Bare Bone Klone (no charge pump)
- earthtonesaudio
- Transistor Tuner
The charge pump in the Klon seems a bit unnecessary to me. The only logical reason it should be used is to keep op-amp U2 from clipping.
But here's the thing. When the gain pot is at maximum, the only signal hitting U2A has already been clipped by the diodes, so a typical Ge diode will limit the signal going to U2A to under 1V p-p. For this signal U2A has a gain of about 8.3. Granted, this would clip a TL072 with a 9V single supply. BUT, if you used a rail-rail op-amp, then you'd get no clipping.
That would ensure no signal coming from the distortion stage could cause clipping in U2A, even if you used +9V and 0V as the rails.
From U2A the signal goes into the tone section, where the maximum treble boost gives a gain of about 8.2 to the highest frequencies. Again referring to the situation where the gain pot is at max, that means the signal coming from U2A into U2B is about 5V (.6V p-p signal times U2A's gain of 8.3), so U2B will try to make the highest frequencies into a 40V p-p signal. The power supply can't reproduce anything (even with a rail-rail op-amp) greater than 27V, so the op-amp clips.
Another useful scenario to analyze is when the gain is set to minimum. For high frequencies, the maximum gain from lug 1 of the lower half of the dual gang Gain pot to the output of U2B is 266. An average guitar signal (especially if boosted) can easily make a .1V p-p signal, which would cause op-amp U2B to clip if the gain was set to minimum, and the treble at maximum. Of course, before the signal gets to the that lug of the gain pot, it is attenuated by the input voltage divider as well as the network at the output of U1A.
Anyway, my point is that there is some op-amp clipping possible in the Klon at normal settings, even when using the charge pump. To me, the whole point of using a charge pump is to avoid op-amp clipping. But the Klon uses a charge pump and still clips the op-amps. So why not just ditch the charge pump and use op-amps that sound better when distorted (or will operate at their rails to minimize the amount of time they spend in clipping)?
For those people that would like to fit a Klon into a smaller package (though that is of course limited by the dual gang pot and other hardware), or those who just want to use fewer of the Earth's resources in their pedal building, I think it would be nice to have a Klon alternative which operates from a single supply, with no charge pump.
But here's the thing. When the gain pot is at maximum, the only signal hitting U2A has already been clipped by the diodes, so a typical Ge diode will limit the signal going to U2A to under 1V p-p. For this signal U2A has a gain of about 8.3. Granted, this would clip a TL072 with a 9V single supply. BUT, if you used a rail-rail op-amp, then you'd get no clipping.
That would ensure no signal coming from the distortion stage could cause clipping in U2A, even if you used +9V and 0V as the rails.
From U2A the signal goes into the tone section, where the maximum treble boost gives a gain of about 8.2 to the highest frequencies. Again referring to the situation where the gain pot is at max, that means the signal coming from U2A into U2B is about 5V (.6V p-p signal times U2A's gain of 8.3), so U2B will try to make the highest frequencies into a 40V p-p signal. The power supply can't reproduce anything (even with a rail-rail op-amp) greater than 27V, so the op-amp clips.
Another useful scenario to analyze is when the gain is set to minimum. For high frequencies, the maximum gain from lug 1 of the lower half of the dual gang Gain pot to the output of U2B is 266. An average guitar signal (especially if boosted) can easily make a .1V p-p signal, which would cause op-amp U2B to clip if the gain was set to minimum, and the treble at maximum. Of course, before the signal gets to the that lug of the gain pot, it is attenuated by the input voltage divider as well as the network at the output of U1A.
Anyway, my point is that there is some op-amp clipping possible in the Klon at normal settings, even when using the charge pump. To me, the whole point of using a charge pump is to avoid op-amp clipping. But the Klon uses a charge pump and still clips the op-amps. So why not just ditch the charge pump and use op-amps that sound better when distorted (or will operate at their rails to minimize the amount of time they spend in clipping)?
