Stop copying the Valve Caster circuit, it doesn't always work!

[Gattoconglistivali]

And stop stacking more of them together, it WILL NOT WORK!

I've heard a lot of people saying running tubes at 9 volts "doesn't work"... "sounds bad"...
No, it will work, it works well, maybe not as good as high voltage, but better then most solid states with least voicing efforts required.

I never had a good guitar amp, I don't have space for them, even I had, I don't have space th turn them up, so it will be a waste of their merit.
What I always do is use some amp modeling pedals and IR.

When I built my first Valve Caster, I was astonished how good the sound it creates. It became my primary preamp since then. With the desire of exploring the possibility of such rewarding conditions (easy to build, low costs, tubes last forever... ), I tried build more low voltage tube preamps for different purposes and experiments.

Then there comes the problem: when they work well they sounds fantastic, but sometimes they don't work, sometimes they produce broken sounds. Does that mean those tubes doesn't work at low voltage? It's possible, but before the worst case of throwing your expensive tubes away, there's something could be done(and should be).

That is : don't trust the load resistor values from Valve Caster, or any other circuits, or whatever people tell you. You have to find it yourself. Remember, tubes are unstable components like germanium things, and they differs a lot. You might have heard about building a germanium fuzz pedal, you have to use variable resistor, tune it until you get the voltage cross the transistors right. You should tune the resistance until...well, I don't know.

BIASING THE TUBES:
Every tubes are different, the best way is to make the headroom and "buttroom" equal. Most of the time setting it to 5~6 volts will be good.(cathode short to ground) at this point, a good 12au7 tube offers about 10x voltage gain.And can handle +0.3~0.35v input signal before saturate, and -0.3~-0.35v before cut off .
Lower gain doesn't renders the tube unusable, when a tube is always cut off no matter how high the load resistance, or always saturated no matter how low the load, that's unusable.
The Valve Caster has a relatively high load resistance for the 1st stage to keep it biased properly when cathode resistance changes. Or else when you turn down the gain knob, it will get close to cut off, and possibly start generating broken sound.

Lowering load resistance


Increasing load resistance


A load resistance just fine


ASYMMETRY CLIPPING:
DO NOT try to make a cold clipper stage. We don't have gain for that. And it makes that particular stage no gain left, you will struggle to get more clipping after that. Use the saturation for clipping. Ordinarily amps at high voltage use cut off as clipping methods because tubes can't get saturate, we are now starving the tubes, we can saturate them easily.
Increase the load resistance (anode voltage drop lower) so that the positive lobe of input signal saturate earlier;
decreasing the load resistance will largely decrease the gain and eventually makes the negative lobe cut off earlier.
Some says it makes them the same with semiconductor clipping so the tube sounds are gone, Nope, that's not right, it's still non linear distortion and it's still soft clipping. When you push it hard enough everything is hard clipping even a tube amp. But it's still nothing like a diode clipping.

THE PERFECT FARTHING MACHINE:
STOP STACKING A HELL LOT OF VALVE CASTERS TOGETHER!! It will only fart!
I often heard people saying a tube makes the sound good, more tubes must make it better. No, unless they're used properly.
Like any distortion pedal, you need to push mids into clipping.it means cut low and high before gain stage. When it farts, it's too late to cut anything.
How to make it work? Replace the first input capacitor in the first Valve Caster with a 470pF(code:471) and it will clip beautifully when you add a 2nd Valve Caster, and you can get aggressive distortion sound with a 3rd Valve Caster. If it starts mad cat screaming, connect some 470pf parallel to the load resistors, try and see how many you need.

[Steben]

I understand the essence of your post, but frankly the bottom line is one has to understand how components work and where their comfort zone is.
If fed with higher voltages, triodes are in fact the most stable components out there and the most easy to get into linear working with the least periferal components. Only outclassed with well designed opamp stages (which are tons of components in a box). One of the reasons HiFi purists adore class A triode power amps.

There are low voltage tubes much easier to bias which were designed for car audio.

[Gattoconglistivali]

I understand the essence of your post, but frankly the bottom line is one has to understand how components work and where their comfort zone is.
If fed with higher voltages, triodes are in fact the most stable components out there and the most easy to get into linear working with the least periferal components. Only outclassed with well designed opamp stages (which are tons of components in a box). One of the reasons HiFi purists adore class A triode power amps.

