People often describe each tube in terms of what they hear. I would simply like to see how they reproduce the audio frequency spectrum; much like a speaker 20hz-20khz SPL chart.
Of course these properties get thrown out the window once the output tubes are pushed into overdrive, but my uses are mostly in the clean / non-clipping region.
True, some people describe different tube types in terms of what they think they hear.
Some people even claim that different tube brands sound different. Or even diffrent examples of the same type in the same brand.
Whole textbooks and many papers in learned journals have been written by electronics engineers and psychologists about why people think there are differences in things where there is in fact no difference at all. People imagining non-existent or reversed differences is a major problem in the drug industry and seriously frustrates development and optimisation of medical treatment.
There is a well known difference, easily demonstrated by instruments and known about by electronic engineeers since tetrodes were invented in the 1930's between triodes and tetrodes and pentodes. Pentodes have slightly higher distortion compared to beam tetrodes intended for audio, but the difference is so slight that in various circuits the difference is often reversed.
For tetrodes and pentodes, the sound doesn't inherently depend on the tube. it depends on teh circuit and the quality and type of output transformer.
All the types you nominated are tetrodes (6V6, 6L6) or pentodes (EL34, EL84). However these particular pentodes have higher gain than these particular tetrodes.
There is no intrinsic diffrence in the sound of an EL34/6CA7 compared to the 6L6 for example, but the EL34 has much higher control grid gain (11 mA/V vs 6 mA/V) and different screen characteristics. So, if you start with an amplifier designed for 6L6, and drop in an EL34, the amplifier may become unstable. You hear the instability as distortion. But an amplifier intended to take a EL34 won't be unstable and will sound the same - other aspects (eg transformer design) being equal.
If the amplifier is an ultralinear type there is a measurable (with instruments) difference with 6L6 and EL34 in the same amplifier due to the different screen characteristics. It is quite unlikely that anyone's ears can tell the difference. And the slight difference that is detectable on instruments can be easily reduced or eliminated by adjustment of the transformer screen tapping point.
The human ear cannot detect small changes in volume (1 dB or less) as a change in volume. But we DO hear the change - as a change in bass and clarity. Quite a number of journal articles have appeared that describe how double blind auditioning has identified a diffrence in clarity, bass, or treble, investigation by engineers shown that there was a diffrence in volume. When this was corrected, the auditioners could not detect any difference - even when differences in distortion beween amplifiers were easily detected on instruments.
Due to the peaky nature of music, all but the highest power amplifiers in audio service are routinely overdriven momentarily in any musical selection, unless reproducing vinyl or radio braodcasts, which are compressed. With CD's, even when played at moderate volume for personal listening, momentary overdroves are routine. Even with vinyl, overdrives occaisonally occur except for listening a quite levels.
The ear cannot easily detect occaisonal momentary overloads. Some people cannot detect it at all. (We can all, of course, easily detect overdrive once it occurs continuously.) However, less than perfect amplifier circuit design can result in an amplifier than does not instantly recover from an overdrive, and remains parallised for a short time after. This parallysis is easilty detectable and sounds horrible. A type of overdrive response called re-entrant clipping, once not uncommon in solid state amplifiers, also sounds bad.
Make no mistake. The behavior of audio amplifiers (both tube and solid state) under overdrive is VERY important and is a major reason why some sound a lot better than others, and why some sound different than others. Overdrive behaviour is determined by circuit design, not the tube type. Unless of course the tube is worn out or faulty.
Some people even claim that different tube brands sound different. Or even diffrent examples of the same type in the same brand.
Whole textbooks and many papers in learned journals have been written by electronics engineers and psychologists about why people think there are differences in things where there is in fact no difference at all. People imagining non-existent or reversed differences is a major problem in the drug industry and seriously frustrates development and optimisation of medical treatment.
There is a well known difference, easily demonstrated by instruments and known about by electronic engineeers since tetrodes were invented in the 1930's between triodes and tetrodes and pentodes. Pentodes have slightly higher distortion compared to beam tetrodes intended for audio, but the difference is so slight that in various circuits the difference is often reversed.
For tetrodes and pentodes, the sound doesn't inherently depend on the tube. it depends on teh circuit and the quality and type of output transformer.
