Have you seen the prices of KT88's lately? Yowzer. GU-50's are under $5, and sockets are similar. (I'm starting to sound like Wavebourn.)
..Todd
For a HiFi amp we tend to minimize distortion and for an SE HiFi amp we want most of the distortion to be second harmonic. The rules become more lenient for a guitar amp, and highly user dependent. I doubt however that there is much use for an amp that is fairly distorted all the time even at a low volume.
The HiFi world is populated with people called "tweakers". People who change components, often without full knowledge of the principles involved, in search of audio nirvana. The guitar amp world takes this to new levels. I must admit that in my previous life I was a guitar amp tweaker. I made some unique sounding guitar amps by doing things that were technically wrong like stuffing a Bandmaster with one 6L6GC and one EL34.
Finding the "perfect tone" is as you said an art form that requires some serious tinkering and listening. The results are highly dependent on the user, the amp design, and the speakers involved.
I try to design the guitar power amp from almost a HiFi perspective. I want it to be clean up to a fairly high power level, but it must be capable of being overdriven by 20 db or more without damage. The 20 watt amp that I am working on for the hundred buck amp challenge will do 10 watts at 0.5% distortion. That is measured from the guitar input to the speaker output with the volume and master gain on "1". The amp has WAY more gain and power than needed so it can be cranked into mega - overdrive, and then an attenuator can be engaged between the amp and speaker for quiet playing. The controls can be set for preamp distortion, power amp distortion, or both, then the output level can be set with the attenuator.
That said, I find that quite a high level of distortion is OK, and even sounds good, at low volumes if the distortion is mostly second harmonic. I have another amp that makes about 2 watts cleanly and 2.5 watts at full saturation. It has a measured distortion of 5% or more at any volume level, yet sounds rather clean when playing chords.
IMD distortion will make chords sound very bad since 6 simultaneous notes will have 64 second order IMD components and zillions of higher order components. IMD can't be avoided when the amp is cranked, which is what power chords are all about.
Current production KT88's are priced from about $60/pair retail to double that for premium brands. NOS are unobtainium and priced accordingly.
Of course finding a matched pair of GU50's will probably take a box of 20 to accomplish, but for an SE amp, thankfully that's not a concern.
..Todd
Yup, from what I've read and listened to, you can use a considerable amount of H2 with the H3 and up harmonic spectrum cascading downwards (by adjusting the Vq point upwards from center), and still achieve a nice warm "clean-ish sound" (It's a blues amp I'm making.) Just a taste of 'hair' on it.
SE will give me an abundance of H2 (the reason behind this design goal) that would otherwise be significantly reduced in a push-push output stage.
But the crunchy distortion apparently comes from driving hard and overdriving the tubes in succession as they head towards the output. I play (hah!) a strat with fairly standard pickups so they're not high output and won't overdrive the input tube. I do have a humbucker in the bridge position for lead solos (hah!).
..Todd
Does the first stage of a guitar amp overdrive much? Considering that it is usually a 12AX7 with a 100K-200K anode resistor and an additional tonestack load, I would expect the gain to actually be close to 60.
With a 300V supply and center biased around 170V you will need near 110V peak to start clipping so input would be around 1.8V peak to drive it to clipping.
The initial peak was that high in the P-Scope graphs Tubelab posted (1.4Vpeak), and IIRC someone else posted showing a high peak with an 'E' chord. However the signals decay rapidly.
Hence I don't think you get much overdrive distortion from the first stage. Second stage with the gain turned up should distort much more easily even considering the attenuation of the tonestack.
Beyond that gain must be controlled (attenuated) to prevent clipping so any further stages are not an issue.
Thus I would expect the gain stages typically to start clipping at the PI, working back to the input in succession, and somewhere along the line (gain and volume dependent) the output starts clipping.
If this were not so and the first stage started clipping first, there would not be a lot of additional increase in amplitude to drive the subsiquent stages into clipping.
That said, I would still expect a lot of distortion out of a 12AX7 before it is driven to 250Vp-p with a 300V supply.
With a 300V supply and center biased around 170V you will need near 110V peak to start clipping so input would be around 1.8V peak to drive it to clipping.
The initial peak was that high in the P-Scope graphs Tubelab posted (1.4Vpeak), and IIRC someone else posted showing a high peak with an 'E' chord. However the signals decay rapidly.
Hence I don't think you get much overdrive distortion from the first stage. Second stage with the gain turned up should distort much more easily even considering the attenuation of the tonestack.
Beyond that gain must be controlled (attenuated) to prevent clipping so any further stages are not an issue.
Thus I would expect the gain stages typically to start clipping at the PI, working back to the input in succession, and somewhere along the line (gain and volume dependent) the output starts clipping.
If this were not so and the first stage started clipping first, there would not be a lot of additional increase in amplitude to drive the subsiquent stages into clipping.
That said, I would still expect a lot of distortion out of a 12AX7 before it is driven to 250Vp-p with a 300V supply.
Not the first stage. Unless someone is using REALLY active pickups, or an overdrive pedal.
The Bluesman amp circuit I'm following uses a pentode (6SJ7) for input, then a 12AX7, half for a gain stage and half for cathode follower before the tone stack, then output. I might use an EF86 at the input instead.
I don't think I would ever push it hard enough to overdrive the output section. Except maybe to test it, when the neighbors aren't home.
