Emlid M2 to Trimble Ez Guide 150 Success

PotatoFarmer · July 15, 2020, 5:00am

Hello long time reader, first time poster. Another farmer friend got me into these little Emlids now i am so close to using them for something useful.

So i have kinda solved one of the burning wishes of a lot of tractor guidance posts i have read.

I was able to integrate an Emlid M+ and also my newly arrived Emlid M2 to act as an external GPS radio for an Trimble Ez Guide 150.
So far it has only had success on the bench… due to some onerous protocol conversion.

Long story short once the Emlid Rover gets a fix its NMEA data is sent through tcp to my laptop running NMEA Router software. Then it exits on a usb to serial converter to a ShipModul Multiplexer. The Multiplexer is wired to convert the 232 to its native 422 input. The Multiplexer strips off all the uneeded NMEA sentences leaving me with GGA and VTG. But the wrong GGA and VTG, so it moves through a “Talker Code Mask” to convert “GN” to “GP”. Then it exits the Multiplexer only to get converted from 422 back to 232 again, then to the external Data port of the Trimble 150. It seems very stable and usable at a standstill.

The result is an obsolete Trimble 150 running guidance off an cutting edge emlid. I think this trick will work for the 250 as well but none newer. I am also working on emlid as base for EZ 500 and newer.

Once i get some more conversion hardware i will be able to take the PC out of the picture and attempt to cut my grass and seed potatoes with RTK! A dream hopefully come true.

I was wondering if the development team would ever consider allowing NMEA sentence selection on the M2, and then also being able to change the Talker Codes?

With those 2 abilities I would only have to convert UART TTL to 232 and the signal would be ready for AG guidance. I would also be able to sell a pile of M2’s as the Trimble 150 was built like a tank and will outlive us all.

tatiana.andreeva · July 15, 2020, 9:45am

Hi @PotatoFarmer,

Congrats on your first post! It’d be interesting to see the photos of the full setup once you finish it.

We’re thinking of adding these features. However, we hardly can implement it soon.

PotatoFarmer · July 15, 2020, 8:57pm

@tatiana.andreeva I understand exactly how far your team have brought your product so far, its very impressive. This is why farmers are all trying to use reach for everything.

I have some pictures of the bench test, as soon as its in the tractor i will post the results as well.

This is the NMEA router software and ShipModul configuration software watching the data stream during the test.

This is the Trimble 150 GPS diagnostic screen, as you can see it is reading data no problem.

The M2 with an RTK fix, The M2 rover will control the guidance through NMEA.

A Very exciting moment! I was able to set point A, just waiting to move to set point B to start guidance.

This cluster of electrical debauchery needs to get cleaned up with some converters. The wiring will be more simple and I wont need the computer on the tractor. The best part is all of the Trimble messages ( searching sats, high DOP, Low Update rate 1hz, No guidance, No corrections, Low accuracy float) all work.

Proof of concept in one screen, you do not get signal like that running WAAS.

Field testing will happen as soon as the new hardware is delivered. Then to figure out a way to somewhat weatherproof the components as my Tractor is open station.

EBE111057 · July 16, 2020, 1:01am

Great work !!! It’s rewarding when all things come together !!

tatiana.andreeva · July 16, 2020, 10:08am

Looking forward to seeing the photos from field testing!

Skegpro · July 24, 2020, 1:43am

yes. I wish Emlid would focus on the ag sector more.
Primarily around the integration of connections with legacy auto-steer equipment.
Ie Trimble CFX-750

PotatoFarmer · July 24, 2020, 2:32am

Haha @Skegpro,

I am currently working for solution on your CFX-750. It involves Trimble UAV development boards. They are only slightly more cash than an Emlid M2, but they look like they may create the needed RTCM3.0 or 3.1 Messages.

CFX-750 was first released in 2010, my mind is blown. I thought they only started selling them up here a few years ago. But they still are worth $9000CDN used with RTK unlock, display only.

A little more RTCM message flexibility would make EMLIDs much more easy to integrate into other systems. A little bit of NMEA flexibility and they become a new breath of life in older L1 only systems.

I get when Emlid dropped RTCM2 support, L1 corrections are kinda ancient.
RTCM3.x has 3 versions already. If the .x is wrong they do not work.

But hopefully after they have the Survey enhancement under wraps, Agriculture and Autonomy will be next.

