Robomast Motorized Mobile Radio Antenna Mount

This project started in 2006 shortly after I got my amateur radio license. The idea came to me one day as I drove my truck into the Hartford Bradley Airport parking garage in mid-2006. I had a Comet SBB-5 antenna on the roof with a magnetic mount. The antenna has a foldover option, but the truck was tall enough that it still wouldn't clear. I had to get out and remove the antenna from the mount.

At the time, there were no motorized antenna mounts on the amateur radio market, so I found a linear actuator on eBay and designed Robomast around it. The actuator has 6" of stroke length, but I couldn't use more than 3". Fortunately, the upper limit switch was adjustable. It was also designed for 24 volts, so it would have to run slowly on my truck's 12 volt system.

The first pivoting head was an ugly concoction. It worked, but it was not thought out well and very bulky. That lasted just 7 days until I came up with a major improvement. The 1" x 2" aluminum mast and the actuator geometry worked very well though. The first head was a great opportunity to prove the geometry and work out some design challenges.

The second head design was a major improvement. Built mostly of 3/8" thick 6061 aluminum bar stock, it has a stainless steel axle hidden between the side plates, a stainless pivot shaft on the side for the operating rod and only two exposed fasteners on top. The underside of the NMO connector is protected in a waterproof cavity milled into the top and bottom plates.

I should also mention that soon after building Robomast, I upgraded to a Comet SBB-7 antenna which is 15" taller than the SBB-5. There are now commercially available motorized mounts, but still, none can handle the SBB-7 because of the torque loads caused by its weight and additional leverage. Robomast is very stout and can handle much larger antennas, although I know of none larger for the VHF/UHF amateur bands.

A simple DPDT toggle switch with a center-off position in the cab allowed me to raise and lower the antenna by 90° with ease. Limit switches inside the actuator automatically stop it at the top and bottom positions. One of my early design ideas was to limit or eliminate any carbon steel since the unit would live outdoors, exposed to the elements. There were a few steel components early on, but they were soon replaced with aluminum. Today, Robomast is constructed entirely of aluminum and stainless steel with composite axle bearings. This has worked well as it has endured the sun, wind, rain, snow and ice for years without a single failure. It has been cycled thousands of times.

First Head Design, September 2006	Second Head Design
Head In Down Position	Head In Up Position

Weaknesses

The one thing I've never cared for is the exposed coax cable. That has always bugged me, but the same piece of coax has withstood years of service and harsh exposure without any issues. Still, a completely redesigned Robomast has been rolling around in my mind for a long time. There is a prototype of it under construction and that has no exposed operating rod or coax. Everything is inside the mast. But that's for another web page . . .

Also, Robomast has no overload provision for tree strikes. A clutch of some sort or a telescoping operating rod with a friction lock would allow it to pivot if the antenna strikes an immovable object. That's difficult to implement. So far, I've only had one major incident though. More about that below.

Another thing that bugged me was not knowing for sure whether the actuator was in motion. Mounted behind the cab on the driver's side, it's hard to see whether the antenna is still traversing. The 24 volt actuator running on 12 volts also moves a bit slower than I'd like. Transit time is around 12 seconds which is about 5 seconds too slow for my preference. But that problem is now solved. More on that below too.

And of course, Robomast is a tall vertical unit specifically designed to be installed in the front corner of a pickup truck bed. The commercially available units are much smaller and can be bolted to a roof rack, so they're more versatile. But for my purposes, Robomast is perfect.

In October 2010, the truck that Robomast was mounted to blew its transmission nearly 2,000 miles from home (during a cross-country move!). I ended up stripping my personal effects from the truck and leaving it behind to be sold later. Robomast was hastily cut free and transported home. Soon after, I got a new truck, but didn't want to drill a hole in the bed for mounting it. The bed mount had always been a weakness anyway.

Over the years, Robomast had accumulated a lot of grime and some minor surface pitting. It had never been painted and was just bare aluminum. I started refurbishing the unit, which included disassembly, a thorough cleaning and rounding off many of the sharp corners in the milling machine.

The top of the actuator is suffering a bit from UV exposure. That will probably be the limit to Robomast's lifespan, but I believe I can protect it to last many more years.

Years Of Grime On Top	More Grime Below	Actuator Detail (After Refurbishment)
All Parts After Refinishing

At the same time, I was working on a new control with visual feedback so I could tell when it was transiting. A new power system was in the works as well. Then life got in the way and the project got derailed for several months.

