Maker Pro
Maker Pro

Poor performance with extended aerial

I have an 868MHz design incorporating a reference design for the
transceiver. The aerial is connected to a PCB mounted SMA connector.
The board sits in a metal box and the aerial passes partly through the
box to attach to the board.

Range is acceptable if not great. The aerial manufacturer retuned the
aerial for our box and said that unfortunately the most sensitive part
of the aerial was inside the box thus we were losing range.

We tried removing the sma connector from the board and attaching a
short 3-4 inches of coax with a bulk head mounting SMA connector to
original mounting holes of the pcb mount connector. range was reduce
to about 4 feet. We tried in and out of the box with no success. On
looking at the IQ signals it appears that the Q signal is largely
unaffected by the change, may be slighly higher amplitude but the I
signal goes from good to collapsing over the 4 feet distance. The
cable with connector is off the shelf and we hae tried several cables
and several boards. Any clues what may be happening by adding the
additional cable.

Thanks for any help

Paul
 
G

Genome

Jan 1, 1970
0
[snip blah]
Thanks for any help

Paul

In simple terms, having extended your aerial you have to play with the
tuning dial(s) and other parts to optimise response.

Trust me, I'm not an expert.

DNA
 
F

Fred Bloggs

Jan 1, 1970
0
I have an 868MHz design incorporating a reference design for the
transceiver. The aerial is connected to a PCB mounted SMA connector.
The board sits in a metal box and the aerial passes partly through the
box to attach to the board.

Range is acceptable if not great. The aerial manufacturer retuned the
aerial for our box and said that unfortunately the most sensitive part
of the aerial was inside the box thus we were losing range.

We tried removing the sma connector from the board and attaching a
short 3-4 inches of coax with a bulk head mounting SMA connector to
original mounting holes of the pcb mount connector. range was reduce
to about 4 feet. We tried in and out of the box with no success. On
looking at the IQ signals it appears that the Q signal is largely
unaffected by the change, may be slighly higher amplitude but the I
signal goes from good to collapsing over the 4 feet distance. The
cable with connector is off the shelf and we hae tried several cables
and several boards. Any clues what may be happening by adding the
additional cable.

The reference design most likely is tuning out the "aerial" reactance
over a narrow frequency band. You have perturbed this reactance with the
impedance transformation resulting from the section of coax extension
line. Retune the xmit reactance element, if at all possible, for best
results.
 
M

mpm

Jan 1, 1970
0
The aerial manufacturer retuned the
aerial for our box and said that unfortunately the most sensitive part
of the aerial was inside the box thus we were losing range.
 Any clues what may be happening by adding the
additional cable.

If you want to do this the right way:

Rent (or find someone with) an HP8753C Vector Network Analyzer (or
equiv).
If you are unfamiliar with how to set one up for the measurement,
you'll need a qualified operator.
Perform an S11 parameter test and locate (tune, or adjust) the antenna/
feedline for best Return Loss. That will give you your best
throughput.

Next, you can deal with directional pattern problems that may be
introduced by your enclosure or near-field antenna obstructions or
other parasitic elements, if any.

-mpm
 
If you want to do this the right way:

Rent (or find someone with) an HP8753C Vector Network Analyzer (or
equiv).
If you are unfamiliar with how to set one up for the measurement,
you'll need a qualified operator.
Perform an S11 parameter test and locate (tune, or adjust) the antenna/
feedline for best Return Loss. That will give you your best
throughput.

Next, you can deal with directional pattern problems that may be
introduced by your enclosure or near-field antenna obstructions or
other parasitic elements, if any.

-mpm

The transceiver has an LC network to splitting the incomming signal
into a refA refB signal are these the components I need to alter or is
it best done at the aerial.

Also If I tune everything up, If I then want to enable the use of an
extension aerial are we saying it will be no good just having a length
of cable plus say an 868 Yagi aerial as I would probably get more
loss than just that of the cable and connectors, but that I would also
have to either tune the aerial for the length of cable fitted or
provide a tuning components selectable on what aerial/cable is
fitted. We didn't seem to need to do this on the old 173MHz system.

Paul
 
I have an 868MHz design incorporating a reference design for the
transceiver. The aerial is connected to a PCB mounted SMA connector.
<<SNIP>>

Another thing, I'm not just seeing poor performance -6db -12db
reduction I'm seeing terrible performance -30db.

Why does it only affect Iamp shouldn't it affect both as the same
signal entering the tranceiver is mixed with the in and quad phase
local oscillators.

Another thing actually I tried using a SMA to SMA to original aerial
no coax involved and the same result. Now as I removed each section,
first the aerial and then the SMA-SMA connector the signal first moved
from being only on the Qamp to only being on the Iamp to only being on
the Qamp again.

I don't understand.

Paul
 
Top