A Trekdesk and a couple of monitors over my treadmill. With this, I measure conference calls in terms of miles, not minutes!
This past weekend I finally started building the ultimate case to house the LNA and FM BCB filter for the rooftop multireceiver project.
Last night I did the final bit of wiring, installed a DC-DC buck converter to take +12 vdc and knock it down to 5.0 vdc.
Here’s some pics of the project.
The two separate die-cast enclosures are the LNA (lower) and the FM BCB notch filter (upper). There’s a LM2596 DC-DC converter in the lower left, and a weathertight Ethernet connection in the center-right wall. All external RF connectors are N. All internal connectors are TNC. All coax is RG316 double-shielded. I used an Ethernet connection to get power to the box and to allow the future addition of either an Arduino or an Raspberry Pi for telemetry purposes.
The overall enclosure with installed components. Got this particular enclosure for about $22 delivered – someone in Santa Maria CA didn’t want it. Was missing the base plate upon which I mounted all the components – fortunately, I had a piece of Al in the garage that was a near-perfect fit! Would have cost nearly $70 new.
The left N connector is the antenna input, the right N is the assembly output to the shack. The Ethernet connector is for future expansion and dc power.
Installed on the roof temporarily (will be mounted on the tripod to right ultimately) with power supplied by batteries in the Tupperware container.
The batteries (Eneloop AA x 10) are charged with a newly modified solar panel install. Again, over the weekend I cut some Al extrusions to replace the old way I’d attached a single 5 w solar panel, now it supports two 5 w panels for a total of 10 w. This almost guarantees sufficient charge for long winter nights.
In addition, the box mounted on the tripod already had a PoE Ethernet connection from the main rooftop unit. I use that 12.6 volt PoE through a single 1N4001 rectifier to source power to the LNA box when the solar panels can’t provide enough charge to keep the batteries up. I want to add some telemetry (through a future Arduino or similar on the roof) to measure voltages, temperatures, etc.
As I was rebuilding a pair of Windows boxes here, I ran across a reddit thread that pointed me to this nice anti-virus comparison table that seems to get updated regularly.
The new Raspberry Pi 3 goes on sale today. I already have a number of R-Pis running stuff around the house, a couple for ADS-B receivers up on the roof, and while the built-in wired connectivity was always reliable and a terrific feature, the new wireless connectivity, especially the integrated IEEE 802.11n, means that I can remote these bad pups even further, maybe using solar or other wireless power sources.
Oh, and Happy Leap Day! Google has a cute cartoon to commemorate this calendar artifact…
Everyone, including Elon M,, seems to be hot and heavy to go to Mars. I know Mr. Musk is envisioning a world to terraform and retire to, but the Moon is so much closer and so easy to work with. And, from the moon, you can do advertising.
Up on the roof is a GigE 8 port switch, two IP cameras, an ancient Linksys WRT54G router to provide Wi-Fi coverage around the property, a Raspberry Pi and SDR USB dongle running as an ADS-B receiver, and a UBNT 5 GHz bullet supplying a link to another Bullet over on the roof of the garage, to get my local network out to stuff in the garage. In addition, in the box on the roof, there’s a 12 vdc to 5 vdc converter to supply power to the Raspberry Pi. All told, this is only about 20-25 w of power consumption, which at the fused voltage of 12 vdc, should only be 2 A or so. So I can’t see the reason that a 10 A fuse is getting taken out.
I supply both 12 vdc and 48 vdc to the roof via a cable from the shack. There’s a 12 vdc to 48 vdc 150 w dc-dc converter in the shack to generate the 48 vdc. Both that and the raw 12 vdc are sourced from the 12 vdc deep-cycle battery in the shack, through a single 10 A fuse.
For some reason, yesterday I discovered the 10 A fuse had blown. Not good. I replaced the fuse and it appeared to start working again. However, I see this morning that the cameras and wx station were last heard from around 0641 local, so I suspect the fuse blew again.
Looks like this afternoon will include a visit to the roof to see what’s up.
Welcome back to the WirelessJon blog – I’d taken a long respite from blogging, but it’s time to dive back in.
Six weeks ago I decided that it was high time that the WirelessJon household had a professional-grade weather station. I’d been toying with $90 LaCrosse Technology all-in-one wx stations from Costco for many years, and each year or two or three I’d have to replace the hardware because something had stopped working. As well, the LaCrosse stuff didn’t have a nice way to fire the data to WeatherUnderground (or any other online service), so I had to run WUHU on a PC that was also running the LaCrosse Heavy Weather software. It was a kludge at best!
