Has anyone used a Cradlepoint/Red Lion cellular modem for remote control panel comms?

Hi all,

I have a client that has a water tank with a loop powered level sensor; the 4-20mA loop goes through a telephone line, half a mile away to the tank. Surprisingly, the client is complaining about a noisy signal!

We're discussing various options to upgrade this signal, as you might imagine. Radio link is a no-go, path survey came back pretty bad due to terrain and we'd need to put up 50' at treatment plant, 40' at tank. Trenching a dedicated fiber optic line would be prohibitively expensive, not surprisingly.

One potential solution we're discussing is getting a cellular based modem (like a Red Lion or Cradlepoint) and using that to send data. Another local client is doing exactly that, and it sounds like a good solution. However, I've never set up anything like that and have no idea how one would set up comms. I realize the modem would need a static IP address, but beyond that I have no idea. Wondering if anyone here has set up something similar and could provide some insight? Thanks!

a couple of these: https://arcelect.com/FSKshorthaulmodem.html.
and a couple of MicroLogix 1400 controllers. no monthly bill with cellular.
You can do Point to point or point to multipoint if you have any additional water tanks, pump stations or whatever.

Leased lines from the telco are not uncommon in Water/Wastewater. It probably has been working fine for decades and they're definitely paying for it - maybe they can contact the telco and ask them to look into it?

TWS said:

a couple of these: https://arcelect.com/FSKshorthaulmodem.html.
and a couple of MicroLogix 1400 controllers. no monthly bill with cellular.
You can do Point to point or point to multipoint if you have any additional water tanks, pump stations or whatever.

Click to expand...

Interesting idea, though the PLCs on site are GE RX3i so I'd use a Versamax controller. I'll take a look, thanks!

bjh said:

Leased lines from the telco are not uncommon in Water/Wastewater. It probably has been working fine for decades and they're definitely paying for it - maybe they can contact the telco and ask them to look into it?

Click to expand...

It's actually not a leased line, it's literally the 4-20mA signal being sent through a twisted pair in the telco cable, not shielded. How AT&T even allowed it to happen blows my mind.

Quote: It's actually not a leased line, it's literally the 4-20mA signal being sent through a twisted pair in the telco cable, not shielded. How AT&T even allowed it to happen blows my mind. This type of communication line was VERY common in the 1930's until the use of fiber replaced copper lines.
Original transmitters were mercury time pulse transmitters with a cam that would rotate and a mercury switch to complete a circuit 3-to 25 seconds that would correspond to 0-to 100% of full scale. Since the phone lines were lease lines the voltage could be as high as the phone system voltage. On the other end of the transmitter you would have a circular chart recorder or gauge with the pen driven by the same rotating cam arrangement that would drive the needle of the gauge or the pen of a chart recorder.

TWS said:

This type of communication line was VERY common in the 1930's until the use of fiber replaced copper lines.
Original transmitters were mercury time pulse transmitters with a cam that would rotate and a mercury switch to complete a circuit 3-to 25 seconds that would correspond to 0-to 100% of full scale. Since the phone lines were lease lines the voltage could be as high as the phone system voltage. On the other end of the transmitter you would have a circular chart recorder or gauge with the pen driven by the same rotating cam arrangement that would drive the needle of the gauge or the pen of a chart recorder.

Click to expand...

Wow, fascinating bit of history! Thanks for the information, I can now see how this system evolved.

I can also see how the term "leased line" could mean different things; I thought a leased line specifically meant data was being transmitted in a modern sense (i.e., bits sent via modems), but it seems there is history to that as well.

1/2 mile? I have some horrible radio paths that work beautifully with serial UHF radio modems. Here's a path that is 6.3 miles where the yagi at the remote site is 12' off the ground pointed at a 100' tall cliff and I still get > 80% message completion polling it ~ every 10 seconds along with 13 other nodes. At 1/2 mile, I would expect to get well over 90% success rate with antennas laying on the ground pointing away from each other.

OkiePC said:

1/2 mile? I have some horrible radio paths that work beautifully with serial UHF radio modems. Here's a path that is 6.3 miles where the yagi at the remote site is 12' off the ground pointed at a 100' tall cliff and I still get > 80% message completion polling it ~ every 10 seconds along with 13 other nodes. At 1/2 mile, I would expect to get well over 90% success rate with antennas laying on the ground pointing away from each other.

Click to expand...

Interesting. The radio path survey indicated it would be a real bear, but hell there's nothing as simple as trying, right? The one issue I'd worry about is the area was burned over a few years ago during a wildland fire, so the vegetation will only get thicker and thicker and so far as I understand it, UHF is prone to vegetation blockage. I'll post the path survey below.

I use Cradlepoint and they work fine. They are tilted toward supporting PC's as opposed to PLCs so keep that in mind. Their support staff is a little weak if you have to support something wierd, but they will work with you.

Prosoft have a similar process line I have not explored but I have been happy with all their other stuff I have used.

Whilst that path is definitely obstructed, I'd just give it a whirl and use the packet transmit test function in the trio JR900s to see how many dropped packets you get with say, 10,000 packets.

The path study shows -79dBm receive signal - I know that's with the high antennas on both ends, but that is plenty of fade margin from the Trio receiver minimum RSSI of -120dBm (from memory).

If you get less than -100dBm both ends, and better than 99% success rate on the packet test, I'd be happy that the link should be reliable enough for messaging one level reading from tank to plant.

I'd be happy to run another simulation for you if you want to pm me the coordinates, using a more achievable antenna pole height.

You could use an FSK modem like others have suggested, but I've started to recommend clients move away from using leased lines here in that manner where possible... its only a matter of time before the Telcos start abandoning the old copper circuits, and this sort of thing won't fly once the area is upgraded to fibre (as is happening reasonably quickly here)

defcon.klaxon said:

Interesting. The radio path survey indicated it would be a real bear, but hell there's nothing as simple as trying, right? The one issue I'd worry about is the area was burned over a few years ago during a wildland fire, so the vegetation will only get thicker and thicker and so far as I understand it, UHF is prone to vegetation blockage. I'll post the path survey below.

Click to expand...

Although point to point is often desirable, due to having less equipment. Usually this would mean using a lower frequency radio to help with interference, which also equates to a slower network link. If terrain permits, could you put in a repeater station that both sites have a more line of site path to?

chelton said:

Although point to point is often desirable, due to having less equipment. Usually this would mean using a lower frequency radio to help with interference, which also equates to a slower network link. If terrain permits, could you put in a repeater station that both sites have a more line of site path to?

Click to expand...

For reading a handful of registers from an RTU or PLC, speed really isn't an issue in this application. You could do this particular job at 300bps if you really needed to. They only need a level reading, and probably a few auxiliary items like DC supply voltage, mains power status etc.

Actually, thinking about that... does the site have a power source currently, or was this a loop powered level sensor?

Saffa said:

For reading a handful of registers from an RTU or PLC, speed really isn't an issue in this application. You could do this particular job at 300bps if you really needed to. They only need a level reading, and probably a few auxiliary items like DC supply voltage, mains power status etc.

Actually, thinking about that... does the site have a power source currently, or was this a loop powered level sensor?

Click to expand...

I agree completely, but if you have to get into the licensed band below 900Mhz, for a reliable link, sometimes using the licence free band with a repeater is a viable alternative.

I would not expect good results using 900Mhz without line of sight, but still perhaps worth a trial.