..Todd
Last edited:
The first stage does clip the peaks a bit when hit hard but as stated doesn't distort much. I read in a book somewhere that the actual stage gain of a 12AX7 in a typical guitar amp circuit is 45 to 55. Not being one to believe everything I read, I wired up a 12AX7 with a bypassed pot in the cathode, the typical 68K grid stopper, a 100K plate load, and a 500K pot for a load. I measured a gain of 45 to 50 depending on the tube I tried. The pot was tweaked for a balance between gain and maximum undistorted output with 250 volts of B+.
The tone stack can have as much as 20 db of loss (play with the Duncan Amps tone stack simulator). This is a voltage loss of 100. So it is possible to have less signal at the second stage, and therefore less distortion.
My 5 tube amp for the hundred buck amp challenge doesn't use 12AX7's for cost reasons, so I have even less gain. To fix this I moved the tone stack all the way back to the PI and use 4 gain stages.
This is what you want.
In some cases this is also what you want. An amp is said to be expressive if you can adjust it such that you can be clean but on the edge of distortion, and just by playing a bit harder push the amp into distortion. A bit harder and you get multiple stage distortion. This way the player can control the amp with his picking style. This usually requires a volume control up front and a master gain just before the PI.
Look at the circuit for the small amp I am building for the challenge. See post #1021.
http://www.diyaudio.com/forums/instruments-amps/190738-hundred-buck-amp-challenge.html
The input pentode is similar to the 6AU6 which is like a 6SJ7 in a 7 pin bottle. The 6AV6 is half of a 12AX7. The tone control circuit is rather unique and can go from max treble with no bass to lots of bass with no treble, and doesn't have too much loss. I did not design it but I can't remember where I found it either. There is no PI. The output stage is wired like push pull but only one tube is driven. With no cathode bypass cap the output stages functions like an unbalanced P-P so there is lots of second harmonic. Switching in a bypass cap makes it an SE design but using a P-P OPT (lots cheaper). I don't like the fully SE sound so I ditched the bypass cap, but I may experiment with a pot in series with the cap later.
The "self split" output stage has been used for HiFi designs because it sounds somewhat like SE but uses a cheap OPT. I never liked it for HiFi because of high overall distortion, but it works good here. A CCS instead of the resistor in the tail will make this a full P-P design but must operate in class A. Good for HiFi but sounds pretty ragged when pushed into clipping, so not so good for HiFi.
Does this loadline look reasonable and correctly calculated, or is my comprehension still fundamentally flawed? 
Hi costis_n,
Thanks for the feedback. Your lower current is a good thing, but I didn't realize you could move so far to the right, forcing it to be so asymmetrical. How much H2 is there? Does it sound okay like that? I guess it would or you wouldn't have done that.
Where did you get the output transformer for yours?
..Todd
Thanks for the feedback. Your lower current is a good thing, but I didn't realize you could move so far to the right, forcing it to be so asymmetrical. How much H2 is there? Does it sound okay like that? I guess it would or you wouldn't have done that.
Where did you get the output transformer for yours?
..Todd
Attachments
Last edited:
Well look
a. with the speaker impedance varying so widely with frequency and being inaccurate in general, calculating loadlines is just educated guesses. Don't be fussy. Just keep it generally ok. For example, I had already the transformers, so the voltage was already set.
b. Guitar amplifier , more H2=better
c. My loadlines usually fall over to the right side of the graphs, cause i tend to go with high voltage high load for my HiFi amp too. Valves do not cut off as easily as V/I graphs imply, it is very gradual.
d. Lower current+ higher loadline= longer life, esp in SE. Mr tubelab has explained it earlier. more cathode current = more stress to the cathode coating.
I get my transformers custom wound from a shop 2 blocks away from my home. Or from one like 10 minutes away. 30 euros apiece. (the benefits of living in Athens) Want some ?
As to how it sounds, I cant tell right now, It is a bunch of wires an hums too loud. Sounds like a guitar amplifier, I think though.
a. with the speaker impedance varying so widely with frequency and being inaccurate in general, calculating loadlines is just educated guesses. Don't be fussy. Just keep it generally ok. For example, I had already the transformers, so the voltage was already set.
b. Guitar amplifier , more H2=better
c. My loadlines usually fall over to the right side of the graphs, cause i tend to go with high voltage high load for my HiFi amp too. Valves do not cut off as easily as V/I graphs imply, it is very gradual.
d. Lower current+ higher loadline= longer life, esp in SE. Mr tubelab has explained it earlier. more cathode current = more stress to the cathode coating.
I get my transformers custom wound from a shop 2 blocks away from my home. Or from one like 10 minutes away. 30 euros apiece. (the benefits of living in Athens) Want some ?
As to how it sounds, I cant tell right now, It is a bunch of wires an hums too loud. Sounds like a guitar amplifier, I think though.
Last edited:
Ha ha. Okay thanks. That gives me more freedom to play with things.
Speaking of transformers, if the amp generates 14 watts output power, would a 15W output transformer be adequate, or should I move up to 25W? If it was for hi-fi the choice would be obvious, but I understand there might be good reason to use a smaller one for a guitar amp.
..Todd
Speaking of transformers, if the amp generates 14 watts output power, would a 15W output transformer be adequate, or should I move up to 25W? If it was for hi-fi the choice would be obvious, but I understand there might be good reason to use a smaller one for a guitar amp.
..Todd
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.