All the types you nominated are tetrodes (6V6, 6L6) or pentodes (EL34, EL84). However these particular pentodes have higher gain than these particular tetrodes.
There is no intrinsic diffrence in the sound of an EL34/6CA7 compared to the 6L6 for example, but the EL34 has much higher control grid gain (11 mA/V vs 6 mA/V) and different screen characteristics. So, if you start with an amplifier designed for 6L6, and drop in an EL34, the amplifier may become unstable. You hear the instability as distortion. But an amplifier intended to take a EL34 won't be unstable and will sound the same - other aspects (eg transformer design) being equal.
If the amplifier is an ultralinear type there is a measurable (with instruments) difference with 6L6 and EL34 in the same amplifier due to the different screen characteristics. It is quite unlikely that anyone's ears can tell the difference. And the slight difference that is detectable on instruments can be easily reduced or eliminated by adjustment of the transformer screen tapping point.
The human ear cannot detect small changes in volume (1 dB or less) as a change in volume. But we DO hear the change - as a change in bass and clarity. Quite a number of journal articles have appeared that describe how double blind auditioning has identified a diffrence in clarity, bass, or treble, investigation by engineers shown that there was a diffrence in volume. When this was corrected, the auditioners could not detect any difference - even when differences in distortion beween amplifiers were easily detected on instruments.
Due to the peaky nature of music, all but the highest power amplifiers in audio service are routinely overdriven momentarily in any musical selection, unless reproducing vinyl or radio braodcasts, which are compressed. With CD's, even when played at moderate volume for personal listening, momentary overdroves are routine. Even with vinyl, overdrives occaisonally occur except for listening a quite levels.
The ear cannot easily detect occaisonal momentary overloads. Some people cannot detect it at all. (We can all, of course, easily detect overdrive once it occurs continuously.) However, less than perfect amplifier circuit design can result in an amplifier than does not instantly recover from an overdrive, and remains parallised for a short time after. This parallysis is easilty detectable and sounds horrible. A type of overdrive response called re-entrant clipping, once not uncommon in solid state amplifiers, also sounds bad.
Make no mistake. The behavior of audio amplifiers (both tube and solid state) under overdrive is VERY important and is a major reason why some sound a lot better than others, and why some sound different than others. Overdrive behaviour is determined by circuit design, not the tube type. Unless of course the tube is worn out or faulty.
Last edited:
Thank you very much for this answer. That is something I had always assumed was part of the equation, but never read it anywhere.
Basically, a circuit designed for a 6l6 will sound like it has more treble or presence with el34 because the circuit has now changed? But if two amps were each optimized for each tube it would be very difficult to tell the difference?
Basically, a circuit designed for a 6l6 will sound like it has more treble or presence with el34 because the circuit has now changed? But if two amps were each optimized for each tube it would be very difficult to tell the difference?
The Frequency response is dependent on a number of parameters...
Not entirely the tube, but each of the tubes you mention can have the same frequency response if the circuit is optimized for a given set of requirements...
Guitar amps are a bit more non-linear.... since you work primarily in square wave most of the time.... and square wave in the output stage spend 98% of their time in the "triode" region of the plate curves, left of the knee...This drops the effective plate resistance down to 400 to 800 ohms range.... thus changing the frequency response in a dynamic way....
Not entirely the tube, but each of the tubes you mention can have the same frequency response if the circuit is optimized for a given set of requirements...
Guitar amps are a bit more non-linear.... since you work primarily in square wave most of the time.... and square wave in the output stage spend 98% of their time in the "triode" region of the plate curves, left of the knee...This drops the effective plate resistance down to 400 to 800 ohms range.... thus changing the frequency response in a dynamic way....
Both posts #8 (Nigel) and #9 (cerrem) are correct (through cerrem's 98% figure is an exaggeration). Note however that none of what they said implies that one tube type will react differently to another.
In any case the drop in plate resistance causes an increase in bandwidth, which is not much affected by the tube type. Frequency response is determined by the output transformer, the grid coupling capacitors, and any cathode bypass capacitors.
In any case the drop in plate resistance causes an increase in bandwidth, which is not much affected by the tube type. Frequency response is determined by the output transformer, the grid coupling capacitors, and any cathode bypass capacitors.
No...not really an exaggeration..OK, maybe slightly ......
