Tube Amp Forum: The Ultimate Tone

Full Version: Submini tubes
You're currently viewing a stripped down version of our content. View the full version with proper formatting.
Hey there, 
I'm a pedal builder who is looking to get into using tubes (specifically subminiature tubes)for pedals and hopefully preamps eventually. I know a fair amount about pedal design but the use of tubes is not something that I come across very often and know next to nothing about how to begin. Does anyone have experience with designing with submini tubes or know where I can go to learn?
Cheers 
Steve
Hi Steve

First I will re-iterate my usual recommendation that tubes under foot are not a good idea. That does not stop people from doing it, however Smile

Most of the very tiny tubes were meant for use in harsh environments, so you have that working in your favour. Your tube pedal will likely have to rely on mains power rather than batteries, so that relieves you of some constraints, as well. The biggest electrical concern with tubes is the heater power required. since it literally is just "wasted" as heat. Of course, without that heat the cathode would not get to a temperature that allows liberation of electrons and for the tube to function, so the heat and power for the heater is a necessity.

The plate voltage often does not have to be too high. If you want "tubey" voltages, use an AC wall-wart for power and then internally use a reverse-wired small PT. You can get up to 200V using the smallest devices. The amount of plate power required for one or two gain stages is low, as you likely won't even be at 1mA per plate. The small PT will restrict the voltage conversion, which you will see is not the usual factor of root-2. This has to do with the basic design of tiny PTs compared to larger ones. If you use something greater than 10VA, then root-2 magically works in this application and the raw Va can be about 3200Vdc.

As far as selecting circuit values, you will see examples using quite low resistances and extremely high resistances. Since the tubes are optimised for lower plate voltages, and often for high-frequency use, lower values are what the tube expects and this is reflected in the values that arise using the plate curves to test or find load lines. Fortunately with tubes there are generally no "wrong" circuit values as they are very forgiving and robust. This leaves it open for you to begin with intuitive values and go from their, taking an empirical design route, measuring results, or listening, and making value changes then re-evaluating the new circuit.

Have fun
Hey Kevin, 
Thanks for the reply. I understand you advising against tube pedals. Whenever I see 12AX7 or 12AT7 tubes sticking out of pedal enclosures I immediately think, "that's gonna break quick." That one of the reasons I chose to go with subminis. They seem easier to protect, being so tiny. I'll be using JAN 6948s, guided missle tubes, so yeah, they should be a bit tougher than your average HIFI tube  Big Grin 

Also, pedal builders seem to ignore these tubes as a resource, so what better gimmick to have  Tongue

I'm a bit wary of kicking things in a pedal to 200V. I always find electrocuting customers is a bad business model lol. I was thinking of using a small voltage converter to supply ths heater(s) with 60V to maybe as high as 90V from a 12V or 18V wall wart.

Your last paragraph seems to suggest just taking a basic 12AX7 circuit and changing the tube to a JAN 6948. I must be reading that wrong, because in my experience things are never that simple. 

Lastly, I'm assuming since I'm using such low voltages, the standard 1/4W metal film resistors I use in my pedals would be sufficient?
Cheers
Steve
(12-25-2021, 06:21 AM)SGillespie Wrote: [ -> ]Hey Kevin, 
Thanks for the reply. I understand you advising against tube pedals. Whenever I see 12AX7 or 12AT7 tubes sticking out of pedal enclosures I immediately think, "that's gonna break quick." That one of the reasons I chose to go with subminis. They seem easier to protect, being so tiny. I'll be using JAN 6948s, guided missle tubes, so yeah, they should be a bit tougher than your average HIFI tube  Big Grin 

Also, pedal builders seem to ignore these tubes as a resource, so what better gimmick to have  Tongue

I'm a bit wary of kicking things in a pedal to 200V. I always find electrocuting customers is a bad business model lol. I was thinking of using a small voltage converter to supply ths heater(s) with 60V to maybe as high as 90V from a 12V or 18V wall wart.

Your last paragraph seems to suggest just taking a basic 12AX7 circuit and changing the tube to a JAN 6948. I must be reading that wrong, because in my experience things are never that simple. 

Lastly, I'm assuming since I'm using such low voltages, the standard 1/4W metal film resistors I use in my pedals would be sufficient?
Cheers
Steve
Cool project.....My suggestion would be not to go too low on the voltages to the tubes....60V might not sound very good. The data sheet for the 6948 has example circuits that would be a good starting point and data for plate voltages for 100V to 250V.  Since .6W or 1/2W metal film resistors are so tiny I'd personally stick with those or if I was to use 1/4W resistors I'd calculate the dissipation  to be certain it's in spec.
Hi Guys

I was just looking up the 6948 data which is actually the 6112 - turned out I already had that Smile

As Makinrose said, the data sheet has b0th the curves and very complete application tables for operation at 100V and 200V supplies, so it is easiest to start with those.

