Xbee Ping Test behaving strangley: Jumper 1 on AXE210 "doesn't work"

lahaye

Member
Xbee Ping Test behaving strangely: Jumper 1 on AXE210 "doesn't work"

Hello,
I have been playing with the Xbees, only some basic range/reliability stuff but I am already stuck.
Please let me explain:
The basic problem that I have is that the Xbees echo back whatever they received regardless of the position of jumper 1 (unless removed completely). When in top position ping shouldn't work but in my case it does at least 90% of the cases (moving the jumper back and forth a bit seems to alleviate the problem slightly, but after a short while ping works again, although the jumper is fittet in the top position)
So, this time around, something that shouldn't work does.

Setup:
-Two XBee connect Boards, jumper 1 top-position on both boards
-one board (A) connected to PC (MAX3232)
-one board (B)with ONLY the XBee module (no MAX3232, no PICAXE)
- if jumper in top position (both boards), I get a successful ping (90% of the time) when I shouldn't.
- if jumper in bottom position same as above but this is as expected.


Some Tests:

Jumper1 is removed
Continuity:
-PIC
only between jumper1-bottom-pin and DIP-18 socket-pin 16 (in7)
-XBee
between Dout and jumper1-bottom-pin
between Din and jumper1-middle-pin
NO cont. between Dout and Din ' as observed => Ping fails when jumper is in this position (removed)

so there is always a connection between Dout and in7

Resistance:
Jumper1-top-pin (top is stereo-connector side)
to socket-pin 3 (serin) : 51K Ohm
to socket-pin 4 (reset): 51K Ohm
to socket-pin 5 (gnd) : 20K Ohm
to socket.pin 13 (Out7) : 10K Ohm (part of voltage divider)
to socket-pin 14 (V+) : 47K Ohm
omitted socket-pins: inf. Ohm

Between the jumper1-pins (any combination) there is no continuity and infinite resistance.

Jumper applied:
Jumper 1 top position (normal position)

Continuity/Resistance:
Dout - out7 : no / inf. Ohm
Dout - in7 : yes / 0 Ohm
Din - out7 : no / 10K Ohm
Din - in7 : no / inf. Ohm

Dout - Din : no / inf. Ohm ' so it SHOULD work correctly (as in no ping in this position, righ?)


Jumper 1 bottom position (ping position)

Continuity/Resistance:
Dout - out7 : no / inf. Ohm
Dout - in7 : yes / 0 Ohm
Din - out7 : no / inf. Ohm
Din - in7 : yes / 0 Ohm

Dout - Din : yes / 0 Ohm ' as expected Dout goes back into Din for ping

Dout/Din/in7 are cont.; out7 is not connected (as I would expect during echo/ping/loop-back).


I set up the Xbees as described in th Rev-Ed AXE210 tutorial; at the moment
they are set to unicast with 16 bit addresses (MY set to 1234 and 4321), but this also happens when set to broadcast.
I preformed a factory reset.
The modules are more than 1 meter apart .
All test where preformed on both connect-boards.

Anybody got an idea why I get the echo regardless of the position of jumper1?
What more test could I preform ?(I have the feeling not all of my test are
terribly useful)
Anything else catches your attention?
Have I overlooked something basic?
Has anybody had similar problems?

Any ideas are highly welcome

Florian

P.S.: I realize my description is not perfectly clear so if you want me to rephrase/clarify something please let me know. The fat-print is the important/paradox part.
 
Last edited:

lahaye

Member
Interesting Observation

Ok think I just made a "Discovery"
Happened when I scoped the echoing XBee. I'll keep it short this time.
Setup Hardware:
Jumper1 top position (connecting jumper1-top-pin to jumper1-middle-pin), jumper connected to probe A, jumper1-bottom-pin connected to probe B
Tests:
A) no probes connected
B) probes connected set to 1/10th (I don't know how to call it but it measures 0,3V in place of 3V)
C) probes connected set to 1/1 (measures 3V as 3V)

Results
A) Ping works 'which is a bad result because jumper 1 is in top position
B) Ping works BUT Rx is corrupted (see Quote)
C) Ping doesn't work 'which is good because it should NOT work because jumper 1 is in top position

TX: XBee Ping Test
RX: (doesn't show, looks like small r[mirrored] /90 degrees angle)

TX: XBee Ping Test
RX: ¡`

TX: XBee Ping Test
RX: $$ €

I'll try to do some more tests, but I wanted to post this quickly. So sorry if my spelling and language is even worse than above.

Florian
 
Last edited:

lahaye

Member
I made some more tests and...well here are the results, sort of:

The hole scope-probe 1/1, 1/10 issue got me thinking that introducing some capacitance into the system (specifically the jumper1-area) might help. As far as I know the benefits of this "voltage-dilation" is to reduce influence to the system to be measured (at the expense of resolution: Heisenberg ?), in this case by reducing probe capacitance. I wouldn't really know though.

So this is what I did:
-Connect a 470 nF ceramic cap between jumper1 (who was in top/no echo position) and ground.
-Then ping

Testing scenarios:
- nothing atached to jumper
- no cap, but crocodile clip (clip + wire has some C)
- cap not connected to ground
- cap between jumper and ground
- last two repeated with 100 pF


Results:

To keep it short:
-The 470 nF between ground and jumper seems to work reliable, this means:
a) Ping does not work (nothing is echoed)
b) RSSI LED of echoing XBee is lit, other stays off (The echoing XBee is hearing the message [correct behaviour], but doesn't send it back [correct behaviour]). In 149 out of 150 times this was the case. Once did the echoing board not get the message (RSSI LED remained off).

-everything else:
doesn't work reliable, BUT better than nothing at all.

Details:

nothing: ping works +95% of the time, if no ping then usually after a ping-pause (circa one minute), repetitive pinging yields near 100% ping success (which we do not want)

crocodile clip: some "beneficial"-effects, ping usually works but the data is corrupted (slightly to totally); but the wizard does says Ping success. No ping is sometimes due to no packet arriving at the echoing board (which is bad, we want data to get here, we just don't want anything send back again).
cap not connected to ground: see crocodile clip, slightly "better":less Ping success/more corruption.

cap between jumper and ground:
470 nF:
works reliable (see above)
100 pF:
not reliable: although ping usually doesn't work the RSSI LED of echoing board is not always lit which indicates a lost packet rather than a non-echoed one.

I hope the 470 nF cap proves to be reliable.


Florian
 
Top