Sony SRF M37V image problem solved - Update


Hi Stephan:

Gre aus Amerika! Ich habe Ihre Artikel ber die 7600GR geniessen.

False alarm! I forgot that the M37V uses a 450 khz filter, and so the
images are only 900 khz away. Sure enough, this morning my neighborhood
stations at 1590 and 1680 are solidly represented on 690 and 780, and so
the image problem is still there. I too was wondering how IF filtering
could solve an RF stage problem...

Sorry for the confusion!

73 - Kevin S
Bainbridge Island, WA

--- In ultralightdx@..., "dhsatyadhana" <satya@...> wrote:
Hi all:
Jim Kenny and I were talking on the phone and each pulled out our M37V
with a 4.5 khz retrofitted filter installed. Both of us realized that,
unlike the stock M37V, there were no images from strong locals (i.e.,
phantom 680 signal from local 1590, etc.). I then turned on my Sony
7600GR, and the images from my strong locals were blanketing the dial,
but nothing on the M37V!

Do keep in mind that the '7600GR with its wideband MW/LW
frontend will be more easily overloaded by strong AM
locals than a set with frontend tuning. (Its image
rejection is no greater than on shortwave, i.e. an
estimated 30..40 dB, for the same reason.)

But in any case, improved image rejection with a better
IF filter is a puzzler. Image rejection should be a
function solely of the frontend, i.e. RF tuning and mixer
(or, for most sets, RF tuning only, as regular mixers
have no inherent image rejection).

(A 680 kHz response from 1590 definitely is an image
if we're looking at a single conversion set with the LO
running above the tuned frequency - as it's normally the
case - and an IF of 455 kHz.)

Depending on achieved Q (and thus ultimately parts
quality), image rejection near the bottom end of MW for
sets with a single tuned IF circuit is about 30 to 50 dB.
It drops quite a bit towards the high end, but as there
are no strong signals directly above the band, this is
not an issue.

Anyway, I fail to see how the IF filter would play in
here. It can't really be spurious responses either.

Hmm. *scratches head*


Stephan Grossklass

--- In ultralightdx@..., satya@... wrote:

False alarm! I forgot that the M37V uses a 450 khz filter, and so
the images are only 900 khz away.
BTW, you know what the reason for 450 kHz is? With a 3 or 9 kHz
PLL reference frequency (and thus granularity), you'd never hit
the 9 kHz frequency grid precisely when using 455 kHz. (Same
goes for a 10 kHz reference in 10 kHz territory.)