Upping the heater voltage on a starved plate preamp tube

[Deude_Band]

So I have been reading up on starved plate preamps used for overdrive pedals, such as the Behringer VT999. I have read the debates about how not running the tube at the intended voltages (200-300V) compromises the sound quality. I have also listened to the differences in some of the starved plate and high voltage pedals (like the Blackstar HT pedals).

Here is my question... In the preamp tubes like a 12AT7 it looks like the heater plates are electrically isolated from the remainder of the tube. Is there any sonic benefit to running a separate, higher DC voltage to the heater circuits?

For example, take a Behringer VT999. The heater wires could be isolated from the rest of the circuit and a DC/DC converter or the like could be used to increase their voltage so they run hotter. I am assuming of course that this would have no effect on the rest of the circuitry.

Thanks in advance.

[mozz]

If you run it over 10% higher than 6.3v, such as 6.9 or more, you run the risk of burning the tube out, or drastically shortening it's life. I don't think you are going to gain anything. I would try 12au7 or 12at7 for starved plate designs, they work much better than a 12ax7. There are multiple variations of those tubes to try for cheap: 5963,5965, 6201,6211, 12av7, 12az7, others.

[Deude_Band]

OK thanks that answers my question.

[merlinb]

[Deude_Band]

OK so this brings up a related question... I have seen everything from 6V to 12V for the heaters. However I am confused as to how each is implemented. Correct me if I am wrong here...

1. 12V is used if the two heaters are wired in series without using the ground pin. That is, 12V is wired across pins 4 and 5.

2. 6V (or 6.3V) is used if the two heaters are wired in parallel by using the ground pin. That is pins 4 and 5 get 6V and pin 9 is grounded.

So, are 1 and 2 both correct, is only 1 or 2 correct, or are both incorrect?

Also, I have seen comments that even though both work, method 2 may be preferable since using that scheme introduces common mode noise rejection so that any noise on the DC power supply is reduced. Is that correct?

And finally, if method 2 is used with 12V in parallel on the two heaters, then that's too much power and will burn out the heaters, at least prematurely. Is that correct?

Thanks

[vigilante397]

1. 12V is used if the two heaters are wired in series without using the ground pin. That is, 12V is wired across pins 4 and 5.

2. 6V (or 6.3V) is used if the two heaters are wired in parallel by using the ground pin. That is pins 4 and 5 get 6V and pin 9 is grounded.

That is so very close to correct. The only thing you're missing is that pin 9 isn't a ground pin, it's the heater center tap. It doesn't matter which voltage gets connected where, as long as the difference between the two pins is correct. Heater filaments are like a very low value resistor, they aren't polarized and you can connect the positive voltage to whichever pin is most convenient.

You are absolutely correct that wiring the heaters in parallel and supplying 12V is a bad idea, and your tube will not like you.

The best way I've found to think of tube heaters is as a simple voltage divider. Pin 4 needs to be 6.3V different (higher or lower, doesn't matter) than pin 9, AND pin 5 needs to be 6.3V different than pin 9. An easy way to do this is to put 12.6V (12V is close enough) between pins 4 and 5, and since they are close enough to the same resistance you will get 6.3V on pin 9. Another way to do this is to short pins 4 and 5 and put 6.3V between those pins and pin 9.

The biggest advantage to using 12V is current. On a 12AX7 your heaters will pull 300mA at 6.3V versus 150mA at 12.6V, which is easy to understand if you think about the voltage divider concept.

[duck_arse]

"ground pin" is incorrect.

you have two choices - 6V3 or 12V6.

you have two choices - parallel or series connection.

if 6V3, then parallel. if 12V, then series.

in each case, for DC supply, something gets grounded, something gets supply. current drawn for series will be half the current drawn by parallel.

[Deude_Band]

OK thanks I understand the heater wiring now. Also thanks for the distinction on pin 9 being a center tap and not a 'ground' per se. So for a parallel wiring setup it would get the negative voltage from the heater supply and the two heater pins would get positive voltage from the same supply.

What about the common mode rejection configuration where they are wired in parallel (method 2)? Is there any truth to that? The better approach would be to make sure the heater power supply is quiet, but if I have choice 1 or 2, and 2 is more noise immune, then I might go with that despite the current increase.

[vigilante397]

If you're running DC heaters chances are they're pretty well going to be fairly quiet anyway, so I will continue to recommend whatever is most convenient for your design.

[PRR]

It is two 6V "lamps" on three wires/pins. Since there are two leads per lamp, obviously they have to share a lead, a "common".

Put 12V on a 6V lamp and the life is thousands of times less. For tubes, say an hour. Maybe less: the ~~3X heat is liable to shatter glass seals.

[amptramp]

If you go to this site:

http://diyaudioprojects.com/mirror/members.aol.com/sbench101/

you will see a couple of articles on running tubes with starved filament voltages.  Normally, a tube is run in the space-charge region where there are more than enough electrons between the cathode and the grid that it can support the maximum level of current being demanded of it with some reserve available.

If you reduce the filament voltage, you go from the space-charge region to the temperature limited region where the number of electrons available between the cathode and grid limiits the current that can be drawn.  This region of operation results in more noise even though temperature is a component of the noise equation because you are limited to the shot noise of the individual electrons being emitted rather than the lower ratio of noise to cathode current associated with the greater number of electrons in the space charge region.

The examples at this site show Directly heated cathodes but similar arguments apply to cathodes using separate heaters.  There are graphs of harmonic content for each harmonic and there are some sweet spots at unexpectedly low voltages.  Note also that with directly heated cathodes, the bias on the tube changes with the position on the cathode since there is a voltage gradient along the filament and this tends to change how the tube responds to signal changes versus the tubes with separate heaters and cathodes (like the 12 A*7 series) where all parts of the cathode operate at the same bias.

Generally, best linearity occurs at lower than standard filament voltages because at the high end of the current, it tends to approach current limiting.