[ARC5] IMD performance

[Geoff]

> The topic of hollow-state receiver performance is a very
> interesting one. Unfortunately most of the modern approach
> to IMD testing, etc, postdates the times when tube receiver
> design was at its peak.
>
> One small exception appears to be the EF183/EF184 aka
> 6EJ7/6EH7, designed for low IMD use in color TV IF chains.
> Their data sheet has a small section showing IMD performance.
> Also, if I recall correctly, the RF input impedance versus frequency
> is discussed to some extent.
>
> Another contender for very good intermod performance would
> appear to be the beam deflection tubes. To me, differential plate
> current versus beam deflection voltage appears to be very linear,
> but eyeballing these things is not very scientific.
>
> There must be other published information and/or measurements
> on this area of tube performance, surely? Anyone seen something
> like this?
>
> Chris Trask made comparisons between tubes of different
> linearity but I don't know whether he had a way of doing this
> quantitatively from data sheets, or whether it was subjective.
>
> My latest tube receiver effort has a 6EJ7 RF amplifier mostly
> to act as a buffer between two stages of preselection. This
> at least gets rid of lots of interfering junk that is some way out
> from the desired signal while giving reasonable IMD performance.
>
> One of these years I guess I should read the literature on how to
> make IMD/IP3 measurements and see how well these glass front
> ends hold up against published figures for solid state gear.
>
> My basic thought behind all this is that a figure of merit for tubes
> may not be worse than that for transistors: solid state devices
> have much higher gain-BW products but much more nonlinear
> characteristics in general.
>
> 73, ian K3IMW
>

SS as used in the CATV line amps were extremely linear in the presence of 
500MHz of TV channels as far back as the mid 70's. The 2N5109 is a case in 
point. Ive used it as a post mixer stage in several projects to bring the NF 
way down and add no additional IMD. The FET and its various iterations can 
give small signal IMD3 down to -50dB or so; the U-310 is still a very 
useable device and the more common J-310 is not too bad. Ive often wondered 
how well a Class A driver stage TO-220 RF bipolar and FET would do as an HF 
preamp?

I can state positively that the 7360's I used in both 75A4 mixers in 1965 
turned an overload dog into a contest battle cruiser. I still have it. The 
best I could do for the RF amp was a 6GM6 since I didnt want to hack a 9 pin 
socket into it.

My next major mod will be a 6ES8 Pullen mixer in a 75A3 as many reports put 
it up in the realm of the 100 mpg carburetor....but no one is trying to sell 
you anything.
Since I have enough NIB 7360's the 75A4 will be left alone...unless the 
Pullen really is more than just a lil bit better.

Carl