At 07:33 PM 9/02/2003 GMT, you wrote:
G'day all
A huge thanks to SP5ENA, K8GG, K8ND, W8CAR, G4VGO, AA1K, K6SE, N7RT N1EU,
DF2PY and ZL3IX for their excellent contributions to my query about what
kind of filters to buy to improve the signal/noise performance of my Yaesu
FT1000MP, currently fitted with a 500Hz stock filter in the 8MHz IF and a
Collins 500Hz in the 455KHz IF.
First, a few general comments.
N1EU and ZL3IX stressed the point that the best filter we have is the one
between the ears - and judging by the responses, everyone's brain/ears seem
to work slightly differently/have been trained differently, which affects
the kind of narrow filters we prefer. ZL3IX noted: "In my opinion, when
the background is just white noise, the human ear does far better than any
filter with a bandwidth less than about 500 Hz. A senior colleague of mine
did some tests about 25 years ago, in which he demonstrated that a well
trained human operator can copy CW at signal to noise ratios down to around
minus 15 dB in a 2.5 kHz SSB bandwidth. There was little or no improvement
with a narrow filter. The ear/brain combination is a wonderful thing!"
SP5ENA and others remarked really narrow filters can make pulling signals
out of QRN worse, owing to ringing. AA1K notes his pair of 250Hz filters
are best for weak signals, but under heavy QRN, his wider pairs of filters
are better.
SP5ENA feels 250Hz is the optimum bandwidth, but this situation depends
upon the quality of the filters (i.e. pulse response, ripple in the
bandpass and slope). W8CAR notes there are times when his pair of 250Hz
filters put out signals that his 500Hz filter pair doesn't.
Some like K6SE, K8ND and DF2PY find the 125Hz IF filters a great help with
signal/noise ratio on weak signals, but others like G4VGO and K1EU prefer
wider IF bandwidths (in G4VGO's case, backed up with a 125Hz DSP audio
filter).
On the Inrad 125Hz filter, K8ND notes : "ethereal signals become copiable".
Yuri K3BU makes the observation that 125Hz bandwidth IF filters can improve
signal/noise ratio, but he also has experienced occasions when 250 and
500Hz filters provide better readibility. "When signals get a bit mushy
(multi-path?), it helps to have a wider bandwidth for discrimination - with
my ears/brain at least".
Now, some practicalities. To improve on the basic 500/500Hz combination in
the FT1000MP, the cheapest/simplest most-bang-for-bucks solution seems to
be to add a 250Hz filter in either the 8MHz or 455KHz IFs. K8GG suggests
the 8MHz IF - the earliest IF it is possible to add more filtering to - as
the one to go for.
The ultimate solution is the one proposed by K6SE - 500Hz and 250Hz filters
in both IFs, with an Inrad 125Hz filter at 455KHz, using the 2.0KHz filter
slot. Earl really likes the filter panel on the FT1000MP, making it easy
to select any combination of filters. He adds that he finds the 125Hz
filter offers big s/n improvements compared to using the pair of 250Hz
filters - and also uses the 60Hz dsp filter bandwidth that is selectable in
the MP.
DF2PY has substituted a 250 Hz filter in the 455KHz IF with a 125 HZ
filter. Wolf points out that this has about
2 db more insertion loss than the 250Hz filter and has compensated for this
by putting in a class A amp with a bipolar transistor ahead of the filter.
He also notes the extra narrow filter is worth it.
After weighting up all of the above, the VK6VZ decision is to purse a
gradual upgrade path and to first buy an Inrad 250Hz filter for the 8MHz IF.
Then, if I get some money back from the taxman, next year, I'll add a 250Hz
filter for the 455KHz IF. The following one, if the same thing happens,
I'll put in an Inrad 125Hz filter in the 455KHz IF.
Hopefully, by 2005, near the bottom of the solar cycle, I'll have achieved
the K6SE ultimate solution - for the ultimate 160m conditions!
Vy 73
Steve, VK6VZ
|