Non-simultaneous Pentode anode/G2 turn-on

[Klimon]

Here's a difficult one (atleast to me, obviously):

I'm building a PP with separate HT supply for anode and screen grids; output tubes are el95 in pentode mode. Anode supply will have Gz34 or GZ 32 rectifier (Indirectly heated); G2 supply az1 (old tube with side-contact socket, directly heated, similar to 5y3). I just realised it will take some seconds for the anode supply to reach high voltage and catch up with the (direct heated rectifier) G2 supply. Will this harm / destroy the output tubes? -my uneducated guess is that max. dissipation of G2 will be exceeded.

As an alternative I'm thinking of using a direct heated octal instead of GZ32/4, still leaving the possibility that the AZ1 will turn-on a little faster, or maybe a separate switch to activate screen supply after anode supply has reached it's working voltage.

btw The output tubes are biased using an lm317 as ccs.

Muchos gracias for any suggestions!

Simon

[ak_47_boy]

That is an interesting question, i will be interested in the answer.

One possible solution is to delay the el95's filaments a few seconds.

[Tom Bavis]

Delaying the screen supply is probably a good idea - if you slave it to the main B+ it will be fail-safe in case of a rectifier failure, etc.

[pmillett]

Yup, I'm pretty sure you'll make the screen grid light up.

Unsophoisticated, but you could just put a silicon diode between the screen supply and plate supply... once the plate supply is above the screen supply the diode will be reverse biased...

Pete

[Klimon]

thanks for your suggestions.

As I understand it, when installing a diode the anodes will draw current from the screen supply untill it's own PS has reached working voltage. Sounds elegant but I'm afraid my screen supply trafo will be temporarily overloaded by >200% each time I power up, and it's an old chunk of iron Separate slave switch it'll be; expanded ignition ritual

Edit: Am I right to presume that the anodes will draw more current when the screens aren't at their desired voltage? (embarrassed)

Simon

[Fuling]

Am I right to presume that the anodes will draw more current when the screens aren't at their desired voltage?

More like the opposite, with 0V on the screens the plate current will be very low.

[Klimon]

thanks Mr. Fuling; that means I can resume building and should get back to tube basics

Simon

[ray_moth]

As others have said, you must not allow the screen voltage to be applied before the plate voltage, or you will destroy the tube through excessiver screen current.

You may like to regulate the screen supply, which is a good idea to maximize linearity. You can build a simple regulator from a single MOSFET, a few zeners, a couple of resistors and a couple of capacitors. It's easy to arrange a delay on the screen supply, to ensure that the plate will reach full voltage before the screen does.