October 2015 - PSDR - mailman.hamwan.org

Does anybody have a digitizing microscope?
by Bart Kus 29 Oct '15

29 Oct '15

Hello, I wanna take some pix of the 9HPn PCB layout. The components and markings are extremely small, and my camera can't resolve them. Does anyone have a microscope with picture capability to do a mosaic? Thanks, --Bart

4 7

Perhaps 900MHz isn't a complete write-off
by Bart Kus 29 Oct '15

29 Oct '15

Some of you may know I did a survey of 900MHz recently, along with the real-world modem tests. The spectrum analyzer inside the modem showed a very dire picture of the band, with the noise floor averaging about -60dBm. Today, in preparation for 1.2GHz antenna installs, I ran some coax to the roof and was able to see the 900MHz vertical there directly on a proper spectrum analyzer. The spectrum story is quite different! Here's a view of a typical spectrum from 902-928MHz: floor The reference line at the top is -20dBm and each vertical graticule is 10dB. The noise floor looks to be about 7 graticules below -20dBm, or -90dBm. There is about 2dB loss in the 45ft feedline, so let's call the floor -88dBm. There are however some sparse transmissions in this band. I wish I could find a way to measure their duty cycle or plot their average power. Sadly the best I can do is to record their maximum power using the peak-hold mode of the spectrum analyzer. I captured about a 10 minute sample of air time, and here are the results: peaks The bright dot is the peak marker, and it reads -42dBm, which is really -40dBm at the antenna after you account for the coax loss. The peaks are all suspiciously uniform in power, so this may be a single transmitter just jumping around the spectrum with each burst. I can't explain why the Mikrotik 9HPn reports such a vastly different spectrum. It may be mixing the 860-900MHz cell band into the 900MHz band? I need to perform some further tests to figure this out. There are definitely strong signals in the 860-900MHz range (not shown in these photos). --Bart

3 3

Re: [HamWAN PSDR] Any tower climbers available for this Saturday @ Gold Mtn?
by KL7WM＠aol.com 16 Oct '15

16 Oct '15

I'm teaching I can be available on Sunday Daniel Stevens KL7WM In a message dated 10/15/2015 10:59:32 P.M. Pacific Daylight Time, me(a)bartk.us writes: It's by Bremerton. Please let me know if you can climb and install stuff. We have a full cell site to deploy. Thanks, --Bart _______________________________________________ PSDR mailing list PSDR(a)hamwan.org http://mail.hamwan.net/mailman/listinfo/psdr

4 3

SnoDEM (Paine) is well-connected again
by Bart Kus 16 Oct '15

16 Oct '15

Hi All, Nigel and I did a trip today to SnoDEM to re-align a couple of the dishes. We've got nice strong signals now from East Tiger and Haystack going to SnoDEM. After Saturday's Gold Mtn deploy we'll need to take 1 more trip to align a SnoDEM dish to aim back @ Gold Mtn. This should make SnoDEM triple redundant, which is great for an EOC. Things got super sketchy there for a while, as multiple events coincided to wreak havoc on SnoDEM's connectivity: 1) Nigel moved and his uplink was disconnected as a result. 2) The CapitolPark-SnoDEM link faded into the useless signal range. We disabled it since it could hardly pass any packets. 3) A modem died @ Haystack which killed the SnoDEM feed from there. 4) My uplink to SnoDEM got tree-faded into uselessness and was disabled. And that's how a quad-redundant site suddenly became a disconnected site! --Bart

2 3

Any tower climbers available for this Saturday @ Gold Mtn?
by Bart Kus 15 Oct '15

15 Oct '15

It's by Bremerton. Please let me know if you can climb and install stuff. We have a full cell site to deploy. Thanks, --Bart

1 0

Council House signal strength tests
by Bart Kus 13 Oct '15

13 Oct '15

Council House is at (47.614479°, -122.310476°). Council-Gold: -58dBm @ 20MHz dish Council-SnoDEM: -62dBm @ 5MHz dish Council-Baldi: -68dBm @ 5MHz sector Coundil-Baldi: -56dBm @ 20MHz dish Council-Haystack: -65dBm @ 5MHz sector Tests were conducted with a 30dBi Poynting dish and 5SHPn modem. This will make a fine hub site for tying the network together and giving rise to some ring structure (graph cycles <https://en.wikipedia.org/wiki/Cycle_%28graph_theory%29>). --Bart

