Lifting 12ax7 cathode

[EFK]

Is it somehow detrimental to tube life to lift an operating 12ax7 cathode connection to ground? Some of the old Ampeg jet schematics show a tremolo switch between the ground side of the cathode resistor and ground, which I suppose is a good way to kill the trem. However, the tube is still being supplied with B+ so I'm wondering if this is damaging to the tube in some way?

[Chuck H]

It's done all the time with power tubes. I don't see how it should cause a problem at all. My only reservation would be if the Vp rises above max when the tube is not drawing current. Not sure if this could cause internal arcs or not. Obviously it's not a dissapation issue but internal spacing could be. Probably not though.

[mooreamps]

I don't lift cathodes ; preamp tubes or otherwise.. I just think that's a mistake..
I prefer rather to shunt signals to ground via tant caps...

-g

[R.G.]

Is it somehow detrimental to tube life to lift an operating 12ax7 cathode connection to ground? ...the tube is still being supplied with B+ so I'm wondering if this is damaging to the tube in some way?

At the voltages in guitar amps, no, it's not damaging in any significant way. There is a damage mechanism at much higher voltages.

I don't lift cathodes ; preamp tubes or otherwise.. I just think that's a mistake..
I prefer rather to shunt signals to ground via tant caps...

Is there a technical reason for this, or is it purely personal preference, like not liking broccoli or seafoam green wall paint?

[mooreamps]

It's mostly personal preference.. But, I understand the switching of the cathodes on the power tubes of the Mesa Blue Angel has caused some issues.. Also, signal shunting tends to be somewhat noiseless when doing channel switching..

http://www.tubefreak.com/bluangpw.gif

-g

[Wilder Amplification]

What I usually do is put a high value resistor in line with the cathode resistor, then wire up a switch across the resistor. When the switch unshunts it, that elevates the cathode to a voltage high enough to place that stage into cutoff.

[mooreamps]

That's also a fine way to do it.. I'm sure either way works the same...

-g

[Merlinb]

It could lead to cathode poisoning, but there's no evidence currently available to suggest that modern tubes are prone to this.

[Chuck H]

What I usually do is put a high value resistor in line with the cathode resistor, then wire up a switch across the resistor. When the switch unshunts it, that elevates the cathode to a voltage high enough to place that stage into cutoff.

+++ Send the tube into cutoff and keep the 0V reference.

[R.G.]

It could lead to cathode poisoning, but there's no evidence currently available to suggest that modern tubes are prone to this.

That was the issue I alluded to in "at higher voltages". There are two damage mechanisms you can get with odd turnoff mechanisms.

I've read (N.B. - in actual tube textbooks, not just on the internet; the internet is the world's largest source of rumor and innuendo) that at voltages in the 600V and up range you can get cathode stripping from positive ion bombardment. The voltage is important because the stripping mechanism is the energy left by the ions whamming into the cathode. Apparently it proceeds at a minuscule rate at lower voltages, some kind of exponential or threshold effect. The voltages in guitar preamps are uniformly under 400V, so cathode stripping is almost never an effect.

There is a second damage mechanism that happens at low voltages, which was referred to as "sleeping sickness". This was noticed in tubes used in computer logic circuits. It's a migration of high resistance oxides to the cathode surface, where they form sort of an extremely leaky capacitor within the cathode. Changes the way the tube responds to fast signals. Computer uses would sometimes keep one tube turned off substantially all the time, deep into cutoff. This does have the possibility to cause issues if the off-time is really substantial; however, it's unlikely approaching unheard of to have enough hours on a cut off tube in a guitar amp to do this. The old tube computers were run hard and continuously. We'll never get that many hours on a guitar amp.

And for all their flaws, modern tubes are made with purer materials than they used to be. I've never seen any references to modern tubes being prone to it either.

Hey - I ought to see if I could induce sleeping sickness in 12AX7s deliberately, and market them as "pre-conditioned to produce the brown sound of original vintage tubes; this secret process adapts the microstructure of the tube cathode to give the smooth, silky sound of real, old manufacture tubes from the Golden Age... yada, yada, yada..."

[Wilder Amplification]

+++ Send the tube into cutoff and keep the 0V reference.

Exactly. Came up with that when I was adapting a blackface tremolo to a Bluesbreaker reissue. Due to the borrowed bias supply and the high value of the range resistor (150K) the original design that used the 2.2M to feed the fixed bias voltage to the oscillator grid wouldn't work as it would make the bias voltage more positive by a factor of 2 volts every time you grounded out the 2.2M to engage the tremolo, which would cause a HUGE change in idle bias current. So I thought "Why not cutoff bias it via cathode biasing?".

The schematic ended up something like this -

[EFK]

I realize this is a resurrection but I just wanted to offer a "thanks" and state that the elevation to cutoff method is a PERFECT solution to a current project requiring a footswitch without adding noise (i.e., doing things via grid).

[Chuck H]

+1 This isn't a bad thread to re read. For the record... When the tube is in cutoff it's not conducting. So Vp rises to the source voltage, correct? Obviously this happens for 50% of the time when a tube clipping center biased or colder. But for prolonged periods of cutoff, is there any danger exeeding the max Vp while the grid and cathode still have a 0V reference? (he wondered aloud) Or does that not matter when the tube isn't conducting?

[trobbins]

Design centre level for plate-cathode is 300V on the 12AX7 datasheets I just looked at, but the Mazda also indicates 550V at zero anode current.

The heater cathode voltage (180V) and the grid negative voltage (50V) are also spec limits to cause concern if lifting the cathode, so capacitive coupling and stray cathode-heater resistance would split anode=B+ voltage down to a 'floating cathode' and then to ground (which is most likely also = grid with any normal grid leak resistor).

[J M Fahey]

I think that a very interesting question, which nobody answered so far (not even asked, I wonder why), is what *actual* voltage a floating cathode rises to.
Before talking about "300V, "600V" , "400V , "180V" and "550V" which were all mentioned in different posts, let's grab our trusty multimeter and post some real world values.
I *guess* we won't be too far from 10V or less. 20V or more would really surprise me.
Of course, I do not discard anything, I kneel before Queen Reality.
Any takers?

EDIT: voltage across 220K would also be interesting ... and I'm sure in that case it won't easily surpass 10 V.