For those people that would like to fit a Klon into a smaller package (though that is of course limited by the dual gang pot and other hardware), or those who just want to use fewer of the Earth's resources in their pedal building, I think it would be nice to have a Klon alternative which operates from a single supply, with no charge pump.
rocklander wrote:hairsplitting and semantics aren't exactly the same thing though.. we may need two contests for that.
- culturejam
- Old Solderhand
Information
- RnFR
- Old Solderhand
Information
i think soulsonic already has this in the bag. at least he mentioned it to me that he was working on it last time we talked.
"You've converted me to Cubic thinking. Where do I sign up for the newsletter? I need to learn more about how I can break free from ONEism Death Math." - Soulsonic
Blog-APOCALYPSE AUDIO
Blog-APOCALYPSE AUDIO
- earthtonesaudio
- Transistor Tuner
Cool. He has a good set of ears so I bet it will sound nice.
I was thinking, for starters, that simply replacing IC2 with a rail-rail op-amp that clips gracefully (TLC2262 perhaps) and running it off the same 9V as IC1 would get you mostly there. If you were tenacious you could re-do the gain of IC2A and IC2B slightly so that the op-amp clips with the same input signal as the original Klon.
If Martin's working on this I hope he publishes his results here, it would be interesting to see what he does to the circuit.
I was thinking, for starters, that simply replacing IC2 with a rail-rail op-amp that clips gracefully (TLC2262 perhaps) and running it off the same 9V as IC1 would get you mostly there. If you were tenacious you could re-do the gain of IC2A and IC2B slightly so that the op-amp clips with the same input signal as the original Klon.
If Martin's working on this I hope he publishes his results here, it would be interesting to see what he does to the circuit.
rocklander wrote:hairsplitting and semantics aren't exactly the same thing though.. we may need two contests for that.
- soulsonic
- Old Solderhand
Information
The thing I'm working on still uses the charge pump, but I took out the voltage doubler part.
I'm pretty much re-doing the entire circuit from scratch to be functionally similar to the Klon but work in a different way.
So, if you're wanting to try it without a charge pump, that seems like a worthwhile thing to pursue.
I'm pretty much re-doing the entire circuit from scratch to be functionally similar to the Klon but work in a different way.
So, if you're wanting to try it without a charge pump, that seems like a worthwhile thing to pursue.
"Analog electronics in music is dead. Analog effects pedal design is a dead art." - Fran
- bajaman
- Old Solderhand
Information
- Posts: 4549
- Joined: 26 Jun 2007, 21:18
- Location: New Brighton, Christchurch, NZ
- Has thanked: 595 times
- Been thanked: 2056 times
This looks interesting - less current drain too
bajaman
bajaman
be kind to all animals - especially human beings
- blanik
- Solder Soldier
it will be a greener klon... we could call it the Kleen, it' a Klon, it's green, and it's clean... it's the Kleen!bajaman wrote:This looks interesting - less current drain too
bajaman
maybe Bono could be our spokesperson!
- earthtonesaudio
- Transistor Tuner
Haha!
rocklander wrote:hairsplitting and semantics aren't exactly the same thing though.. we may need two contests for that.
- mictester
- Old Solderhand
Information
You are entirely right. The approach I've taken is to attenuate the post-clipped signal, so that the final stage gets a smaller signal, and doesn't "hit the rails". Experiments on this topology have shown that it sounds just the same as the "original", but with a slightly lower output level. Who needs 16 V of audio at the output of a stompbox, anyway?earthtonesaudio wrote:The charge pump in the Klon seems a bit unnecessary to me. The only logical reason it should be used is to keep op-amp U2 from clipping.
But here's the thing. When the gain pot is at maximum, the only signal hitting U2A has already been clipped by the diodes, so a typical Ge diode will limit the signal going to U2A to under 1V p-p. For this signal U2A has a gain of about 8.3. Granted, this would clip a TL072 with a 9V single supply. BUT, if you used a rail-rail op-amp, then you'd get no clipping.