The protocol converters are starting to show up, my soldering iron is waiting. Just waiting to see how the M2 actually functions in the field.

PotatoFarmer · July 28, 2020, 9:53pm

Getting very close to field testing, i think the bare minimum to make the rover work has finally arrived.

In preparation built and Installed a permanent base T post. I want to make sure all my root vegetables are straight in line for the rest of time.

Also will be a good test of lora radios abilities at 2m off the ground.

PotatoFarmer · August 2, 2020, 9:32pm

Ok the bare minimum to make it work without PC has arrived, and Oh Boy Oh Boy what fun.

Here is the inside of what i like to call the “Back to the Future Box” Taking GNSS NMEA and stripping extra sentences and converting to GPS NMEA in real time.

I was not 100% sure everything was going to go well with testing so the external GPS is zip tied to the tractor ROPS. The first time i plugged in the Emlid M2 on the tractor it refused to work. I find that it is not so much the length of the cable, or the power source (though 2A or greater helps), but the wire size inside the USB cable that causes the M2 to stall on an orange light at startup. I still have a 12v to USB power supply and strain reliefs coming to tidy the GPS enclosure up a bit. As well as some cables to permanently mount the Lora and GPS antenna with more separation.

Here is the solar powered base station, still waiting on cables. Not the most ideal mounting place with the lora and M2 zip tied to the solar panel. It was 31C out and no issues with overheating. Even in a yard full of buildings and trees and elevation changes the Lora never dropped once. Had GPS, Glonass, Galileo, and Beidou all set to 1hz; ARP Coordinates at 0.1Hz. I plan on taking the rover for a drive soon to see how far I can be away from the Base station .

Waiting on magnetic mounts, so taped a Bic Pen to the old WAAS antenna base, then taped the Tallysman to it, and wrapped a rag around it for support. This setup looked horrible but worked pretty good. I had to completely re-tune the old Trimble 150, at first it was swerving everywhere uncontrollably. Soon got it down to +/- 2 inches driving.

Out in the open there is no shortage of signal at all. The M2 never dropped the fix under normal field conditions, even right beside tall trees on one side. Driving under tree canopy to get to the lawn it would loose fix but immediately recover once out in the open again. Very impressed. Way better than L1 WAAS, was lucky to get 8 satellites, and had a 10min correction warmup.

Trying to get the max driving accuracy I tried something most farmers up this far north will never do, move the antenna from the top of the cab to the nose of the tractor. The cantilever tractor canopy also shakes pretty bad so I thought this would gain some repeatability. To my surprise it barely affected the number of satellites detected way to go Tallysman Helical on the M2! On another field tuned the Trimble 150 driving accuracy down to +/- 1 inch!!! But it was leaving 1/2" strips every once and a while with the Trimble set to 2" overlap. I lifted the 72" mower to measure the blades and it was actually 70.5" wide blade tip to blade tip. Staring at 1"…0"…0"…1" for an hour I switched everything to metric for more measurement resolution.

This is the worst area for GPS ever tall trees to the North, South, East and West. Plus the road slopes up about 12’ on the south side where I am standing. L1 WAAS on only two occasions could only very briefly hold a line in the middle. The M2 only went to float here twice and wandered a little, but I found that you have to wait a few seconds on either open end or the Trimble would oscillate at the start of the swath. I think its safe to say the 15 year old Ez Guide 150 chimera is a success at +/- 2cm steering accuracy. By the time that steering shift makes it back 3 meters to the implement it is basically 1cm or less.

This fall I will also be testing this system on a large 4WD tractor Discing and also a Swather running Ez Guide 150’s. Still working on Integrating this box to talk TSIP for the Trimble 250 and 500.

I want to mount the Tallysman to a small Ardupilot vibration isolation platform and test for improvements. As well as give the whole system a bit more polish as the parts arrive.

In conclusion was it worth it? Yes totally. But I do not recommend unless your really good at tuning control loops, the steering setup of the Ez guide 150 is not the most user friendly.

Emlid M2 really can make if feasible and affordable to upgrade your older L1 GPS system, to modern L1/L2 GNSS.

jp-drain-sol · August 3, 2020, 10:52am

What are the dimensions of your project box in the tractor? Where did you buy it and do they offer different sizes?

