[HBR] Another Receiver Project -- HBR-4, Part 23

[waltah at earthlink.net]

'Learning ... something ... I think ...'

The high pass filter idea didn't work, but adjusting the premixer and 
mixer balance made all the various spurs tolerably small.   I still 
haven't gone to work on the crystal oscillator.

Time for serious testing of dynamic range.   I spent about a week 
thinking about theory and reading about issues.   Then I did the 
tests at various signal levels to get a handle on receiver issues vs. 
test setup issues.   I got good consistent results at about IFDR = 
78 db.   That's better than a whole lot of vacuum tube receivers, not 
quite as good as the R-390A, and certainly not what I wanted or 
even expected.

Back to head scratching.  

I have a hard time believing that the problem is in the mixer; the 
6JH8 is linear for deflection plate inputs to about 40 volts peak-to-
peak.  I don't think the 6ES8 RF stage should be that much worse; 
while it doesn't have so great a range as the mixer, it's way up 
there as RF stage tubes go, and the sections are in push pull, so 
second-order distortion should be mostly cancelled out.  Plus 
signals are smaller there.  

That doesn't leave much.  The most obvious candidate is those 
wonderful FT-101E tunable front end coils: cup-and-core 
construction, very high Q, but the cores are very small -- under 
6mm diameter; could saturation be the issue? The FT-101's IFDR 
was *40* db; coils that would have limited the range to ~80 db 
would have been adequate and then some.

With a good spectrum analyzer the answer would come pretty 
easily -- at least to the level of what stage the problem shows up 
in.  I don't have that, but another approach would be using a signal 
generator to measure the IFDR and third order intercept (sort of the 
maximum power level that can be handled) at each stage.   I may 
not have a powerful enough sig gen to do that either, but it's worth 
a try.   

Maybe at the mixer input.   Sensitivity there is probably around 10 
uV and the signal generators go over 100,000 uV -- there's 80 db, 
so it should be close at worst.

Yet another way -- and confirmation of the diagnosis if the signal 
generator test suggests the problem is between the RF and mixer 
stages -- is a temporary external tuned circuit with a more robust 
coil.   That one's messy, due to the push-pull output of the RF 
stage.   Too soon to think about the 'fix' if this really is the problem.

Walt
KJ4KV