83 rectifier peak repetitive current?

[plovati]

I'm in trouble to exploit all the 83 mercury vapour full wave
rectifier capability. The limit on the input cap is due to the max repetitive peak current on the normal working condition and on the one cycle peak or surge peak current at startup (with preheated cathode). The post already searched in diyaudio are not resolutive.
I need to known the max repetitive peak current (the peak current allowed for each 100Hz charging cycle) and the max surge peak current allowed at very low duty cycle (example 2.7A for max 200ms for 5U4, typically at power-on)
I'm not referring to the max DC current that is 225mA.

I've collected the following datasheets around the world:

1) 83 (and 82) Sylvania, french manual, 1944
2) 83 (and 82) Tung Sol, 1940
3) 83 RCA, 1937
4) 83 RCA, 1963
5) 83 (and 83V) Brimar, 1958
6) 83 (and many others) Universal Vademecum, 1960

I was not able to find out a consistent value for the so called "steady state peak current" rating for this tube.

According to the above mentioned datasheets but the Tung Sol one, the "maximum peak plate current" is 1A (per plate).
Well, according to the venerable RC15, RC18 and RC20 tube manual, the "Maximum peak plate current is the highest instantaneous plate current that a tube can safely carry recurrently in the direction of normal current flow."

So, at first sight, it seems to me that, as for the word "recurrently", that "steady state peak plate current" should be 1A (design center maximum).

But on the 1940 Tung Sol's datasheet there's another, different, data:
the "Maximum steady state peak plate current", and it's rated at
675mA (per plate), and there's no other rating on this datasheet which has as value 1A.

What is the real values for the peak currents (repetitive, non recurrent) for 83 diode?

[Colt45]

just use choke input so you don't have to worry about it