There are low voltage tubes much easier to bias which were designed for car audio.

Cool information, interested about your sources.
As I used to know, tube sound comes from it's non linear characteristics. if it's perfectly linear, the sound would be no difference than solid states.(a lot of tube amp experts said so)

Tubes get stable at high voltage that's right, or we should say low voltage makes them less stable, but they are far from "the most stable" components. And they even change over time.(never had to bias an OPA, or replace a transistor only because it got old). Again, very interested to see sources suggest otherwise

And I've never needed a tone of components to outclass tube amplifiers when I want it linear , sometimes a power amp only takes a single IC with no more than 10 components.

What takes a tone of components is to emulate tube sound with semiconductors, and why I'm not interested of this technique is because just build a tube amplifier is a far easier way to get that sound.

[anotherjim]

I think the ton of components referred to are in the internal opamp circuit.

If you assume there is no correcting negative feedback, it always seems to me that any class A might struggle to be linear.
There is a fixed load resistance and if I'm right, the active device is in parallel to the load for AC signal. If you were to put a linear pot as a variable resistor in parallel with a fixed resistor would you expect a linear adjustment response from the combination? So if a 10k resistor is in parallel with a 10k variable, setting the variable to 5k does not make the total 5k. This suggests that to be linear, you either need some corrective feedback or an active device that is non-linear in just the right direction to compensate and produce a linear result with the right value of load resistor.

[Steben]

Yes, indeed. I ment that an opamp is a whole lotta components in a tiny package. But therefor a great building component in rapid design.
If you compare discrete components the classic triode has a portion of its transfer curve which is quite linear compared to solid state components or pentodes (which are closer to transistors than most want to see) for example.
And mentioning "tube tone"... It's way more complicated than that as far as I've noticed in years of listening in discussions and looking from the side.
Without feedback and within the clipping boundaries, transistors have higher distortion component than a triode. Both even as odd order. The transfer curve is relatively more curved (log). This means that (again: simple!) stacked amp stages will result in more distortion and that an array of triode might just sound good because it has less total distortion. Even order distortion has worse IM distortion than odd order.
And do not forget, mainstream tube tech is high supply voltage with components hardly having a gain much higher than 100. Solid state exhibits lower supply voltage and much higher gain per component. The guitar amplitude is the same in both cases! This makes for a very complicated comparison with the tube amp having more components to arrive at the limit of clipping.
In comes solid state tech with lower gain at each stage by some feedback which flattens the transfer curve (less distortion closer to triodes). You'll see a pattern in for example Peavey transtube preamps or the Fetzer valve.
Check this (I love the guy's playing here):

This a clean channel on a Redstripe Bandit (only 2 transistor "preamp stages" in total) which passes a symmetrical transtube clipping circuit with compression. The preamp is not ment to clip here, only the warm up the signal into what follows. Add a treble booster and you'll get very close to Rory's '74 tour sound.

[amptramp]

For low voltage tube amps using 12 volt car radio tubes, see here:

https://web.archive.org/web/20080303033317/http://www.sophtamps.ca/mambo/index.php

[blackieNYC]

Headroom? Buttroom.
I'll allow it.

[Gattoconglistivali]

Yes, indeed. I ment that an opamp is a whole lotta components in a tiny package. But therefor a great building component in rapid design.
If you compare discrete components the classic triode has a portion of its transfer curve which is quite linear compared to solid state components or pentodes (which are closer to transistors than most want to see) for example.
And mentioning "tube tone"... It's way more complicated than that as far as I've noticed in years of listening in discussions and looking from the side.
Without feedback and within the clipping boundaries, transistors have higher distortion component than a triode. Both even as odd order. The transfer curve is relatively more curved (log). This means that (again: simple!) stacked amp stages will result in more distortion and that an array of triode might just sound good because it has less total distortion. Even order distortion has worse IM distortion than odd order.
And do not forget, mainstream tube tech is high supply voltage with components hardly having a gain much higher than 100. Solid state exhibits lower supply voltage and much higher gain per component. The guitar amplitude is the same in both cases! This makes for a very complicated comparison with the tube amp having more components to arrive at the limit of clipping.
In comes solid state tech with lower gain at each stage by some feedback which flattens the transfer curve (less distortion closer to triodes). You'll see a pattern in for example Peavey transtube preamps or the Fetzer valve.
Check this (I love the guy's playing here):

This a clean channel on a Redstripe Bandit (only 2 transistor "preamp stages" in total) which passes a symmetrical transtube clipping circuit with compression. The preamp is not ment to clip here, only the warm up the signal into what follows. Add a treble booster and you'll get very close to Rory's '74 tour sound.

