Black Rock City Wifi Summit 2015

The Black Rock City Wifi Summit 2015 took place yesterday. It was an interesting discussion with a collection of people from inside the Burning Man organization and artists / community who want to provide wifi to their art installation and camps. It was held at the Internet Archive which is a very interesting idea / place.

Tidbits

There are some bits of information that I learned or re-learned while at the summit.

BRC internet basically uses all the available bandwidth coming into Gerlach, around 40Mb/s, to support all the services for a temporary city of 70,000 people.

In 2014, the bandwidth was maxed out on Thursday before the event started.

Internet access is becoming / has become critical for the support of Burning Man. Without it, the event might not be able to take place.

Ubiquiti AirMax protocol deals well with the hidden node problem.

It’s important that the uplink antenna be stable and pointed at the center camp. Movement or misalignment of the antenna causes problems for everyone in that sector by reducing the available bandwidth.

The uplink antenna needs line of sight to the center camp tower.

If you need to use tape to seal anything, use aluminized hvac tape instead of duct tape. The glue on duct tape deteriorates rapidly in the heat and dust of the playa. The aluminized tape will last much longer.

During the weeks before and the event, download is maxed out. The week after, it’s the upload bandwidth that’s maxed out.

At 4:00 and A there’s usually a phone booth that uses VOIP to allow calls to the outside world. Calls are limited to 3 minutes. Because of bandwidth usage, the other side will be able to hear you while you might not be able to hear them. That makes for interesting conversations when you’re stoned.

Don’t use 5Ghz band for access points or other projects. It’s used for the uplinks and putting up local access points (or anything else) in the 5Ghz band will just interfere with the uplinks.

There’s lots of local, on playa bandwidth available. If there’s lots of data that you

Mount the local access points 6 or 7 feet in the air, no higher. This will help reduce interference and the hidden node problem on the local wifi connections.

If there’s any data you want to make available, consider setting up a server on the playa. If you do, let the folks on the BRC wifi mailing lists know about it. They may be able to help let people know about the server by including it on the captive portal page, setting up a dedicated IP address and / or a DNS entry.

Protect the access point and uplink power sources from water and dust. A simple way to do this is to put them in a tupperware type container. Cut a notch near the edge to allow any wires into the container and seal it closed.

There are no guarantees. The bandwidth available is completely subjective to the current needs and wifi in camps and art installations is at the bottom of the priority list.

Lastly and perhaps most importantly, the folks running the BRC network are at Burning Man to enjoy themselves. Leave them alone and talk with other camps and installations about any issues that you may be experiencing.

Setup On The Playa

There are two main components needed to setup wifi on the playa – an uplink radio and an access point. The uplink radio runs in the 5Ghz band and the access point in the 2.4Ghz band. The recommended uplink radio is the NanoBeam from Ubiquiti. A NanoBridge will work if you have a spare one laying about. The recommended access point is a Ubiquiti Unifi AP.

To start, reset the NanoBeam to factory settings before bringing to the playa can help you save some hassles and wasted time on playa. With it reset to factory defaults, mount it with a clear line of sight to the center camp tower and power it up. It should connect and start provisioning. This can take a while, as in hours depending on what else is going on. Use a laptop to monitor the setup and configuration.

After the NanoBeam has connected, been provisioned and received it’s IP address, connect and power up the local access point.

Consider a Unifi AP Outdoor. The rain last year helped several people discover that the Unifi AP’s aren’t waterproof. Water and playa dust will do corrosive wonders to exposed connections.

Note that the UniFi AP Outdoor 5G runs in the 5Ghz band, don’t use it.. Why? Because 5Ghz is used for the uplinks.

The configuration of the Unifi AP’s that’s downloaded from the playa NOC will most likely include a captive portal that times out after some period. This will help manage the bandwidth needed since you’ll have to re-accept the terms and conditions page to connect back to the Internet.

