Hi everyone, I have an interesting scenario that I'm hoping an expert can weigh in on. I've looked through documentation and forums and can't quite find the answer I'm looking for.
My scenario is this:
Site A:
2 AP's, timing from CMM4 GPS sync
Site B:
1 SM connected to Site A
1 AP, timing from CMM4 GPS sync
The problem (I think) is that I don't have enough vertical seperation at site B, and the AP is transmitting while the SM is receiving, and AP is receiving while the SM is transmitting. My connections are very intermittent.
So my thought was... when an AP receives timing from an SM using the Aux port with RJ12:
- Does the AP sync its transmit with the SM transmit of the unit it is tied to via RJ12?
If so, using this principal, could I bring a sync cable down at Site B from the Aux port of the SM, and tie it into the CMM4 on the slave setting via the Aux port on the CMM4. This sync timing would then be used by the AP at site B, theoretically syncronizing the transmit/receive of the AP and SM at site B.
Thank you :)
A couple of questions may help shed some light on this:
1) What model radios are you using? Are they all the same?
2) What frequencies are you using? Site A( AP 1 and AP 2), Site B?
3) What is the application you have requiring Site B to have an SM and AP colocated? Are you doing a "micropop" kind of scenario feeding the AP with data from the SM?
What you suggest doing will not change the timing behavior. Using GPS timing, whether via the CMM4 or the AUX/timing port will still provide the 1 PPS tick, and, by design, all APs will transmit and receive in sync (with the SMs doing the opposite).
Without knowing more about your setup, I would suggest that increasing physical isolation (whether moving them apart or putting shielding in between them) will help some. Better would be to look at the frequencies that are being used by the neighboring sites and separating them as much as possible.
Matt
Thanks Matt.
1) Site A: PMP100 AP, and PMP450i AP. Frame caculator has been used to align timing. Both AP's are playing well together. Both AP's have SM's connected that are solid, except for Site B's SM connected to Site A.
2) I have tried multiple frequency ranges, right now I've settled on 906 for the PMP100 and 904.5Mhz for the PMP450i for Site A. The modulation is mostly 2x/2x except for the furthest client on the PMP100, and 8x/6x for both chains on the PMP450. Other frequencies have a very high noise floor and I've had trouble with trying them. Site B is using 914Mhz at 5Mhz.
3) This is a SCADA implementation, where sites were too far to directly connect to Site A, so Site B has both equipment connecting to Site A, and is acting as a micropop as you suggest, feeding the AP, through a switch.
Bascially I can have Sites A and B working flawlessly if it turn off the AP at Site B disconnecting further downstream SM's.. Or I can disable the Site B SM to Site A to have things working flawlessly.
I'm thinking the physicall isolation is the issue for sure.
How much isolation is recommended? I realize this is probably an open ended question... a calculator I used says at 900Mhz I would get 58dB isolation at 6 feet seperation, 46dB at 3 feet, etc. Right now they are too close I'm sure, at about 1-2 feet.
What would be a good shielding option? I've never actually seen a marketed shielding product (not to say that they aren't out there).
Thanks again!
Ah, this is 900 MHz... lots of constraint on frequency selection.
Is there any possibility to use a higher frequency (higher than 914) for the AP at Site B? More frequency separation should help as well. Alternatively, could another band be used somewhere (such as 2.4, 3 or 5 GHz)?
In terms of isolation, I don't know of any commercially sold products either. Usually customers come up with creative solutions using objects at the site. Sorry I can't be of more help here, but vertical separation should help.
Out of pure curiosity, could you please explain the details of how sync propogation works via the Aux port?
If I had the following scenario;
AP -> SM -> AP (timing from SM) -> SM -> AP (timing from SM)
And lets say each site is linked in a long chain, each being about 6.5 miles or 70,000 ns away from the next.
Would the sync timing eventually "slip" by the same amount as the air delay at each leg? So the first AP that receives timing from an SM has a PPS tick that is 0.07ms (70,000ns) later than the original AP? The next would have a 0.14 ms later timing than the orginal? And so on.
Or would the SM caculate a new "tick" by subtracting the air delay that it knows between it and it's AP, and then pass that on as the new tick?
Or something else entirely?