Oh, I got it!

But comparing a 400v tube with a 10v transistor and saying tubes are linear up to 100v and transistor only goes 10, is unfair. There are high voltage transistors up to 1000v and will be linear all the way to the top(before distortion), in comparison, tubes always suffer (or "enjoy " in the case of guitar amps) some non liner distortion even in its most linear conditions.

You just don't see them in guitar amps, why? They're not needed.

A speaker is a low voltage device (and high current). That's why we need output transformers in tube amps. Tubes need high voltage, speakers need low voltage (high current), so we need transformers. But with transistors' high current capabilities, we can now just drive the speakers, high voltage is not needed in a solid state amp anymore.

↑Just to be clear tubes are not the most stable and linear components,they are not even stable or linear at all

Back to the valve caster, it's not my own design, and I believe it's designed this way because being an overdrive pedal, we just don't want it linear and we don't want headroom.

[Steben]

I lost the link to comparison between Triode, jFET and BJT. I keep looking...

[Steben]

https://www.its.caltech.edu/~musiclab/feedback-paper-acrobat.pdf

found it!

Check the comparison of harmonic spectrum bewteen triodes, jfets and bjts.... And how feedback through source or emitter resistors reduces the harmonics and how jFET (square law) and BJT (log)comes closer to a CLEANER triode (1,5 law).
Now think about what I ment with using class A transistors stages with some feedback reducing their gains (in relation to lower supply voltage) and reducing higher harmonic content (flattening the square or log function).

Bottom line: many old school tube preamps have a decent design by nature for guitar use. Yet solid state in preamps can sound great by understanding the component and typology. It needs careful biasing and feedback to shape the response and "tame the gain".

[PRR]

...unfair. There are high voltage transistors up to 1000v and will be linear all the way to the top(before distortion).....

Absolute voltage does NOT matter. We audio-techs can always cut-back or boost-up to suit any requirement. The primary limit is universal hiss. But it works out that higher threshold voltage devices have higher hiss voltage. Tubes are at a slight disadvantage because they run HOT. But all low-level e-guitar interfaces evolved to account for that.

BJT transistors are *NOT* "linear" in voltage-to-current. (Or with resistor load, in voltage-to-voltage.) They are grossly non-linear above a dozen mV input, or outputs above a few % of supply voltage. BUT they also have FAR higher gain than any of the side-throttle devices. Tubes and FETs, the charges do not touch the grid/gate, they pass alongside and are influenced at a distance. BJTs, the charges flow IN the crystal lattice and are directly "touched", influenced. The transconductance gain is far higher. So high that BJTs are typically NOT DC stable without a heaping helping of NFB. This is often also used to trade-away gain for linearity.

[Steben]

side-throttle devices.

nice

This is often also used to trade-away gain for linearity.

Which is no biggie if gain is what you have plenty.
Let's be honest, if tubes were small, did not use high voltages and did not use ridiculously power sucking heaters... they were almost perfect.

[teemuk]

There is such a device. It's called a transistor. 

And... at it's dying breath the tube industry did develop these things called "nuvistors"; low voltage capability, very low filament current draw and enclosures compact enough to compete in size with early discrete transistors. In fact, nuvistors were by purpose developed to compete with the new transistor technology. Because of their incredibly small size nuvistors even had to be assembled by robots in a vacuum chamber. Nuvistors were employed in Tektronix high end oscilloscopes, Ranger program snd aviation electronics.

Guess which device was still superior in the end and obsoleted the highest of the vacuum tube high-tech in just couple of years?

[FiveseveN]

The killer app for transistors is integration: 10 billion devices on one die, while in tube world the best we got is... two. I'd say you can't make a tube from a handful of atoms but they sure are trying: https://en.wikipedia.org/wiki/Nanoscale_vacuum-channel_transistor

https://www.its.caltech.edu/~musiclab/feedback-paper-acrobat.pdf

found it!

