12AX7 and 12AY7 12 volt heater

[67polara]

I read on a data sheet somewhere that when running the 12AY7 on 12 volts it should not be A/C because of hum considerations.  Is this true and why would it make a difference?  Is this true for the 12AX7 also?

Tony

[FYL]

You may use AC: heater current is halved, radiated fields are much lower and compensate the unbalanced supply.

[bigsbybender]

I've run them in close quarters at 12.6v AC without noise issues.  I also have had a couple of old tube Reel-to-Reel tape machines that ran that way, I didn't notice heater noise.

j.

[67polara]

Thanks my transformer is 12.6 center taped so I shouldn't have a problem right?  Or should I make a false center tap like we do the 6 volt.  I figured I would just ground the center tap.  I could also use 6.3 using the center tap and 1 leg and make a sudo ground I think, its marked 6.3 0 6.3 with a bracket over the top stating 12.6v.  I thought I ordered a 6.3 volt filament transformer W CT, guess I did it wrong LOL.

Tony

[PRR]

> a data sheet somewhere ... 12AY7 ... the 12AX7 also?

Where somewhere?

Anyway: the 12AX7 was made for analog computers. NObody needs that much gain at audio. When they did, and hit noise (both hum and hiss) problems, the 12AY7 was introduced and app-noted for best audio performance. "Best" is often more than we need, and even RCA didn't run their 12AY7 with datasheet loading. At guitar level, sufficiently obsessive heater wiring gives low-enough hum.

Given a choice, I think voltage matters more than current and 6V ought to be quieter. It does depend on the circuit layout and impedances. There surely will be cases where 12V "can" be quieter.

[CircuitButcher]

Posted by: PRR

> a data sheet somewhere ... 12AY7 ... the 12AX7 also?

Where somewhere?

Anyway: the 12AX7 was made for analog computers. NObody needs that much gain at audio. When they did, and hit noise (both hum and hiss) problems, the 12AY7 was introduced and app-noted for best audio performance.

Always informative PRR - not knowing the genesis of the 12A*7 dual troide, what precipitated the audio designed 12AY7 to be replaced by the 12AX7, rationalization of production?

[67polara]

Here is where I read it at the bottom of page one.

http://www.drtube.com/datasheets/12ay7-rca1953.pdf

Tony

[HotBluePlates]

Here is where I read it at the bottom of page one.

http://www.drtube.com/datasheets/12ay7-rca1953.pdf

Tony

Ahhh... That 'splains it!

Right above this qoute is the basing diagram. When you operate with 6.3v, you use pins 4/5 (tied together) and pin 9. That is a humbucking filament connection. 12.6v wiring uses pin 4 on one side and pin 5 on the other; pin 9 remains unused. That is not "humbucking". So RCA wanted to cover themselves by suggesting the humbucking connection be used when you want the lowest possible hum.

Which doesn't really mean 12.6v will result in hum. But you will not be using this tube in a super-low-level signal stage anyway within a guitar amp. You might see super-low-level in something like a voltmeter for very small a.c. voltages (less than 1mV). When building something like that, the engineers (if the final meter price can justify it) will attack every possible hum source.

For example, Hewlett-Packard introduced a vacuum-tube meter which could measure very small d.c. voltages. A.C. hum might upset the circuits doing the measuring, so they regulated the power supply. 60Hz hum might still leak in and mess up the process, so they hit on an idea to convert the incoming d.c. to light, which was chopped mechanically at a rate that was unrelated to the line frequency or any harmonic of the line frequency (call it a.c. light). Amplification of the small signal could happen of an electrical a.c. signal created from the chopper, then reconverted to d.c. at a higher level (relatively immune to noise). The reconversion Process responded only to the unique frequency created by the chopper, and strongly rejected any hum component or harmonic.

We have no call for such extreme measures in a guitar amp. And RCA's note is more targeted to those folks who might have cause to take such extremes.

[RicharD]

>Ahhh... That 'splains it!

Well HEY stranger!  Where ya been hiding?

[PRR]

> vacuum-tube meter which could measure very small d.c. voltages.

The chopper's essential function was to eliminate DC drift.

Do you have a real VTVM? Put it on DC, short the leads, and trim it to dead-zero. Leave it on. Check it morning and night for a week. There's 100mV drift in the first hour, maybe 10mV/day and 10mV with wall-variations. That's why the lowest scale is commonly 1.5V.

Do you have a tube analog computer? (Few do; you just might.) Then you know every op-amp (unless you have choppers or auto-zero amps) has a 1.35V Mercury cell and a zero-trim which must be set before every run, or at least every 10 minutes.

There's a lot in the MIT RADAR Amplifiers book about tube drift. You just can't get rid of it. All-day 50mV zero-error is darn hard. Auto-zero for short-run and choppers for long-run or unattentive operators is the answer. (When good clean transistor pairs came out, one feature was 10X or even 100X lower drift.)

That a chopper also can be rigged to reject hum is a bonus. In most cases there's cheaper ways to get hum down.

Good to see you back.

[HotBluePlates]

Do you have a real VTVM?

A bunch of 'em. Too many, including ones using the HP approach mentioned, but also others like the RCA WV-77 which need the zero checked/trimmed before every use.

Do you have a tube analog computer? (Few do; you just might.)

I've seen and read about them; if I had one, I probably would've already ripped it apart to build into something else. To me, the electro-mechanical analog computers using synchros and servos (for aiming battleship guns) seemed much more interesting.

Well HEY stranger!  Where ya been hiding?

Hey! I've been busy getting moved and onto the new assignment with the Army. It has been pretty hectic...