David wrote in
:


David,

You could try to get your hands on a tuneable bandpass filter, eg a
coaxial filter can as used in repeater installations, and place it
inline. If it fixes the problem, it points to IMD, and the need to design
a sufficiently good filter, but not necessarily as good as the can.

Image response only doubles the ambient noise, unless you are actually
hearing image signals above ambient noise levels. The lack of image
rejection points to a lack of front end selectivity, but IMD noise is
another mechanism that creates in band noise (and image noise) from
undesired signals in the environment.

Did you do the test to see if a small amount of attenuation improves S/N,
it should degrade it, but attenuation reduces IM products by more than
the attenuation (two or more times, depending on the mixing process).

I suggest you need a bandpass filter, what makes you think that IM
products are only below the operating frequency?

Owen

I have done some more investigation and found that the input filtering
is very broad (more of a match and a filter).

I am getting quite large responses at the image frequency and a couple
of smaller but still significant responses around 145-147 MHz.

The operating frequency is 151.637 MHz but these other responses are
causing a combined rise in RSSI.

I want to be able to test the design as it is before committing to a
new PCB. Can I simply retrofit a filter in line with the antenna (The
unit transmits also but only at 20dBm).

If this is an option, do you have any suggestions for a basic circuit
that will give good attenuation below say 150 MHz.(Steep cutoff)

Thanks again

Regards

David


Owen Duffy wrote:
David wrote in
news:9SuEh.3761$8U4.3204@news- server.bigpond.net.au:

David,

There are a number of things that might make the amplifier perform
differently if the source impedance changes, and you seem to be
exploring those.

Another thought:

One of the things that a real antenna introduces to the mix (cf a
signal generator) is the real world RF environment with a host of
signals to mix with each other in your receiver.

A quick check for this is to see if the S/N ratio of a weak signal is
improved by inserting an attenuator after the antenna. Obviously,
inserting a large amount of attenuation will degrade S/N, but try 6,
10, 15dB.

You have described the selectivity of the front end as "and then a
tuned circuit". Front end selectivity is very important in preventing
loss of receiver sensitivity by noise from IMD in the rx front end.

Owen

I have just completed the design of a 151 MHz ASK transceiver.

On the bench the unit works fine with sensitivity to around -115 dBm
however, when I attach an antenna, the RSSI level increases
dramatically
and it appears something may have become unstable.

I have tried a commercial ground independent antenna and a 1/4 wave
whip
and the results are the same.

If I disconnect the antenna and leave the RF input floating, it
behaves properly. I can also attach a small wire (say 300mm)
directly to the RF input and it also works fine. It just plays up
when I have a proper antenna connected.

I am not sure if this may be some issue with the VSWR of the antenna
causing some instability (I tried a 3dB pad between RF input and
antenna
and this did not fix the problem), or whether the input is so
sensitive that it is picking up the LO from the radio and causing
regeneration ?

A Spectrum Analyser was used to determine if anything else is being
transmitted in that region and it appears to be clear to well
outside the bandwidth of my receiver.

Any ideas much appreciated for what might be happening or methods of
debugging my design.
(The circuit consists of an RF switch after the Antenna port
(selects TX/RX) then into a 50 Ohm RF Amp, followed by a tuned
circuit into an SA6505 RF IF. The LO is crystal based harmonic
Butler emitter follower.

PS. I also tried bypassing the RF Amp but still no improvement.

Thanks in advance

Regards

David