[1000mp] MP vs MkV vs K2

[Tom Rauch]

> Where are the non-linear products in the MP being generated?  In
> particular, are they in the early stages, prior to the 8.2 and 455 IF
> filters? 

In virtually all systems IM products are generated in the early 
stages of the receivers, before reaching narrow filters. IM 
generation requires non-linear signal amplitude response, so filters 
are almost perfect up to large fractions of a volt or more. Almost 
exclusively, the active devices used in mixers and amplifiers are the 
fault.

In the FT1000MP and D series, the first mixer is a well-designed 
strong mixer. The RF amp is pretty good also. Yaesu unfortunately 
hangs a dual-gate FET right off the 70MHz IF just after the roofing 
filter, and it is by far the worse problem. The FET is used for the 
noise blanker, and it saturates and back-feeds IM products into the 
70 MHz IF even when the noise blanker is OFF. 

It's as if the designer's only goal was looking good in a 20kHz test, 
and as if they didn't care a bit about good operating performance.

After that the FET, the second mixer and RF amp are the next poorest 
stages for IM and blocking dynamic range in the 1000 D and MP. The 
synthesizer has a bit of noise that results in unwanted noise 
production, but that is minor compared to the other problems.

The small signal diodes used to switch circuits in the receiver have 
no effect at all on performance. Changing them is a waste of time and 
money, like taking a sugar pill for an infection.

The K2 actually is old simple technology, with no special effort to 
be good. The mostly just didn't do anything grossly wrong, rather 
than actually having an exceptional design. It would be quite easy to 
beat the K2, or improve it. 

What they did right is use a diode passive mixer, and a reasonable 
post-mixer amplifier after the mixer and before the filter. Just by 
changing the post-mixer amp, the numbers could increase 5dB. For 
about another twenty dollars or so in cost, you could push IM3 and 
blocking up another ten or fifteen dB.

Yeasu, like most manufacturers, screwed up after the roofing 
filter...acting like no one would ever operate within 10kHz of the 
user. The same "careful design thought" obviously went into the CW 
keying waveform, like it does on many other rigs. 

Generally harmonics and intermod products are generated in
> the latter stages, where the signal levels are the highest.  And
> woujld be after one or both IF filters.

That is not correct, unless you are talking about signals within the 
passband of the filters. Even at that, such IM and harmonic 
distortion is not deleterious to communications. The only exception 
is when broadband noise mixers with a noise-floor CW signal and 
muddies it up. That can make it difficult for people with "good CW 
ears" to dig out the signal.  

The real problem is in front of the first **narrow** filters.
 
> The point I'm heading for is:  What is the significance of  20 KHz.
> spaced signals, or even 5 KHz spaced signals, if the nonlinear
> products are generated after the IF filters, where you won't have both
> of those strong signals simultaneously present?

Because the products are NOT generated after the narrow filters, and 
NOT generated IN the narrow filters, and the because the 
receiver has a very wide roofing filter than can make a poor receiver 
look good at wide spacings, test-spacing is critical.

Manufacturers would love to have all receivers tested at 20kHz or 
wider, because it makes the system look better than it is when a 
roofing filter is employed. Users on the other hand would find 1kHz 
spacing more useful for CW, and 3kHz for SSB.

After all, stop and think about it. Who bothers you most often 
(keyclicks aside)? Is it a guy 3kHz up or down on SSB, or a guy 20kHz 
away?

The 20kHz test obviously thought up by someone who cared less what 
the end-application would actually be.   
 
> The assumption in the measurements must be that the nonlinear products
> are generated pre-IF filters.  Does anyone know this for sure?

Absolutely!!

Without any doubt at all the most destructive IM products and 
blocking problems are generated **ahead** of narrow filters. Not in 
them, not after them.73, Tom W8JI
W8JI@contesting.com