If you don’t use a Unifi AP, set your access point SSID to the camp name / location and make it discoverable. Set up a captive portal for your users. It’ll help manage available bandwidth and provide service to the community:

Projects

I’ve been toying with different ideas for project on the playa of which two seem to be most likely to happen this year.

Wifi Kiosks

The first is a local wifi spot and map server but since there will be playamap onsite and there’s lots of local bandwidth available, I think the mapping may get put on the back burner for this year.

It turns out that there were a few like minded people at the summit talking about projects to bring wifi to the masses. I’m planning to work with the folks at http://ki7wv.net/ to create some solar powered wifi kiosks and distribute them on the playa.

Trasher

Set up a Raspberry Pi image that pulls SNMP status from the Ubiquity NanoBeam uplinks. This will allow an uplink operator to get an idea on how well their connections works over time.

I’m currently thinking this can be done with Cacti or Nagios or perhaps just a simple Graphite setup. I think Nagios is way overkill for this but we’ll see. Some experimentation needs to be done…

Tags : , ,

Battery Life, Part II

I previously posted about monitoring how long my vx8-dr would last on different types of batteries. One of the things that I noticed was that there is a relationship between battery voltage level and the output power but I had no way of watching what was happening. So I invested in a Diamond SX-600 SWR & power meter along with an MFJ-261 dummy load and re-ran parts of the battery test.

2,000 mAh lithium ion w/ power

vx8-dr 2000mAh power

relative time voltage wattage
0:00:00 8.2 4.2
1:35:31 8.0 4.2
3:59:07 7.7 4.1
5:50:41 7.6 4.1
7:16:26 7.5 4.1
8:21:35 7.5 4.1
9:19:13 7.4 4.0
11:04:18 7.4 4.1
12:33:02 7.3 4.0
13:56:41 7.2 3.9
14:37:08 7.0 3.8
15:10:24 6.9 3.0

This is pretty much what I would expect. As the power available from the battery goes down, the output power on the radio goes down.

AA alkaline w/ power

vx8-dr aa alkaline power

relative time voltage wattage
0:00:00 4.2 0.90
1:48:19 3.7 0.15
2:40:05 3.7 0.10

This, ladies and gentlemen, is utterly ridiculous! Starting with fresh AA alkaline batteries, the output is less than 1 watt and it quickly deteriorates from there.

I swapped out the AA alkaline for AA lithium ion and found a similar output power level. Since it started with less than 1 watt output power, I didn’t bother running through the whole test. This was not at all what I expected, The input power voltage level is critical for the vx8-dr to perform well.

My original intention was to use AA lithium ion and / or AA alkaline batteries as a backup power source for the vx8-dr in the event of an emergency. This now looks like a bad plan. Running on anything less than the lithium ion battery packs has a detrimental effect on output power. It might do as a backup to the backup power but I don’t think it’s a good idea to count on it.

Right now, I’m thinking multiple 2,000 mAh lithium ion battery packs, a charging dock and perhaps a solar panel to charge it all. Plus an automotive plug adapter. And a plug w/ anderson powerpole connectors.

This was an interesting experiment!


 

Update:

I was asked about the discrepancy between the maximum measured power out of 4.2 watts with the battery voltage of 8.2 and the published maximum of 5.0 watts. I have a couple thoughts on that though they’re not based in much more that aether.

  1. The published specs may be coming from theoretical maximums.
  2. The published specs may be coming from values measured in a lab under ideal conditions.
  3. Maybe the battery’s just not quite fully charged.
  4. The SX-600 power meter may be off calibration.

I did plug the radio into the wall wart that comes w/ the charger. The voltage showed at 10.8v and the output power at 4.3 watts. Using the automotive cigarette plug, the voltage shows at 12.8v and the output power was measured at 4.5 watts.

BTW, the manual for the vx8-dr has specs for the supply voltage maxing out at 12v.

From all this, I take the maximum output wattage as being 4.5 with a more typical of 4.0 to 4.2.

