>>They are simply showing the degradation when the
>> tube is run at its TYPICAL rating of 2KV.
>
> and at much higher output power. The IMD goes bad because
> the tube
> is being pushed to deliver too much power, not because the
> HV is
> higher.
That is correct. People are comparing an entirely different
set of parameters, not just a voltage increase alone.
It is the output power increase that is largely responsible
for higher IM, although a loose screen supply could
contribute if the higher voltage starts to produce a more
dynamically varying load on the screen supply.
> STC data, 'FRX's and my measurements are all at 2kV or
> close. We
> all agree on the linearity at 250-300W pep output. Eimac
> quote
> figures at 470W pep - the IMD is going to be much worse at
> that power.
That is an important point.
> Why - I see nothing to suggest it's a typo. 470W pep (two
> equal
> tones) is 235W average if there's no IMD. At 470W pep from
> a
> single '250 there's going to be considerable flat-topping,
> as
> evidenced by the IMD levels, so the average output power
> will be
> higher, maybe 260-270W. That's entirely reasonable for
> 490W dc input.
It loooked reasonable to me also.
> Only if you try to push the output power too high. Under
> all
> normal circumstances higher voltage will give you higher
> output
> for given linearity or better linearity for a given
> output.
That's a good point and an important point.
> IMD figures need to be tied to power levels (and any other
> factors
> which have a significant effect).
...and just when I started to think only anode voltage
mattered! I was going to run all my tubes at 100 volts, so
they would be really linear. :-)
For a given power output more HV virtually always means
better linearity. However, we should be aware the screen
supplies were not always the best back when the 250 data was
taken. Those figures also were meant only to be a typical
operating point, not an optimal point for any particular
design concern like IM3.
73 Tom
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
|