Filament Voltage?

[Structo]

After building my 100w ODS amp I have noticed that my heater voltages average around 6.9 vac.
Sometimes it is 7 vac which is enough to concern me.

This is a Magnetic Component Fender Twin power transformer.
It specs the heater voltage at 6.5v @ 5.5A

I was doing some reading at various places and the general consensus seems to be if it is within 10% you shouldn't worry.

Then I read another article that stated if you keep the voltage down around 6v that you can almost double the life of the tube, at least the filament life.

Because everytime you turn on the current to those heaters it shocks them with a surge of in rush current.

I don't like the idea of resistors to lower the voltage, but I saw on Merlin's site where you can take two silicon diode rectifiers and wire them in parallel with opposite polarities and use the forward current voltage drop to drop from .5 to 1v depending on the diodes used.

[img:154:60]http://www.freewebs.com/valvewizard/heater7.jpg[/img]

So I checked with the local Radio Shack and the closest thing they carry for what I need were some 6 amp 50v diodes.
Of course no other specs.

[img:345:235]http://rsk.imageg.net/graphics/product_ ... 01w345.jpg[/img]

So I soldered those in and they drop about .7 volts bringing the heater voltage down to 6vac.
I was hoping to drop .5v but no go.

But even though my heater draw is 4.5 amps, these diodes get pretty dam hot!

Is there a way with the artificial center tap resistors to drop some voltage?

Anybody here ever adjust their heater voltage?

[ChrisM]

Your putting around 4.5A through the diodes, that's enough to get them warm. Use a higher speced diode, something like 8A or 10A.
That's what I'd do, I don't like using big power resistors either.

[Colossal]

Tom, you could also try a couple of Rs, one on each leg to drop some voltage. You know the current your pulling and the desired voltage so you can calculate the necessary R...

[David Root]

Tom, the Ancients said +/- 10% from nominal range is OK. so 6.9 to 5.7V. But they weren't pushing the crap out of the tubes like we do.

I tend to buy the 6.0V argument, if only because tubes were so much cheaper in those days. Let me qualify that, tubes that we now call NOS were a lot cheaper then!

Let me further qualify that as applying only to preamp tubes. I would not do that with power tubes, because there is some heavy lifting involved here, electrically speaking. I have found, for example, that a good BF Twin PT or BF Bassman PT will run 6.7V quiescent with only 3 or 4 12AX7s, as in a Dumble.

I've not tried to measure it, but do power tubes sag the heater voltage when you crank 'em?

Bottom line, I think you can safely starve a 12AX7 of heater voltage, but I wouldn't do it with a KT88! Plate voltages and bias points notwithstanding.

[Structo]

Yeah, I'm not sure if the heater voltage sags at higher volume.
My guess is no on this one since I am running 3 12ax7's less than a Fender Twin.

So I'm pulling 4.5 amps on the heaters.
I haven't measured that, just going by specs.

I've kind of been concerned about this since I do run some vintage preamp tubes, I would just as soon have them last a long time.

I know that I am at the upper end of nominal at 6.9v.
I've been doing a lot of reading about it and a good article was the one that somebody posted called, shoot, I can't find it right now.
Something like, Everything you ever wanted to know about vacuum tubes or similar.
Pretty nice PDF book.

[67plexi]

Tom, you could go to the next level Regulated DC heater supply 5 Amp.
Or just add resistors in series with the transformer output to get 6.3v AC.

Steve.

[phsyconoodler]

When was the last time you had a filament failure in a preamp tube?

I can't remember myself.The voltage is NOT a concern.So they last 6,000 hrs instead of 10,000 hrs.
If you use the amp three hours a day it will take 5.5 years for those filaments to go bad.Hmm.....I have some 10 year old Sovteks that still work and my filament voltage is a constant 7v.
I believe it's over blown about a little bit of extra voltage.

[Guitarman18]

Tom,

This has helped me out in the past.


http://www.tubecad.com/july2000/


Hope it helps,

Paul.

[Structo]

Thanks, that has some good information.

I realize I may be over analyzing this but it's been bugging me, plus I get to learn along the way.

To use resistors to drop the voltage.
My calculations come up with 2 resistors at .1 ohm

Now my amp with 4x 6L6 and 3x 12ax7 draws 4.5A

So for the wattage I think it would be P = R x I2 which gives me 2.025 watts

That would mean two .1ohm 5 watt resistors correct?

The bonus with using resistors is the in rush resistance on cold filaments.

Another thing I was looking into is using a NTC thermister.
But finding the correct value seems to be a bit troublesome.

[Structo]

Good article on inrush current.

http://www.sound.westhost.com/articles/inrush.htm

[Cliff Schecht]

In my experience the heater voltages do not sag under heavy loading. If they do sag more than a few percent then you are using an underrated power transformer.

For dropping heater voltage, I've really come to prefer the power resistor method. I don't like shoving diodes in, they clip the AC and can reduce balance, which increases noise. Even in a balanced configuration, clipping the diodes produces EMI and adds more "hash" to the amp. This can be settled by adding snubbing caps to the diodes, but now your parts count has doubled.

Resistors don't have these problems. If you calculate the required resistance, half this value and add that value resistor to each leg of the heater you maintain the balance and don't add any distortion to the waveform. In essence, you can better maintain the balance in the heaters and get better noise cancellation. The other nice thing about using the split resistors is power resistors have sturdy leads that can be reliably "floated" in an amp without having to worry about breaking. Most high current diodes have these bigger leads too, but the diode/lead junction is much more sensitive than the resistor/lead junction, especially after anything past minimal soldering heat.

[talbany]

The cool thing about the 100 ohm resistors on the filament supply ( false CT) is they can act as a kind of fuse.. If for some reason as say a heater cathode short and the supply starts to draw too much current you will fry the resistors instead of the PT.. It's never a guarantee but does add some protection vs say a grounded CT supply will.. Thought I would mention for those not aware..As long as I have those resistors in there I don't sweat the extra voltage.. Especially in a Beam power like 6L's..
You can also stick a fuse in the supply as well..I think CE certification requires it

Hope this Helps!!

Tony

[hitchcaster]

factor in PT sag when pushed hard.. and wall line voltage changes too... maybe watch the moving target in different environments before you feel you want to mess with it... just my .02

[Structo]

Just to update this.
I didn't feel comfortagle with the two hot plastic diodes in the heater circuit so I took them out.
Touched up all the solder joints in the chain and the heater voltage now hovers between 6.6v and 6.7 which is fine by me.

I like the idea of resistors there to help with the surge but I don't like the idea of components that get too hot to touch inside the chassis.

A thermister sounds like the ideal thing with it's temperature coefficient properties but again, they get hot. And to find the right value would take a lot of experimentation.

So I will fly with the wind with this one.