[Elecraft] roofing filter questions

[Bill W4ZV]

Dick Green WC1M wrote:
> 
> 
> My friend maintains that the roofing filter would provide selectivity as
> well, but my sense is that unless it's narrower and/or has steeper skirts
> than the IF filters, it won't. I believe the Kenwood roofing filter mod is
> only on the order of 2K-3K wide, so the selectivity shouldn't improve. 
> 
> Even if the roofing filter was the same or narrower than his IF filters,
> would it improve selectivity? Is there an analogy with cascaded IF
> crystals?
> Typically, a filter in the final IF stage (e.g., in the 450KHz range)
> outperforms a filter in an earlier stage (e.g in the 8 MHz range.) Is this
> because it's easier to make better filters at lower frequencies, or
> because
> it's more effective to filter after any amplification by previous IF
> stages?
> 

>From Inrad regarding the 950 roofing filter:

http://www.inrad.net/product.php?productid=88

#########################################################
The TS-950 roofing filter mod consists of a 6-pole, 4 to 5 kHz wide filter
followed by a high dynamic range feedback amplifier. The amplifier provides
enough gain to override the filter insertion loss.

The IMD dynamic range will be improved 5 to 15 dB for some signal spacings.
Also, the blocking dynamic range will be improved up to 10 dB for close in
signals.

The roofing filter mod inserts a narrow band crystal filter between the RF
assembly and the IF board. Reducing the bandwidth at this point in the radio
helps to keep strong off-frequency signals out of the second mixer, where
they can cause intermodulation. Transmission is not changed, as it does not
pass through the roofing filter.
###############################################

If he's being bothered by signals "just outside the passband of his cascaded
IF crystal filters" (presumably 500 Hz each), a 4-5 kHz roofing filter won't
help at all since the roofing filter is ~10X the bandwidth of his IF filters
and will continue to pass the signals bothering him.  If he had a narrow
roofing filter like the K3 (200-500 Hz), that would help IMD (intermod) and
BDR (desense) issues but would *not* reduce transmitted issues from a dirty
interfering TX signal.  Jack K8ZOA is right that it depends on the exact
nature of the interference (i.e. is it transmitted [phase noise or key
clicks] or internal to the receiver [IMD, BDR or LO phase noise]?)

73,  Bill  W4ZV
-- 
View this message in context: http://www.nabble.com/roofing-filter-questions-tp17213673p17216992.html
Sent from the Elecraft mailing list archive at Nabble.com.