Ultralinear, Va, Vg2 and tertiary winding: what's the limit?

[zintolo]

Hi,

with standard UL configuration, where screens are connected to the same voltage supply of the plates, the higher the signal at the input, the lower the plates go compared to the screens (more secondary emission?) but in any case screens will never go negative.

With tertiary windings, like for 6550 where to maximize power you need to connect plates to 600V and screens to 300V, or GU50 with plates at 900V and screens at 300V.

Now if you have classic 43% UL connection, and the plates swing down to 50V, the screens will swing:
- 6550: (600-50) x 0,43 = 236,5 Vp
- GU50: (900-50) x 0,43 = 365,5 Vp

So when the plates will be at 50V the screens will go down to:
- 6550: 300 - 236,5 = 63,5 V (still positive and close to Va)
- GU50: 300 - 365,5 = -65,6 V (negative voltage)

So the first case is ok, there will be almost no secondary emission I would say, and screens should have no issue, but what happens on GU50s with negative voltage on screens?

LTSpice simulations go into oscillation, but what will happen in reality?

Thanks

 

[JonSnell Electronic]

Negative voltage on G2 will stop the valve from emitting.
A lot of commercial amplifiers fit reverse biased diodes to stop the grids and anodes from going negative and absorb excess energy, that can cause arcing, in the transformer.

 

[zintolo]

Thanks JonSnell Electronic,

what can be considered as limit that is better not to exceed? +20V?

EG for the GU50 with Va = 900 V and Vg2 = 300 V, staying in AB1 (supposing Va = 50 V when Vg1 = 0 V) it is better not to exceed:

(300 - 20) / (900 - 50) = 280 / 850 = 33% UL

 

[zintolo]

Can the UL tap with tertiary winding be considered a way to reach very low Va (even in AB2) without having distortion introduced by secondary emission, by targeting Vg2 few volts less than Va when Va is at its minimum?

So for a generic tube with tertiary 43% UL:
Vg2 300V, Va to be calculated.
(300 - 20) = 0,43 ( Va - 50)
Va = (280 / 0,43) + 50 = 702 V

So let's say that 750 V would be better.

And what about secondary emission on UL when there's no tertiare winding?
When we are at full swing, and Vg2 is only 43% of that swing, so 57% higher in voltage?

Thanks again in advance

 

[OldHector]

The desire for UL is to have negative feedback via the screen, and get closer to triode like behaviour. Another mechanism is to use cathode feedback. Patrick Turner has this 6550 amplifier, and discusses his philosophy for using CFB, and regulated screens.
8585-amp-october-2006

The 12.5% CFB has a more dramatic effect on the pure beam tetrode mode harmonic spectra than
UL of any % value except perhaps 100% UL which is triode connection with screens to anode
with series 220r.