Question on Low Voltage Tubes

[bacanador]

Hello Everyone

My name is Jose, this is my first post on this forum, been reading it for some time now.

Recently I decided to start playing with tubes, and to avoid the High Voltage DC levels normally used for tube circuits I started doing some reasearch about using low voltage. Found something called space charge tubes which were designed to work with car batteries, and bought a few to experiment.

I did my homework read the datasheets, did my calculations and started testing circuits, and my calculations were totally off. I found that using tubes at low voltages causes a self biasing effect which will put the grid of the tube at a negative voltage without the need of a resistor in the cathode.

In My case I had a 12 V DC power supply, a 1M resistor on the grid tied to ground a 1k resistor on the cathode and started changing values on the plate resistor until I got a voltage of about 6-7 V DC. turns out that resistor was 1M, which means the plate current is very low because of the self biasing effect, so I removed the 1K resistor on the cathode and tied the cathode to ground. Tested this circuit and it works and amplifies fine, but the valve is almost turned off.

My question is for anybody that has worked with low voltage tubes, and has experience with this stuff, how can I turn on the valve a little more?, the grid is already at -1 V and I want to be able to operate it at -0.3 or -0.4 to get more headroom for amplification.

I don't know if the same thing happens when you use a normal tube such as the 12AX7 at low voltages, If it does what is the solution?

Thanks in advance

Jose

[amptramp]

Welcome to the forum!  Sounds like you have a nice experiment going on there.

The low voltage tubes still have the same effect going on as other tubes: electrons are expelled from the cathode where they land on the grid, driving the grid negative.  With a space-charge grid between the cathode and control grid, this effect is increased because the electrons are accelerated by the positive voltage on the space-charge grid which is the one closest to the cathode.  To reduce the control grid voltage, reduce the value of the grid resistor and the voltage should vary directly with resistance.  You may have to increase the coupling capacitor to the grid to make up for the higher low-frequency turnover point.  The space-charge grid is normally operated at 12 volts and you may want to vary this just to see what happens.

A 12AX7 will have some of the same effect, but it does not have a space-charge grid, so the effect is not as great.  Ususally you can use a 1 Meg resistor and from 1 to 1.5 Kohms for a cathode resistor and a 12AX7 is properly biased.  The old 12AV6 and 12SQ7 that were used in the millions of 5-tube radios built from the 1930's to the 1960's used a 10 Meg grid resistor for grid bias and no cathode resistor.

[petemoore]

 Might want to try...er...whats the max plate voltage?], something closer to that.

[Renegadrian]

thx for reporting! Which kind of tubes are you using!? Some were designed to have 12V both on the heaters AND the plates, some could handle higher voltages at the plates (say up to 90V). The first kind should work good at 12V, I guess...

[bacanador]

Thanks for the info, Will try your suggestions, the problem is that I didn't want to lower the input impedance of the tube stage by using a lower resistor, will try with a higher plate voltage.

The tubes I've been playing with are 12AE6A, 12AJ6, 12AE7, all present the same issue, all are meant to work with 12V on heaters and the plate, maybe I can try using a charge pump dc to dc converter to bring the plate voltage up to 24 on the plates and that will help.

Again guys, thanks for the useful info, will post my results if anyone is interested, I want to try these tubes for a overdrive pedal. but first want to figure out how to use them correctly

Jose

[petemoore]

  maybe I can try using a charge pump dc to dc converter to bring the plate voltage up to 24 on the plates and that will help.
  A thrift store printer [or curb printer]...often comes with a 24vdc or 36vdc power supply, basically safe, probably noisy, but for safety [no mains wiring needed] testing with one will put a larger potential on the plate[s.
  Mind the current draw / components ratings...
  Use power supply filter or consider subtracting more than a few volts from your supply [to run the regulator, they require/drop 2volts] and choose a regulator # [say 7828, 28volt regulator] to take the ripple out of the DC.

[Renegadrian]

http://www.nj7p.org/Tube4.php?tube=12AE6A
http://www.nj7p.org/Tube4.php?tube=12AJ6
http://www.nj7p.org/Tube4.php?tube=12AE7

Max plate V is 30V

[amptramp]

For further explanation:

http://www.junkbox.com/electronics/lowvoltagetubes.shtml

This is a great project site that describes all the features of low-voltage tubes and has projects designed around them.

By dumb luck, I was at a garage sale today where the owner had a 1960 Buick car radio for sale.  Well, wouldn't you just know it, it used 12-volt tubes of the type you have including the 12K5, a driver amp designed to drive the push-pull germanium transistor output stage.  (Power transistors in a TO-3 case, not Fuzz Face material.)  So I plunked down my $5 and got it.  These tubes have some non-linearity for strong signals, so you could make a Fuzz Face equivalent that is not sensitive to temperature nor requires select-on-test parts.

[petemoore]

  Did you try the "Buick" amplifier ?
  What did it have in it to DC block the output ?