Tags :

Battery Life

I had an experience a few weeks ago whereby I couldn’t join the ORCA net on my vx8-dr because of a low battery. From that, I thought I’d run a test to look at how long different types of batteries run in the vx8-dr. The basic protocol is to setup the vx8-dr to transmit APRS location messages on a regular basis and measure the battery voltage over time. The voltage measurement comes from the vx8-dr itself. Yes I’m trusting that it was accurate but with the test running the same way across all the types of batteries, I think it’ll be a valid test. Besides, the reading from the vx8-dr is what I’ll be using in the field.

The vx8-dr can use 4 different types of batteries: 2,000 mAh lithium ion, 1,100 mAh lithium ion, AA alkaline and AA lithium ion. I ran a set of tests over the course of 8 days.

2,000 mAh lithium ion

vx8-dr 2000mAh

relative time voltage
0:00:00 8.2
1:04:28 8.1
2:05:27 8.0
3:14:07 7.8
4:58:20 7.7
6:04:41 7.6
6:46:38 7.6
7:43:17 7.5
8:34:24 7.5
9:30:31 7.4
11:21:29 7.3
12:22:51 7.3

1,100 mAh lithium ion

vx8-dr 1100mAh

relative time voltage
0:00:00 8.2
0:00:00 8.2
2:15:51 7.8
3:06:14 7.7
3:58:57 7.5
6:27:44 7.4
8:01:43 7.1

AA alkaline

vx8-dr aa alkaline

relative time voltage
0:00:00 4.2
0:43:23 3.9
1:26:23 3.8
2:02:19 3.7
2:46:57 3.6
3:43:42 3.6
5:02:25 3.5
8:15:51 3.3

AA lithium ion

vx8-dr aa lithium ion

relative time voltage
0:00:00 4.5
1:23:24 4.3
3:33:47 4.1
5:19:44 4.1
9:04:00 3.9
9:55:36 3.9
20:56:06 3.7

I unintentionally left the vx8-dr running over night and the battery level only dropped 0.2 volts. That was very unexpected.

Conclusion

I came to a few conclusions from this test:

  1. Measuring the voltage will most likely be a good way to keep track of when to change batteries.
  2. Just measuring the voltage doesn’t provide any data on how well the transmitter is working with the available voltage / power.
  3. Judging by the AA lithium ion battery test, I need to re-run the 2,000 mhA and 1,100 mHa test. It looks like the voltage level of the lithium ion batteries may not be linear. That may put a kink in the use of the battery voltage to determine remaining power.

I’ll re-run the 2,000 mHa and 1,100 mHa tests over the next few days to see how long the batteries last. Perhaps lithium ion batteries don’t have a linear voltage level over regular usage.

More importantly, I need to expand the tests to determine the output power level as the batteries are used. I’m not yet sure how to do that. Maybe it’s time to invest in an RF power meter.

 

Tags :

Portable APRS Server

I had a thought yesterday, in the event of an emergency the internet connections will be down. Wouldn’t it be useful to have a localized APRS display to show the location of various people. Set up a radio to listen for APRS transmissions, send them to a small computer, run a webserver on the computer w/ OpenStreetMaps on it, have a wifi connection and power connections. Drop it all into a portable case of some sort and it can be made available to a local group of people who need access to the data.

There’s a bunch of unknowns here but it could be an interesting project. Would it be useful? Reliable? What else is needed?

Tags :

Net Troubles

I wasn’t able to get on the ORCA net last Thursday. It’s worked before and I haven’t changed anything on my radio. Very strange.

In talking things through with Peter (KO6R), we realized that the batteries ran low. Without enough power, the VX8-DR wasn’t able to transmit far enough to be heard by anyone. I was running with the AA battery pack instead of the lithium ion battery pack. It looks like AA batteries just don’t have very much power in them and they ran out of power much faster than I expected. I’ll need to experiment a bit to see how the various battery options last w/ the HT.

Tags : ,