12ax7 as a diode

[Andy Le Blanc]

I found an old electronics course school text book at a junk store. Its from
the late sixties, and is very rudimentary. Some of the demonstration circuits
are crystal sets to illustrate radio theory and history, and then it replaces the
crystal with a tube rectifier and begins there to introduce active "electronics"
with a brief discussion of the "Flemming Valve" leading to DeForest's triode.
The tube they use is a 6av6, and instruct you to connect the grid to the plate
to get the tube to behave as a "valve diode". A 6av6 is a seven pin, single
triode unit that is an equivalent to half a 12ax7. Being unable to resist, I dug
around in my bone pile and found a carcass with enough working bits to wire
a 12ax7 as instructed to see what happens and yes you can make 12ax7 work
as diode, revealing the basic function of tube theory ETC......

One thing that stood out about the exercise was just how imperfect a diode it
is when a triode is wired this way. There is a very large and measurable
capacitance, even with a crap meter, large enough to surprise. And, just the
simple function of the valve, creates a standing voltage of around 1.2v
across the valve. Just with the heater wired and no other connection other than
the grid being connected to the plate. Neat.

What could you use it for?, other than a detector for a radio set.

[Structo]

Interesting.

I sometimes wish I had been born 20 years earlier.

During my EE course at a community college the instructor skimmed right over tube theory telling us that semiconductors will replace all tubes in a few short years. This was around 1976.
So we looked at tubes for a couple days then moved on to better things....

I still remember going with dad with a sack full of tubes from our black and white TV to the corner drug store and using their tester trying to figure out which one was bad.

[Wayne]

Andy - if you're so inclined and haven't torn down your test jig yet - I wonder if it would work better with the grid tied to the cathode instead. Seems to me with the grid tied to the plate the grid would experience some current, and in a 12ax7 that ain't good.

I thought about this before and never followed it through. A beefier preamp tube (12au7, maybe?) might make a nifty recto for some low powered amp (I'm thinking bedroom power level).

Just a thought...

W

[Andy Le Blanc]

Thank you for the response, hadn't gone that step yet.
The voltage present is still the same regardless, but the two hook ups do
show different behaviors. With the grid tied to the plate, the resistance between
the plate (and grid), and cathode was too small for my meter, but a 1n capacitance
is present across the the valve.
When the grid is tied to the cathode, there is next to no capacitance but there
is around a 900 ohm resistance thru the valve. Its a two for. Someone with
truly accurate gear should measure the capacitance. I'd be reluctant to try
this for power supply rectification but any other small signal application will be
neat to explore.

[gahult]

I still remember going with dad with a sack full of tubes from our black and white TV to the corner drug store and using their tester trying to figure out which one was bad.

I remember waaaaaay back when I was a kid going through trash behind tv repair shops digging for tubes and spare parts chassis and spending time at the drug store tube tester for something to do.
I think I was lucky to never run into a charged power supply cap. ouch.
Not with dad though as he wasn't into electronics.

It was fun and back in that time.
We didn't wear helmets and knee pads out to play.
Amazing how many of us made it through childhood.

Funny how those things pop out of the old memory banks when reminded.
Thanks

Gary

[skyboltone]

I've got a couple of sleeves of 6AL5 that are designed for this process. Detectors. I've noodled using one as a mixer tube. Tie the plates together, apply plate voltage, bias the tube with a cathode resistor on each cathode sized for 60 percent of max tube dissipation. Then feed the two capacitor coupled sources to each cathode and take off the sum at the plates with a coupling cap.

Haven't done it yet.

[Andy Le Blanc]

I found I have an old 6h6 metal type, after seeing the last post. There are
probably gobs of twin diodes in storage. A good way to get the sand out of
some projects. 12ax7 are so common, its a good way to explore, with nothing
more than scunge that's left over on a bench. You can make a diode, try out
grounded plate, cathode, and grounded gird as well as various transformer
output circuits with just a handful of parts. A tube diode would be neat to try
on a screen grid circuit, a clipper circuit, and any other small signal application.

[Wayne]

I'll bet a lower mu tube would exhibit less capacitance with the grid connected to the plate. Just a hunch, based on what I think I understand about how grid spacing affects mu. YMMV.

A now long retired radio op told me once about a transmitter they had installed in a remote location. Everything from the power supply through to the output used the same tube. It sure would make packing for a maintenance trip easier!

W

[Andy Le Blanc]

A bit more of interest, I built a simple circuit that normally uses a LED to provide
a non-linear bias, and replaced the LED with a 12ax7 wired as a diode.
The 12ax7 provided the same function when the grid is connected to the plate.
The circuit works with the anode of the LED tied to ground, with the converted
tube, the function is reversed, with the cathode tied to ground performing as
the anode compared to the LED.

A 12ax7 converted to a diode with the grid tied to the tubes cathode did not
provide a similar function the the LED.