That would ensure no signal coming from the distortion stage could cause clipping in U2A, even if you used +9V and 0V as the rails.
From U2A the signal goes into the tone section, where the maximum treble boost gives a gain of about 8.2 to the highest frequencies. Again referring to the situation where the gain pot is at max, that means the signal coming from U2A into U2B is about 5V (.6V p-p signal times U2A's gain of 8.3), so U2B will try to make the highest frequencies into a 40V p-p signal. The power supply can't reproduce anything (even with a rail-rail op-amp) greater than 27V, so the op-amp clips.
Another useful scenario to analyze is when the gain is set to minimum. For high frequencies, the maximum gain from lug 1 of the lower half of the dual gang Gain pot to the output of U2B is 266. An average guitar signal (especially if boosted) can easily make a .1V p-p signal, which would cause op-amp U2B to clip if the gain was set to minimum, and the treble at maximum. Of course, before the signal gets to the that lug of the gain pot, it is attenuated by the input voltage divider as well as the network at the output of U1A.
Anyway, my point is that there is some op-amp clipping possible in the Klon at normal settings, even when using the charge pump. To me, the whole point of using a charge pump is to avoid op-amp clipping. But the Klon uses a charge pump and still clips the op-amps. So why not just ditch the charge pump and use op-amps that sound better when distorted (or will operate at their rails to minimize the amount of time they spend in clipping)?
For those people that would like to fit a Klon into a smaller package (though that is of course limited by the dual gang pot and other hardware), or those who just want to use fewer of the Earth's resources in their pedal building, I think it would be nice to have a Klon alternative which operates from a single supply, with no charge pump.
I'll put the simplified (and slightly improved) version up in the next day or so. I've got to tidy up the documentation, as it's just pencil scribbles in a notebook!
C.
"Why is it humming?" "Because it doesn't know the words!"
- CHEEZOR
- Diode Debunker
That would be very much appreciated! Thanks!mictester wrote:I'll put the simplified (and slightly improved) version up in the next day or so. I've got to tidy up the documentation, as it's just pencil scribbles in a notebook!
C.
- earthtonesaudio
- Transistor Tuner
While waiting for mictester to publish his charge-pump-less version, a little more analysis:
(referring to Soulsonic's schematic drawing)
The "upper path" around the clipping stage (comprised of R7, R19, and C16) rolls off high end starting at about 106Hz. But this section is effective only at high gain settings.
This is due to C3 and the combined resistance of R6 and P1B. When the gain is at minimum, the junction of C3/R6 rolls off frequencies below 89Hz. When the gain is at max, the roll off is nearly gone (8Hz). This means that when the gain is at minimum, the signal appearing at the junction of R7/C16 is a narrow band between 89 and 106 Hz. Basically the low E string up to the 6th fret. When the gain is increased, more low end is allowed to pass through, although this doesn't matter much unless you tune lower than E.
I was also looking at the output 68k resistors again. I'm having a hard time understanding why R26 and R27 are there. R28 I can understand more because you might worry about the middle lug of that switch floating during the brief switching interval... but I would be interested to know whether these resistors are necessary all the time, or only for problematic pop-prone switches.
(They certainly don't contribute much to the tone, because each side of this switch is driven by an op-amp through a 560 ohm resistor. This impedance is so much lower than 68k that it may as well be an open circuit to the signal.)
(referring to Soulsonic's schematic drawing)
The "upper path" around the clipping stage (comprised of R7, R19, and C16) rolls off high end starting at about 106Hz. But this section is effective only at high gain settings.
This is due to C3 and the combined resistance of R6 and P1B. When the gain is at minimum, the junction of C3/R6 rolls off frequencies below 89Hz. When the gain is at max, the roll off is nearly gone (8Hz). This means that when the gain is at minimum, the signal appearing at the junction of R7/C16 is a narrow band between 89 and 106 Hz. Basically the low E string up to the 6th fret. When the gain is increased, more low end is allowed to pass through, although this doesn't matter much unless you tune lower than E.