One of my long-time friends and fellow ham Dave KD7DR is a local expert on weather stations. He’s been using both Peet Bros. and Davis Instruments stations for a long time now. He suggested to me that the Davis had better hardware quality – he uses a wired Davis Vantage Pro2 atop Pinal Mountain, an ~8,000′ peak about 60 miles east of Phoenix, and it’s survived well over the years. I also had some personal experience with a Peet, as I’d installed one at a 5,000′ site above the Tehachapi Loop in southern California back in 2008, and except for losing the anemometer cups to a heavy ice storm, it’s been a pretty solid unit as well.
After some discussion and consideration, I chose the Davis Vantage Pro2 as the baseline. The next decision was whether it would be the wired or wireless version. The external sensor location is on the roof of the house, and I already had a 2″ cable conduit running out to the general location. However, that conduit is stuffed with other cables, including RF coaxial cable, and I didn’t know whether I wanted to pull yet another cable through the already-tight conduit and didn’t know how the slight leakage from the RF coaxial cables might impact the weather sensor readings. Time for more investigation.
I went through the Davis literature and discovered that they’d had the foresight to put “low-pass” filtering on the sensor lines, so what coupling there might be should be alleviated by the filtering. As well, my RF coaxial cable is either LMR400 or LMR240, double-shielded, very low leakage, and none of my transmitters are more than about 50 watts, so it’s likely that the amount of power deposited onto the wx sensor cable leads would be pretty low. It would also be common mode, so that might help reduce any concerns as well.
The last factor for wireless was the choice of RF frequency that the Davis unit uses. It employs some sort of proprietary frequency-hopping transmitter in the license-exempt 902-928 MHz band. Given that I’m a ham and may want to operate in that band (actually, I do have some equipment but it’s currently receive-only in the house), I decided that I didn’t want to pollute that band any more than it was already.
Importantly, I had to look at the type of cable that connects the Davis weather sensors to the console and what connectors are used. Again, a few checks using Google and I found that the cable supplied by Davis is just 4-wire “phone” cable, much like the old silver-colored soft, flexible wire which connected old-time wired POTS phones to the wall outlet. Except that the Davis wire had a UV-resistant jacket. Kind of necessary for an outdoor install, especially here in Arizona. The connectors were the clear plastic 6P4C “RJ15” type, readily available in my garage, as was the crimp tool. Finally, the way the cable was constructed was straight-through, so there was no need to worry about some custom wiring configuration.
The next consideration was how much weather did I want to sense? Since I tinker with solar power a bit, both on the house and on the truck, I’ve always wanted to know the total solar insolation at the house, and ultimately be able to calculate the peak, average and cumulative power/energy delivered to the ground. And while I was spending all this money, I decided to get the UV sensor as well. Never know. I chose not to get the active aspirated temperature sensor, but I believe I could add that on later if I choose.
One final consideration in the wired vs wireless debate is how to get power to the sensors. Weather sensors are just like any other device – they need some juice to allow them to make measurements and report back the information. The wireless Davis unit employs an add-on small solar panel and built-in battery to collect and store electrical power to run the sensors and the wireless transmitter. According to everything I’d read, it works well and the batteries have a good lifetime. But, here in Phoenix, things exposed to the sun get very warm, and batteries up on the roof are going to be in a fairly harsh summer environment. Also, the failure of the most recent LaCrosse unit appeared to be a failure of the built-in solar cell to charge the little NiMH batteries in the anemometer. Using a cable would allow me to avoid completely any concern over having to change batteries in the future.
Now that the wired vs wireless decision was made, it was time to look at the ability for the Davis to push data to the internet. They have a number of interface dongles which plug into the display unit. One is a serial port connection, which requires a PC to run software to manipulate and push the data to the web, and the other was a thing called the WeatherLink IP, a slightly different dongle which plugged into the same socket on the Davis display head but instead of serial had a nifty RJ45 socket and talked IP. It seemed that this allowed me to divorce the entire process from a PC. The one primary drawback was that the literature indicated that the weather data could only be pushed once every 15 minutes. At this point, that seemed ok (i’d come to be unhappy with this later).
That was that. I called up Andrew Revering at Convective Development and placed the order for the Davis Vantage Pro 2 Plus wired station and the WeatherLink IP dongle. Less than a week later a big box showed up on the doorstep, and the day after Thanksgiving I installed the entire thing.
Later on I’ll talk about some of the good, could-be-improved, and mediocre things I’ve found with the Davis, both mechanically and operationally.