[HamWAN PSDR] Beacon Tower-Capital Park Backbone Link

Randy Neals randy at neals.ca
Fri Mar 30 01:43:18 PDT 2018

Not sure if attachments work via this list, but I'm attaching a PDF of the
3.4 GHz Bay Area Data Network I helped build in 2012.
This was all Rocket M3's. (I used to be KI6TWT then)

It was myself and a few friends from the Bay Area Microwave weak-signal
group that helped Bay Net put this together.
We found the Rocket M3's on eBay, and we bought new sector antennas, shield
kits and dishes as needed.

The Emcomm groups were all focussed on HSMM and building 2.4GHZ mesh. It
was early and they hadn't yet realized how poorly a mesh network performs.
They found our 3.4GHz project unappealing because they worried that our
central site with the sectors could fail.

Questions I saw earlier in the thread..

*GPS.* The versions of Rocket M3 we had were before GPS sync was available.
There are later versions with GPS. We were cheap and buying used/second
hand M3's that were intended for International use. I have no experience
with the GPS rocket M3.

*Standards/Standardization..*I applaud HamWan for it's standards, and the
last thing I would want to see is massively increased technical debt due to
one-off deployments.
There are UBNT pci radio cards for 3.4 GHz. Maybe we run those cards in
Mikrotik routerboards to keep the OS the same.

*Seattle ACS Sites/Seattle ACS Link?*In this case, the two sites in
question are Seattle ACS sites (Both are Seattle housing authority
buildings which Seattle ACS formally has an MOU for using the sites)
Perhaps we build and maintain the non-standard link between Capital Park
and Beacon as a Seattle ACS Project?? (Using the COMT people Doug
Then we provide HamWAN an Ethernet VLAN across the link which joins the
Capital Park and Beacon HamWAN routers?
>From HamWAN's perspective, it's a PtP Ethernet circuit that you run a /30
and OSPF over. You call Seattle ACS when it fails, We fix.

*Bart's 3GHz Link to Haystack..I'd be afraid that we wouldn't be able to
GPS sync the Rocket M3's with your planned equipment for Haystack.*Similarly,
if we deployed XP3 cards in Mikrotiks, also no sync.
Haystack is an important site, and 3GHz makes a lot of sense for that
distance up to Haystack - I don't want to impede that.

The Capital Park and Beacon sites could easily work on 10GHz.
I suspect we could make a path work on 24GHz, but I worry about rain fade
with 24GHz.
The Beacon to EOC like is about as far as I would want to trust 24 GHz


On Thu, Mar 29, 2018 at 7:32 PM, Bart Kus <me at bartk.us> wrote:

> As another data point, I do plan to install an AF-3X based uplink to
> Haystack this summer when the snow clears.  But those modems are like
> $600/ea.  :-(
> - Does the Rocket M3 support GPS-sync?
> - Does it require an M5 or some other modem on the LAN to sync with that
> GPS?  (I don't see it having a GPS port)
> - Are the Rocket M3 GPS-synced TX/RX windows in phase with the AF-3X
> GPS-synced transceivers?  (ie: can they coexist on a site)
> To clarify my earlier email, since I consider it implicit but perhaps it
> should be stated explicitly, I'm not opposed to using other bands.  It is
> healthy though to contribute points of balance to the conversation so that
> good decisions can be made by weighing all the pros / cons.
> Let me also address Bryan's IRC statement of "it's ham radio, just buy
> some stuff and play with it".  While that's totally valid for smaller
> networks, we're finding ourselves stretched very thin in the task of
> maintaining this network.  It's grown pretty large, and too few people know
> how to operate it.  We have to take a multi-faceted approach to keeping
> things operational:
> 1) Onboard more skilled folks who know what to do
> 2) Keep things as simple as possible to minimize management
> 3) Rely on automation to offload manual labor
> While some of us operate far larger networks professionally, those
> networks come with significant financial resources in staffing, vendor
> support contracts, and automation software development / tuning.  We don't
> have any of those benefits here, so we're hitting the wall far earlier.
> Finally, I saw a point come up in this thread from Kenny that said: "We
> have hit a situation that the old approaches are not working".  I haven't
> yet seen strong evidence for that being true.  Where "that" means "using
> the 5GHz band".  For the QueenAnne-CapitolPark example cited in this
> thread, I haven't seen anyone post results of spectral analyzer runs at
> both sites to find a common free channel.  I suspect if this is done, we'll
> be able to find a frequency that works A-OK for that link.  (Earlier point
> about us running out of manual labor capacity exemplified here.)  Given the
> excellent LoS between [Beacon, SeaEOC] and [Beacon, CapitolPark], I
> wouldn't be surprised if both CapitolPark and SeaEOC can link to Beacon via
> the same dish there.  SeaEOC can live on a sidelobe since it's extremely
> close.  This would cut out the requirement to install yet-another-dish @
> Beacon (a pro!), and save money (another pro!).  It may be slower than a
> dedicated [3,10,11,18,24,74,etc]GHz link (a con).
> --Bart
> On 3/29/2018 6:06 PM, Fredric Moses wrote:
> We are just using the RocketM3 $179 with the RD-3G26 Dish $229 and Radome
> $49.00   Now times that by 2 for each side of the link.. we have 19 to
> 26mile links going now.. with 6 more on the todo list here in Michigan
> ourselves..
> We use them as simple layer2 PTP bridges between our site routers...
> --
> Fredric Moses - W8FSM - WQOG498
> fred at moses.bz
> On Mar 29, 2018, at 21:02, Tom Hayward <tom at tomh.us> wrote:
> On Thu, Mar 29, 2018, 16:45 Bryan Fields <Bryan at bryanfields.net> wrote:
>> Why not use 3.4 GHz UBNT radios?  We have a link here in Tampa at 16.2
>> miles
>> across Tampa Bay running at 130 Mbit/s.
>> 3.37 to 3.5 GHz (the frequency range of the M3 radios) is totally unused
>> for
>> the most part.  A complete link is well under $1000 including antennas.
> Have you published a BoM for this? Can you?
> Are you still trying to unload a pile of modems?
> Tom
>> _______________________________________________
> PSDR mailing list
> PSDR at hamwan.org
> http://mail.hamwan.net/mailman/listinfo/psdr
> _______________________________________________
> PSDR mailing listPSDR at hamwan.orghttp://mail.hamwan.net/mailman/listinfo/psdr
> _______________________________________________
> PSDR mailing list
> PSDR at hamwan.org
> http://mail.hamwan.net/mailman/listinfo/psdr
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.hamwan.net/pipermail/psdr/attachments/20180330/54478668/attachment-0001.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: WW6Bay Microwave Data Network.pdf
Type: application/pdf
Size: 637619 bytes
Desc: not available
URL: <http://mail.hamwan.net/pipermail/psdr/attachments/20180330/54478668/attachment-0001.pdf>

More information about the PSDR mailing list