3 4

The Westin->Gold antenna has been tweaked
by Bart Kus 10 Oct '15

10 Oct '15

Can you tell when? :) --Bart

1 0

VOTE: Gold Mtn equipment purchase
by Bart Kus 08 Oct '15

08 Oct '15

Hello, We've installed a successful link from the supremely well connected Westin building to Gold Mtn <http://hamwan.org/t/Gold?structure=HamWAN>. The transfer speeds haven't been optimized yet, but initial testing shows about 40Mbit Westin->Gold and about 15Mbit Gold->Westin. There is a visit scheduled to the Westin for this Friday to tweak the alignment of that antenna and optimize the link. The existing speeds are good enough though to consider building the site out as a full cell, with PtP redistribution. To that end, here is the list of hardware we're going to need: *Description** * *Price** * *Qty** * *Subtotal** * APC AP7900 PDU @ eBay $129.99 1 $129.99 eBay shipping $0.00 1 $0.00 ARC-DA5830SD1 2ft MIMO Dish @ Microcom $99.99 2 $199.98 Microcom shipping $60.44 1 $60.44 UDR235 RF Armor dish shielding @ RF Armor $129.95 2 $259.90 RF Armor shipping $35.97 1 $35.97 RB921UAGS-5SHPacD-NM 2W MIMO modem @ Streakwave $131.00 2 $262.00 Ubiquiti TC-Pro 1000ft CAT5 cable @ Streakwave $127.65 1 $127.65 Ubiquiti AM-5G19-120 Sector @ Streakwave $117.00 2 $234.00 RB/GPOE Mirkotik power injector @ Streakwave $7.00 1 $7.00 Streakwave shipping $73.10 1 $73.10 MW-RPSMA-NM-1M RP-SMA Patch Cable @ Baltic $9.95 4 $39.80 RB912UAG-5HPnD modem @ Baltic $69.00 3 $207.00 Baltic shipping $13.53 1 $13.53 WaveGuard AM5G19 RF shield @ ISP Supplies $44.99 2 $89.98 RF Elements SBX-S-CC-2SMA shielding outdoor radio enclosure @ ISP Supplies $35.90 3 $107.70 ISP Supplies shipping $20.71 1 $20.71 Ubiquiti AM-5G19-120 Sector @ Bart Kus $110.00 1 $110.00 WaveGuard AM5G19 RF shield @ Bart Kus $40.00 1 $40.00 Bart Kus shipping $0.00 1 $0.00 Misc mounting hardware (optional spend) $100.00 1 $100.00 GRAND TOTAL $2118.75 The core site router will be the RB2011 from Baldi that failed. We've determined it works when powered from the PoE port, so we'll just feed it that way. There are RJ45 ends left over from the previous order, so no need for those. I'm deferring server purchase for this site to a later date. I've got 1 new-in-box sector and 1 new-in-box shield, so I'm selling those to HamWAN in this order at a discounted price. The total HamWAN savings between the unit cost and shipping decrease is $27.79, and I get to have less clutter in my lab. The work party is scheduled for Oct 17th, and ground shipping will take 5 days, not counting any handling delays, so let's decide on this quickly. I should have put all this together a couple days back. --Bart

4 7

Haystack-SnoDEM (Paine) is working again
by Bart Kus 07 Oct '15

07 Oct '15

Hello, I went up to Haystack yesterday and replaced the failed modem there. The link between Haystack and SnoDEM (Snohomish County Department of Emergency Management / "Paine") is back up at a decent signal level. We have a trip scheduled to SnoDEM for Oct 15th @ 5:00PM to better align two dishes there, including this SnoDEM-Haystack dish, which was originally aligned with Monroe. Now despite this good news, the link between Westin and Queen Anne (which ultimately feeds Haystack) is suffering due to Amazon Fade, and it's only gonna get worse as the building is completed further and occupied. We may need to re-aim the main Haystack uplink dish at a site different than Queen Anne. No firm plan on this yet, but East Tiger might be an option. Another option might be a new site on Capitol Hill that I'll be doing a signal test at on Saturday. Lots of activity lately. --Bart

2 4

Amazon Fade
by Bart Kus 01 Oct '15

01 Oct '15

Hello, We've discovered a new mode of attenuation we're calling Amazon Fade. It's clearly visible in this signal strength graph: Amazon Fade On Sept 28th, Amazon finished constructing a piece of steel member that is right in the middle of our signal between the Westin building and the Queen Anne site. The chunk of steel is towards the right of the building being constructed here, on the far side. Our Queen Anne site is directly behind it on the horizon. This webcam was setup specifically to monitor for the impact of this event, and I'm pleased to see the data come in. One other interesting this from this experiment is that the signal didn't fade at all as the floors were being built. Our signal shot clear in between them. It was only when that vertical riser was put in place that we measured an impact. I'll keep everyone updated on any future signal changes in our continuing study of Amazon Fade. :) --Bart

1 0