Voltages of 300V and less are considered safe, so do not be afraid to use something in the range listed for the tables. Besides, if you build this on a PCB or actually any manner of construction, the fact it has a tube in it should be warning enough that voltages above 9V may be present.

I would not use a switch-mode converter for the plate supply, or for any part of it. These spray a ton of broadband noise through the conductors in the circuit which you likely won't see with an old oscilloscope but definitely will see with a modern DSO. It is hideous. There will be no proximity effect with the guitar and the switching noise adds an edge to circuit hiss. You may find that acceptable but I do not.

i would recommend using an AC wall-wart of 6-12Vac output. Inside the pedal, use a reverse-wired PT to step up the voltage for the plate supply. Suppose you use 6Vac and a 2VA 6V PT with 120V primary. You would intuitively expect about 170Vdc out, but that's not what you get. because of how low-VA PTs are designed, using them in reverse incurs a loss of twice their regulation rating, so you end up with a much lower output than expected. This is a good thing in this application, where you may see something closer to 100Vdc, which conveniently lets you use the application table values as a starting point.

I know in Belgium the mains is 240Vac, but a dual-primary PT allows wiring the primaries in series or parallel. An alternative is to use a 12V PT fed by 6V, which produces half the output than were it fed 2V. In any case, there are a lot of ways to get the plate voltage.

Whatever you do, DO NOT use half-wave rectification or there will be a noise penalty. Always use full bridges.

In the above scenarios, the incoming AC can feed both the step-up PT and a bridge to generate the heater voltage as DC. The tube heater is 6V and it is best to begin with something higher than 6Vac to achieve the DC value. Say you use 9Vac, which should be a common adapter value. This nominally produces about 12Vdc but there is loss over the bridge and there will be some minimum voltage needed across the regulator regardless of the type used.

Note that the plate voltage does not need regulation per se, although such a nicety would maintain tone against mains variations.

So, now we're at 100V and everything is safe and can be done linearly with none of that nasty switching noise.

Have fun
Hi Guys

Steve: Quarter-watt resistors can be used in most positions in a tube amp but there are two parameters to check for certain positions.

In a tube circuit the plate resistor should have a power rating sufficient to withstand the power resulting from the voltage across the part during normal operation. Say Ra=100k and the voltage across it is 150V. Ohm's law and the Power equations suggest a dissipation of 225mW. The instantaneous power and voltage will be higher and lower than the quiescent values, and we are already too close to the maximum for a 250mW part, so go to 400-600mW.

At 100Va the voltage across RA may be only 50V or less, so a 250mW resistor may be ample. You can figure out the R value where you would have to switch to a higher-rated component by dividing the safe idle power into the voltage-squared. For example, (50x50) / 200mW = 12k5. Pick any power rating that seems reasonabe for the component in question.

All of the other positions in a tube preamp circuit can be 250mW quite easily as the voltage across these is less than a few volts.

The other parameter to check with your resistor purchase is the voltage rating. Yes, resistors have voltage ratings and hardly anyone notices unless they read TUTs Smile

Also, remember that electrolytic cap ratings are WORKING voltage, which means they can be operated reliably and nearly forever at that value. Doing so actually keeps them healthier than derating them.

Have fun
(12-29-2021, 02:26 PM)K O'Connor Wrote: [ -> ]Hi Guys

Steve: Quarter-watt resistors can be used in most positions in a tube amp but there are two parameters to check for certain positions.

In a tube circuit the plate resistor should have a power rating sufficient to withstand the power resulting from the voltage across the part during normal operation. Say Ra=100k and the voltage across it is 150V. Ohm's law and the Power equations suggest a dissipation of 225mW. The instantaneous power and voltage will be higher and lower than the quiescent values, and we are already too close to the maximum for a 250mW part, so go to 400-600mW.

At 100Va the voltage across RA may be only 50V or less, so a 250mW resistor may be ample. You can figure out the R value where you would have to switch to a higher-rated component by dividing the safe idle power into the voltage-squared. For example,  (50x50) / 200mW = 12k5. Pick any power rating that seems reasonabe for the component in question.

All of the other positions in a tube preamp circuit can be 250mW quite easily as the voltage across these is less than a few volts.

The other parameter to check with your resistor purchase is the voltage rating. Yes, resistors have voltage ratings and hardly anyone notices unless they read TUTs Smile

Also, remember that electrolytic cap ratings are WORKING voltage, which means they can be operated reliably and nearly forever at that value. Doing so actually keeps them healthier than derating them.

Have fun
Hey,
Thannks for taking the trouble to answer my questions. If anything it's taught me that sticking to JFETs and BJTs are probably my best bet, unless I can find someone who knows what they're doing to hire for the design work  Tongue
Cheers
S
Hey Man

Why give up so easy ? Tube circuits are easy to learn and you can't really damage much in them specially just a pre. I'd follow Ozzy's advice and Just do it man Big Grin If I could install Power Scaling into my Marshall I figure anybody can do it. I did some mods before that big job and learned a lot about how the pre and power amp work and that was at 500V. 100V is a peace of cake

happy new year