Once back on track, I sanded and cleaned all the surfaces, then applied several coats of Duracoat titanium paint and several more coats of clear protectant. Duracoat is a tough chemically-cured finish, so that was allowed to set for a couple of weeks. Then I fabricated new mounts and attached it to my ugly green toolbox. This eliminated the need to drill holes in the truck's bed, although I did drill one in the belly pan under the back seat to allow the coax and power wires to go inside. This is a sealed penetration and the cables have a drip loop before entering liquid tight conduit.

The solution to the slow transit time was to power the actuator at 24 volts. This required a 12VDC-to-24VDC inverter. Since the inverter draws current when it's idle, I connected a relay so when the ignition is off, the inverter is unpowered. The relay connects 12V to the control switch to allow Robomast to work even when the key is off, although at the slower speed. When the key is on, the relay powers the inverter and switches it into the circuit so Robomast runs faster. The new transit time is about 5.5 seconds! It is perfect and not so fast that it puts extreme loads on the mechanism.

The solution to the control feedback came from my friend Willy (N1NKM) in Connecticut. I had a new rocker switch with LED's in the up and down positions, but I needed a circuit to turn them on and off when the actuator is moving. I really didn't want to run additional wires for that either. Willy's idea was to wrap a coil around a reed switch. The current draw from the actuator creates a magnetic field in the coil which closes the reed switch to turn on the LED's. The polarity of the circuit that determines the direction of the actuator also determines which LED lights up. Perfect! It took very little experimentation to come up with a working solution.

24 Volt Inverter	Current Sensing Relay
New Control Circuits	Installed Switch (1.8 MB video)
Robomast In Action! (1.1 MB video)

CRASH!

By the time I got Robomast installed on the "new" truck, 17 months had passed. I had never had a serious accident with it before, but of course on the first trip after installing the new and improved Robomast, I drove into a parking garage and forgot to lower the antenna. DOH!!! Fortunately, the expensive antenna survived (I had replaced the vertical elements with titanium and had to replace the base due to a failed coil, so I have a bit of money in the antenna itself). Fortunately for me, only the top deck of the head and the NMO connector were destroyed. It took a half day of work to machine a new deck and paint it.

Destroyed Deck & Connector

The New Deck & The Old One

New NMO Connector

New Cable Bracket

While I was making the new deck, I decided I had to do something about the unrestrained coax. As a partial answer to the problem (which is really only cosmetic), I made a bracket to fix the coax to the mast and restrict the movement only to the last few inches. This bracket was machined from a solid block of 6061 aluminum bar stock. In these photos, it sports a BNC passthrough connector, but they are not waterproof (I had them on hand), so TNC connectors are on order along with new coax. The old coax has served well, but my original design called for RG-59 size cable instead of RG-58. The new cable will be Times Microwave LMR-240 Ultraflex and will not require a spacer sleeve at the entrance to the head.

Cable Bracket Before Milling

On The Milling Machine

Closeup, Installed

Looking Up

The Next Generations

I did some experimentation and prototyping on a "Roborack" idea which was a 4-foot long aluminum tube between two vertical posts. The tube could handle multiple antennas and positioning would be computer controlled. You can see that work on this page. That project is where the 24V inverter came from.

I lost steam on the project when I realized that there was no overload provision and a tree or building strike could be catastrophic for the whole system. Adding a clutch mechanism could be possible, but not very practical. Shear pins might work, but I couldn't accurately design it to shear at just the right amount of force. But much of the research has worked its way into other projects.

Robomast V2 (or V3, depending on how I think about it) is a major aesthetic improvement. It still uses the same 1" x 2" rectangular tubing, but everything is internal to the tube. The coax would be completely hidden and protected and it has an easy-to-implement overload provision to allow it to survive tree strikes. It does require some really complex machining on an internal rotor, but I've already finished that work. It has a couple of minor shortcomings, but I may end up finishing and deploying it at some point.

One thing that really interests me is an idea that came about during the RoboRack project. That is to let a computer automatically raise and lower the antenna. My vision is that as I approach an obstacle, when I hit the switch to raise or lower the antenna, a computer would note my location and heading on GPS and ask if I want to remember that location. If I store the location (a single press of a button), the next time I am near that spot heading the same direction, the computer would set the antenna height without me having to remember it.

This ties into another project which involves having a full time GPS running aboard the truck anyway. That GPS engine will drive several other applications. I intend for this to run on a dedicated microcontroller -- probably a Fez Panda board with a 32-bit ARM7 chip. I have a separate math coprocessor to decode NMEA sentences from the GPS and to do the proximity calculations. I have also prototyped the user interface on a small color touch screen. The graphics are ray-traced for 3D realism. Still, much work remains to bring it to life.

Color Touch Screen	Display Microcontroller
NXP ARM7 Microcontroller	User Interface Demo (3.8 MB video)