[amptramp]

Just got it today, haven't had time to do anything with it yet.  I intend to find or trace out the schematic so I know what I have.

[PRR]

> Found something called space charge tubes
> been playing with are 12AE6A, 12AJ6, 12AE7

12AE6 is NOT a Space-Charge tube.

It is a plain simple 300V triode (nearly a 12AU7) slightly optimized for low-volt use.

> started changing values on the plate resistor

No. The tube must drive a Load. Start from there. 220K is typical in high voltage gitar amps. Hint: for this lower voltage, aim at 100K. Pick a tube with show-off plate resistance spec 4 to 25 times smaller. This gives 100K/4= 25K to 100K/25= 4K. 12AE6 boasts 13K-20K and may be suitable. Pick your plate DC resistor 2 to 5 times smaller than that AND 2 to 5 times larger than the tube's plate resistance. Values of 50K to 30K seem useful. Take 47K as standard value.

Omit grid and cathode resistors for now. A large negative grid voltage should drive the plate very near the supply rail. Groundng the grid will flow the maximum useful plate current and lowest useful plate voltage. This will not be the negative rail but something higher. Maybe 3V in this case.

Split the difference. If the extremes are 3V and 11V, aim for 7V idle point so the audio can swing both ways somewhat equally.

Add a cathode resistor and trim to get the plate to around 7V. Reading the 12E6 data, I squint 0.6V and 0.1mA, the cathode resistor should be near 0.6V/0.1mA or 6K.

Now un-ground the grid and insert a 1Meg resistor (the 12AE6's rated max for self-bias). AND bypass it to ground with a good cap; 0.1uFd will probably be OK.

If you don't AC-ground the grid with a cap, the sensitive grid will suck-up all the hum, buzz, and radio in the room. If it goes 0.6V higher than cathode it is clamped by the grid, but strong signals may go any amount negative. This asymmetric garbage drives the grid negative and drives plate current toward zero. (This may be the flaw in your breadboard.)

Also ground the 12AE6's diode plates. In most cases this is not essential, but good practice. In a low-volt case it might matter.

The 1Meg resistor "should not" affect bias. That's what the spec means: any value from zero to 1Meg will be all the "same". It is normal to find a small negative voltage with large grid resistor; but the spec value should give such a small change that you do not need to care. If it makes a difference: cook the tube. A long-stored tube has stray gas. The getter (silver spot) will absorb this but only when hot. A very gassy tube may take days to clean-up. This tube should run 10,000 hours; cook it for 4 days straight (100 hours). Heater power alone is sufficient. Leaving plate power on won't hurt.

Sylvania's 12AJ6 sheet shows a resistance-coupled amplifier plan. The suggested 1Meg plate load is rather high for anything except driving a nearby power pentode grid.

12AE7 is quite interesting for another project.... thanks.

> a 1960 Buick car radio for sale.

Find the top of the volume control. Run guitar through booster to here. Suss-out the speaker wiring (it may not be as expected). It may want a full 5 Amperes of 12VDC. It was a good radio, it may be a good music-amp.

[bacanador]

> Found something called space charge tubes
> been playing with are 12AE6A, 12AJ6, 12AE7

12AE6 is NOT a Space-Charge tube.

It is a plain simple 300V triode (nearly a 12AU7) slightly optimized for low-volt use.

> started changing values on the plate resistor

No. The tube must drive a Load. Start from there. 220K is typical in high voltage gitar amps. Hint: for this lower voltage, aim at 100K. Pick a tube with show-off plate resistance spec 4 to 25 times smaller. This gives 100K/4= 25K to 100K/25= 4K. 12AE6 boasts 13K-20K and may be suitable. Pick your plate DC resistor 2 to 5 times smaller than that AND 2 to 5 times larger than the tube's plate resistance. Values of 50K to 30K seem useful. Take 47K as standard value.

Omit grid and cathode resistors for now. A large negative grid voltage should drive the plate very near the supply rail. Groundng the grid will flow the maximum useful plate current and lowest useful plate voltage. This will not be the negative rail but something higher. Maybe 3V in this case.

Split the difference. If the extremes are 3V and 11V, aim for 7V idle point so the audio can swing both ways somewhat equally.

Add a cathode resistor and trim to get the plate to around 7V. Reading the 12E6 data, I squint 0.6V and 0.1mA, the cathode resistor should be near 0.6V/0.1mA or 6K.

Now un-ground the grid and insert a 1Meg resistor (the 12AE6's rated max for self-bias). AND bypass it to ground with a good cap; 0.1uFd will probably be OK.

If you don't AC-ground the grid with a cap, the sensitive grid will suck-up all the hum, buzz, and radio in the room. If it goes 0.6V higher than cathode it is clamped by the grid, but strong signals may go any amount negative. This asymmetric garbage drives the grid negative and drives plate current toward zero. (This may be the flaw in your breadboard.)

Also ground the 12AE6's diode plates. In most cases this is not essential, but good practice. In a low-volt case it might matter.