Let's not forget how a grounded-emitter amp was the latest boutique unicorn pedal not long ago. FFT says "poor performance", people hear "rich, crispy, sparkly". Good advertising will trump your carefully crafted transfer functions any day .

[teemuk]

"Transparent overdrive". 

So you stomp on the switch expecting great distortion tone and it... ...does nothing? 

[FiveseveN]

Presumably the point is to give you "more of what you already have", i.e. to not filter or distort the signal to the extent that it imposes its own "character". A trivial task that was solved half a century ago, yet the demand for "transparent overdrives" (be they distinct or derivative) looks to be on the rise. Because what people really want is some high-pass and some saturation and there's plenty of variation in that space to keep builders and players entertained. And since an overdrive/boost is usually used in an interactive way, different amps (or whatever is being overdriven) and guitars serve to expand that space of possibilities and its relation to taste. Funny how about none of them sound much like a triode (or more), but people keep insisting that's what they were trying to do.

[teemuk]

Yeah, but what's that "triode sound" anyway. If you look back at the history of distortion effects the first ones were solid-state and built to mimic the archetypal sound of broken equipment, and yes, possibly also overdriven tube amps mixing consoles, mic preamps et cetera.


But none of this "triode sound" was yet even important: a half-assed circuit like the FZ-1 or Fuzzrite could produce the most important: Distortion. The effect - nevertheless what it was or what nuances it had - was simply marketed with less glorifying terms such as "fuzz", "buzz" or even "mud". No one was marketing "tube sound". (Well, some effects were marketed of being able to produce "mud" of tube amps).

It took several years before someone even invented they can use tubes for producing distortion (aside from cranking amps to inconvenient volume levels). It took several years more to invent the idea that "tube distortion" is the only good sounding distortion. It took a few more years to invent the idea that "soft clipping" or "asymmetric clipping" are not actually realistic tube distortion emulations. It took even a few years longer to invent the idea that this "distortion" should actually sound like a complete amplifier system distorting, etc.

So, I kinda think that most of these ideas how distortion should sound like ideally are more or less just gimmicks. "Triode sound"... do we even have a generic consensus what that is? I don't think so. It's just a vague marketing phrase like "transparent overdrive".

[PRR]

...things called "nuvistors"; low voltage capability, very low filament current draw and enclosures compact enough to compete in size ...assembled by robots in a vacuum chamber.....

I worked with them when new. They were incrementally better than existing types. 6CW4 was the most common consumer type. The 6.3V 0.135A heater is just a skooch less than half a 12AX7, and the Gm and Rp are maybe 50% better. The operating voltage was typically 70V to 110V. Intermodulation was less than a naked BJT, but a fed-back BJT was clean-enough and a heap cooler, a JFET better than either.
http://www.r-type.org/pdfs/6cw4.pdf

Ampex MR-70 and AKG C12a seem to be the notable audio uses. CBS used one in a compressor but only as a buffer in the side-chain (meter) system. Fisher sold a ton of FM tuners with Nuvistors in the RF end but that may have been as much marketing as engineering.

Nobody working on "cheap" (pedals were always cheap inside) thought Nuvistors had any use. Even though Nuvistors faded-away before pedals took over their world (and thus may have been available cheap-bulk).

Robotic, because they left-off the suction pip. So had to be assembled in vacuum. That did not suit human assemblers. And yes, some of the improved performance is reliable high precision assembly, which is always tough, but when you train a robot it stays trained.

[Steben]

Yeah, but what's that "triode sound" anyway.

If it is anything, it is more linear than a discrete solid state stage, but just a tad non linear if not driven of course.
In the tube HiFi world, triodes are ALL about being the "most" linear.
Looking back at the golden ages you're d"mn right "rock" sound is easily connected to fuzz boxes and boosted power amps... many times with another small "log" solid state device...

You can put your question in another light: what is the Fender blackface sound... ? Nice cleans you say? Now what would they mean with clean? 

[anotherjim]

We could probably make some Nuvistors. Made by Roberts. Our Deadastronaut is perfect for the job. 
I expect all the high gain Nuvistors have been snapped up for boutique condenser microphones. IIRC the ones Tektronics used were low gain around 30 and I'm sure the data also allowed for 5v heater supply for an extra saving.