I was also looking at the output 68k resistors again. I'm having a hard time understanding why R26 and R27 are there. R28 I can understand more because you might worry about the middle lug of that switch floating during the brief switching interval... but I would be interested to know whether these resistors are necessary all the time, or only for problematic pop-prone switches.
(They certainly don't contribute much to the tone, because each side of this switch is driven by an op-amp through a 560 ohm resistor. This impedance is so much lower than 68k that it may as well be an open circuit to the signal.)
rocklander wrote:hairsplitting and semantics aren't exactly the same thing though.. we may need two contests for that.
- mictester
- Old Solderhand
Information
As promised: You'll see that the mixer stage has reduced gain, but the time constant of its feedback circuit remains the same. I've also got rid of some of the "mojo-value" passive components, and changed them for standard values, sometimes by putting two in parallel (R5 and R5A for example). I also put two standard value capacitors in parallel to get the "right" value. This will reduce the component cost substantially - especially for anyone (like me) with complete sets of E12 or E24 resistor values.earthtonesaudio wrote:While waiting for mictester to publish his charge-pump-less version, a little more analysis:
(referring to Soulsonic's schematic drawing)
Further analysis of the circuit shows that the time constants aren't anything special. I'm fairly sure that the "designer" just used capacitor values on hand that "sounded good" and continued to use them.
I've experimented with this simplified circuit, with the "original" clone circuit and with the real thing. Any competently constructed clone will give that "magic" sound that the Klon is meant to have. There really isn't much that's special about the circuit.
Anyway - have fun. Build the simplified version and save on batteries!!
"Why is it humming?" "Because it doesn't know the words!"
- earthtonesaudio
- Transistor Tuner
Nice.mictester wrote:As promised:
rocklander wrote:hairsplitting and semantics aren't exactly the same thing though.. we may need two contests for that.
- earthtonesaudio
- Transistor Tuner
Hey mictester, I looked a little closer at your latest schematic here and it appears that the time constants are in fact altered from stock. You said they "aren't anything special," but does that mean they're different but both good, or does it mean there is no audible difference?
Totally unrelated, but I'm toying with the idea of building a simplified, SMD Klon so I can fit one inside a 1590B alongside a Valvecaster and some sophisticated switching. The big question I'm wrestling with right now is whether or not to use a charge pump for both, or neither.
Totally unrelated, but I'm toying with the idea of building a simplified, SMD Klon so I can fit one inside a 1590B alongside a Valvecaster and some sophisticated switching. The big question I'm wrestling with right now is whether or not to use a charge pump for both, or neither.
rocklander wrote:hairsplitting and semantics aren't exactly the same thing though.. we may need two contests for that.
- quaternotetriplet
- Resistor Ronker
is this scheme already verified?
- mictester
- Old Solderhand
Information
Yes. I've built about seven or eight of them so far. A well-known guitarist friend of mine auditioned one against his genuine, outrageously expensive one, and could hear no differences. Set up for testing was stock Strat into passive A/B footswitch, then pedals, then into Roland JC120 and then later, Gibson S1 into passive A/B footswitch, then pedals into the same amp. Neither of us could hear any differences, except than the single rail version was quieter - less hiss!quaternotetriplet wrote:is this scheme already verified?
"Why is it humming?" "Because it doesn't know the words!"
- mictester
- Old Solderhand
Information
The time constants are exactly the same as the originals. Where resistors were changed, the corresponding capacitor was recalculated to keep the time constants exactly the same.earthtonesaudio wrote:Hey mictester, I looked a little closer at your latest schematic here and it appears that the time constants are in fact altered from stock. You said they "aren't anything special," but does that mean they're different but both good, or does it mean there is no audible difference?
"Why is it humming?" "Because it doesn't know the words!"