The 1Meg resistor "should not" affect bias. That's what the spec means: any value from zero to 1Meg will be all the "same". It is normal to find a small negative voltage with large grid resistor; but the spec value should give such a small change that you do not need to care. If it makes a difference: cook the tube. A long-stored tube has stray gas. The getter (silver spot) will absorb this but only when hot. A very gassy tube may take days to clean-up. This tube should run 10,000 hours; cook it for 4 days straight (100 hours). Heater power alone is sufficient. Leaving plate power on won't hurt.

Sylvania's 12AJ6 sheet shows a resistance-coupled amplifier plan. The suggested 1Meg plate load is rather high for anything except driving a nearby power pentode grid.

12AE7 is quite interesting for another project.... thanks.

> a 1960 Buick car radio for sale.

Find the top of the volume control. Run guitar through booster to here. Suss-out the speaker wiring (it may not be as expected). It may want a full 5 Amperes of 12VDC. It was a good radio, it may be a good music-amp.

Thanks for the useful info.

The problem is that as soon as you add the grid resistor, all changes, the grid becomes negative and the valve is already biased, so I can omit the cathode resistor and the thing works.

With the 12AE6A tube (which seems to be a different tube), I have a plate resistor of 220k, cathode directly to ground, and a 1M resistor on the grid. Looking at the signal at plate resistor, I have about a gain of 6x, the waveform is not clipped.

I decided to try higher voltages, using a 35 V supply, the negative voltage on the grid disappears, and works more like a what I would expect. I wanted to keep it simpler but this might be the way to go, or use the previous circuit without the cathode resistor. I guess a lot of experimentation and listening will decide.

I found designs on the web that seem to use this self bias  effect on low voltage tubes

http://www.geocities.jp/dads_projects/radio/12K5.html

The page is in japanese but the schematic is very simple, and as you can see the first valve has no cathode resistor, my guess is that it is because of the self biasing effect.

[user]

This is confusing to me. Im a noob myself. But 12AU7 isnt a space charge tube either. Then how come it has been utilized for the Valvecaster project, running at 9-12V?

[bacanador]

I have never done it, but there are a few designs out there that use regular tubes (such as the 12AX7) on low voltage circuits. Some people complain that they don't sound good tho, but I have heard some sound clips of the Shaka Tube (you can find it somewhere on this website) and I like it.

[Renegadrian]

This is confusing to me. Im a noob myself. But 12AU7 isnt a space charge tube either. Then how come it has been utilized for the Valvecaster project, running at 9-12V?

Space Charge were developed to be used wit h a car battery (so 12V) - other tubes run much higher voltages as you know.
How much?! Well, it's not a rule of thumb...So the Valvy was born from experimenting, as other projects too...
I had good results with AU7 and AT7, AX7 didn't give me a decent sound, altough some reported it as a good tube in a Valvy.

[bacanador]

Well. the problem is solved on the 12AE6-A tube, with plate voltages higher than 20, the self biasing effect is gone. However the 12AJ6 still presents the same issue. I guess the higher the mu of the triode the problem is worse (haven't tried any other high mu space charge tubes just guessing).

Now about using regular tubes at 9V, why are these criticized so much all over the place, found a few posts on here, and others on other forums saying that the sound is not good.

How does this valvecaster sound?

I heard the sound samples of the shaka tube, and I really liked the tone, the playing of course helped. In that same thread I saw a criticism about running the 12AX7 at a low voltage.

[Renegadrian]

My humble experience and experiments with tubes at low voltage...
Valvys sound GREAT! a good booster/overdrive. and the use of the tube at 12V is not a limit for the 12AU7 inside. Some other tubes work good at low voltages, even at 12V. Some are good only at slightly higher voltages (I have a multiplier that goes to 50V) and up to high V territory. I consider 12V to be the standard with tube experiments, as a common start point of the whole LOW VOLTAGE TUBE CLUB is the well known 12AU7. Not that the tube is gonna die if you feed 9V to it, but as the heaters are intended to receive 12V (and those 3V more on the plates are welcome) and a variable voltage WW is a weapon of choice, I'd reccomend 12V all the time. Another thing I noticed is that A LOT of the ending result depends on the tube used - not the type, but THE tube - I have several 12AU7 and you can bet your mother every one of them has a different "voice".
So among several tubes of the same kind, some may work good even at 9V, some will not (as reported). Also some 12AX7 will work ok at 12V, some will not (as reported). My experiments with the AX7 didn't make me scream in tube heaven, but I cannot tell that is IMPOSSIBLE to have good results.

So my notes above are not "criticisms" but "tales from the road" as I dealt with tubes before and tried new ways.
It's up to you (and everybody) to try some more and share your experiments!
Don't believe what the others want you to take for granted!!!
If we had to follow rule of thumbs, no Valvecaster would exist...

[bacanador]

Thanks for the valuable info. I am going to get some 12Au7 tubes and start experimenting