12AX7 max input signal voltage ?

[Colas LeGrippa]

Hi, I am experiencing a new problem. Class A amp, 2 X 12ax7 and 2 X EL34 in parallel. First 12ax7 is two triodes in cascade. Second 12ax7 is a gain stage and a cathode follower. Following ? If I feed the third gain stage with the signal right from the second plate, the sound cuts shortly after emitting crackling sound. The cure ? A voltage divider . Seems that the signal voltage is too high ( 5o vac ) for the second 12ax7 to do its work. Cant find specs under 12AX7 MAX GRID VOLTAGE.
What would be in general the max grid input voltage for a dual triode ?
 
Colas

[frus]

did you forget to include a grid resistor (usually 1M) from third grid to ground (when there was no voltage divider yet)?


anyway, you could also try putting a large grid stopper and/or a smaller coupling cap between 2nd and 3rd stage.

[Colas LeGrippa]

Hi Frus,
 
I have rarely seen a grid leak resistor at this emplacement, most common is a voltage divider. Which is the same thing in fact, the second resistor of the voltage divider acts as a grid leak resistor. I will try taking off the first one and see what will happen. My question though concerned the max AC voltage ( signal ) that could be present at a grid, and no data sheet talks about it.
 
Muchas gracias !
 
Colas

[SILVERGUN]

If I feed the third gain stage with the signal right from the second plate, the sound cuts shortly after emitting crackling sound.

Do you have a coupling cap between the plate of the second stage and the grid of the third stage?

Max positive DC V on grid is 0   

[Colas LeGrippa]

Max positive DC V on grid is 0

 
False.  The grid of a cathode follower is fed directly by the plate of the preceding stage.

[SILVERGUN]

Max positive DC V on grid is 0

 False.  The grid of a cathode follower is fed directly by the plate of the preceding stage.

Sorry Colas,
I was referring to the cathode biased third stage that you were feeding the signal to, not the cathode follower stage...

And, just trying to be helpful,,,no disrespect intended.
If you had no coupling cap, you'd get cutting out and crackling sound.

[SILVERGUN]

First 12ax7 is two triodes in cascade. Second 12ax7 is a gain stage and a cathode follower. Following ? If I feed the third gain stage with the signal right from the second plate, the sound cuts shortly after emitting crackling sound. The cure ? A voltage divider . Seems that the signal voltage is too high ( 5o vac )

The amplitude of the AC signal voltage going into the third stage would "normally" be below 5Vp-p at that point in the signal chain, so the fact that you're seeing 50V is a problem in itself...are you seeing this on a scope?

A max. AC voltage rating is probably never given because it is not a practical guideline....the bias conditions of the triode determine the acceptable input signal swing, and if you exceed it, you just push the tube into cutoff.

There are more technical explanations, from smarter guys,,, but that's as simple as I see it.

[HotBluePlates]

... Cant find specs under 12AX7 MAX GRID VOLTAGE. ...

Because this is never specified. You find it by looking at the plate curves and plotting a loadline.

But look at the 12AX7 curves below. See the G₁ = -5v line? It runs into the 0mA horizontal axis right near 400v on the plate.

You should interpret that as "If I have 400v plate-to-cathode on my 12AX7 and apply -5vdc to the grid, I will shut the tube off." If the input signal should be able to swing as far up as down, that implies 5v/2 = 2.5v peak or ~1.77v RMS. For most gain stages, you will not have 400v of supply voltage (or maybe even 300v) which will make the input limit before cutoff smaller.

What would be in general the max grid input voltage for a dual triode ?

There is no one maximum for all dual triodes. For other tubes, look at the curves and not that the G₁ lines are in increments of 2v or 5v or more. You will notice a different G₁ voltage for cutoff at ~400v, which tells you the tube could handle a larger input signal.

You will find that tubes with less gain (12AU7) can accept bigger input signals than tubes with higher gain (12AX7).

You will find you can also booger the bias and make a 12AX7 with 400v plate-to-cathode only accept a 20mV signal before plate current cutoff (too big a cathode resistance, too low idle current).

[Colas LeGrippa]

And, just trying to be helpful,,,no disrespect intended.

it was not my intention either silvergun.

[printer2]

philips1954 datasheet

 -Vg = maximum 50V.
Vg(Ig=+0, 3uA) = max -1,3 V

http://www.drtube.com/datasheets/ecc83-philips1954.pdf

[SILVERGUN]

it was not my intention either silvergun.

 

[HotBluePlates]

philips1954 datasheet

-Vg = maximum 50V.
Vg(Ig=+0, 3uA) = max -1,3 V

Okay, there are those. However, these don't speak to maximum input signal which is what Colas wanted. So what do they describe?

The -50v maximum is provided in the same section as the 550v maximum plate voltage, so this is an absolute value of voltage which is never to be exceeded under any circumstances. As I showed with the 12AX7 curves above, with 400v across the 12AX7 the tube cut off with -5v on the grid so the extra -45v that could be applied before reaching the absolute maximum don't really matter, do they?

What about the -1.3v max? Why have a maximum of V_g = -50v and also V_g = -1.3v? Notice in parentheses it says "I_g = + 0.3uA". That means grid current is 3 microamps, which is close enough to "zero grid current" for you and me. If the grid is well-negative of the cathode, the grid looks to external circuitry like an open-circuit and does not flow any current. But this particular data sheet entry tells you that once the grid is 1.3v below cathode voltage, that grid current begins to flow; it will increase as the grid approaches the same voltage as the cathode (V_g=0v) or becomes positive of the cathode.

So if anything, you might interpret the -5v cutoff point I showed earlier and the -1.3v point for the onset of grid current to 3.7v peak-to-peak maximum possible input signal. But in reality, many circuits run the tube closer to V_g=0v because grid current stays small unless the grid goes positive of the cathode.

See how those don't directly state "maximum grid input"? At least not like you might think?

[jojokeo]

I wonder what the reasons are to even want or allow signal to get so close to the max input signal in the first place when it's not necessary? But I realize that Colas was having an issue so thinking he may be into cutoff as a possibility, just making a rhetorical statement. Forgetting to add a grid leak or coupling cap can easily happen when creative juices are happening on the fly or when tired after lots of tweaking's been going on. I've never had anything like this happen!