If you account for the Rise and Fall times of the square wave passing through the plate curves with respect to time.. that leave a whole lot of "dwell" time up in the "triode" region of the curves... Quite obvious from scope plot measurements as well as calculated in time domain...
Integrate the plate resistance in degrees or Radians, whatever you prefer... The average will fall mostly in the value of slope of the triode region...
More than a slight exaggeration.
While in certain styles of music guitarists like to drive their amps as hard as they can, a halfway decent amp engineer arranges that the drive tube overdrives just before, or at the same time as, the output stage. This avoids grid cap and other effects that don't (usually) sound musical - eg a sound glitch/hessitancy just as the amp is comming out of overdrive. I have to include that word "(usually)" or course (see note below), as if there is an effect of some kind, some really creative guy will find a way to exploit it. But that's really the province of fuzz boxes, where the harmonics are generated at low level where it's more controllable and repeatable to do so.
And, for a signal cycle of 360 degrees, the maximum time a tube in a push pull pair (or in SE) can be fully "on" is 180 degrees i.e., a maximum of 50% of the time. During the other 180 degrees the tube is cutoff and open circuit.
Note: Sometimes creativity happens entirely beyond the musician's control. The first use of serious use of distortion is said to have occurred in about 1952. An American band (whose name I should know, but I forget and my song i.d.'ing app refuses to run just now for some reason...) got their very first booking at a recording studio. They set off all piled into the one car - the bass guitarist's amp strapped on the roof. On the way it poured with rain and, running late, negotiating a corner the amp fell off on to the road. In those days you either recorded within the booked time or you didn't record. On arrival the speaker was a write-off. The frame was bent and the cone torn and soggy. Or so it seemed a write-off. On switching on and playing, there was sound - horrible mangled distorted sound. And a run away steaming winner!
Our ears/brains have a peculiar but sometimes useful trick: if we hear the harmonics of a bass note but not the note itself then we think we can hear the note plus its harmonics. That means that a distorting circuit can sound like it has a bass response which extends to lower frequencies than it really does. This may partly explain why some people believe (quite wrongly) that different valves have different audio frequency responses. As others have said, most valves work from DC up to low VHF - it is the circuit which sets the frequency response.
True.
And the song exploiting distortion, regarded as the first to do so, that I couldn't remember before was Rocket 88 recorded in 1951 by Jackie Brenston and The Delta Cats. If you listen to it, you can clearly hear that torn speaker cone flapping & buzzing about. It made the song, an otherwise fairly ordinary black group song, into No1 on Billboard. The torn cone killed the proper bass response, but you still hear a sort of bass, in the manner DF96 said.
And the song exploiting distortion, regarded as the first to do so, that I couldn't remember before was Rocket 88 recorded in 1951 by Jackie Brenston and The Delta Cats. If you listen to it, you can clearly hear that torn speaker cone flapping & buzzing about. It made the song, an otherwise fairly ordinary black group song, into No1 on Billboard. The torn cone killed the proper bass response, but you still hear a sort of bass, in the manner DF96 said.
The 50% of cycle only happens when the bias is set near cutoff. The amp will opperate in Class A until the input signal goes negative enough to push the output down to the cutoff level.
Last edited:
Pauldune,
Under the theorectically severest possible overdrive, a tube is either hard on or hard off - and spends equal time in each state. Simple and obvious.
When a tube is hard on, it is operating at an anode voltage below the point where the anode current is substantially independent of the anode voltage - so it becomes very dependent - so the dynamic anode resistance falls to a low value as Cerrem said. When the tube is cut off, it draws no current - so the dynamic anode resistance rises to infinity.
How does it look incorrect to you? In what way?
Under the theorectically severest possible overdrive, a tube is either hard on or hard off - and spends equal time in each state. Simple and obvious.
When a tube is hard on, it is operating at an anode voltage below the point where the anode current is substantially independent of the anode voltage - so it becomes very dependent - so the dynamic anode resistance falls to a low value as Cerrem said. When the tube is cut off, it draws no current - so the dynamic anode resistance rises to infinity.
How does it look incorrect to you? In what way?
I know that...
Keit even states it is so for SE
Thats why I asked if I missed something.
This is what I missed. I can agree with you for an extremely overdriven class ab push pull amp. But not for class a, and certainly not for SE.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.