I want to do something similar with an M2 setup as a bases station but I don’t need it quite so large. If heat is or becomes an issue I plan on installing a computer processor fan in the box to circulate air

PotatoFarmer · August 3, 2020, 12:26pm

I think the project box is about 10”x10”x3”

I get them, plus most of my electronic parts off of Aliexpress from china. They are extremely reasonable. But there sometimes is a couple month wait. Ive never had anything not turn up. They come in all types of shapes colours and sizes, different mounting options, some already have data ports and fan holes cut if needed.

If i need them quicker we also have a Electronics shop in the nearest city that sells the same ones for substantially more.

Just search “metal project box” on ali, one that size is about $16US free shipping.

PotatoFarmer · August 23, 2020, 7:35pm

Update; Cellular Test, Vibration Damping, Update Rate Issue

I have done some more testing and upgrades of this system.

I solved an issue with NMEA update rate. The position data to the 150 would become erratic jumping between 2HZ and 6Hz. I fixed this by cranking the input rate to the shipmodul to 57600bps. This allows it more time to strip the unneeded NMEA sentences from the data and convert them before exiting at 38400bps. I have finally started learning C# and feel I will not need a shipmodul any longer. I should be able to write a converter for Arduino. Was surprised to find C# is how my graphing calculator used to program in highschool, so already kinda learned the basics.

I just recently installed a permanent RTK basestation at @Skegpro farm. It has a cellular modem so connected the M2 through the iPhone hot spot. This location is 30km away, it truly predictably performed at 7mm+1mm/km even in these trees. Half the time you think the base station was under 10km, but then it would wander just under 2". Still plenty good performance results for most grain farming operations.

I had to try vibration isolation. Total Success DO IT!!!

Its a very inexpensive ardupilot isolation platform. Modified it so that the antenna cable hangs freely through a large hole in the center. Then the bracket is held by four small magnets in case of accidental contact.

Watching the position on the Emlid app you could really see the difference. Without isolation it is constantly chewing through position numbers, with isolation it rarely changes. The difference in machine control is very noticeable. It now performs like much more modern RTK guidance. Just wish the 150 could do 10hz, but i guess that was just not a thing 15 years ago.

Tried skip a swath pattern off the local M2 base. The system with the vibration damping now holds a 1cm or less line in ideal conditions, and a 2cm line on a 12degree or 20% side grade.
drifts a tiny bit more in heavy trees.

I am starting to get some attention from other farmers who think I am paying the $1200 RTK fee to cut my grass, HAHA no chance. They seem pretty dumbfounded that its an obsolete unit with drone GNSS.

Next Update will be real farming with the M2. I love cutting lawn but the M2 needs to control much more horsepower and many more hours of continuous testing.

Could not be happier, truly elated with the Emlid M2’s Precision Ag performance.

tatiana.andreeva · August 24, 2020, 12:41pm

I love reading all the updates on your project It’s always a pleasure to know how our users incorporate our devices for their needs!

Skegpro · August 28, 2020, 1:52am

@PotatoFarmer
New base station is working great, to bad it isn’t a emlid

Skegpro · August 28, 2020, 1:54am

PotatoFarmer · August 28, 2020, 2:09am

Good to see your new gear is working great!

Amazing what rtcm3.x legacy messages can do.

Hopefully with Agopengps I can make a Emlid only solution.

davehofer1993 · August 28, 2020, 2:28pm

Nice pics . Isn’t that canola a bit green yet

PotatoFarmer · September 1, 2020, 2:53pm

NMEA conversion on Emlid !!!

Big News, i posted it here to make it easier to find. Emlid has added NMEA Talker code transformation and message selection to their long to do list!!!

The countryside will look a whole lot straighter!

I think the boating community will be excited as well.

PotatoFarmer · September 25, 2020, 5:58am

Final UPDATE!!! M2 Controls 325HP 4WD Tractor!!!, Running with no Corrections, Pass to Pass Explained, No time for WAAS, Mega Satellite Reception

Hello this is the final instalment for this project. Thank you to everyone reading along, hopefully you will have fun reading about my next two upcoming projects; Emlid to TSIP to Trimble Ez Guide 250, and External Radio Corrections for M2.

The Trimble 150 conversion was very successful. After an almost flawless summer of cutting grass and doing drainage with the M2 I decided to lend it to a friend who uses an EZ 150 for larger scale grain farming. After Combining he wanted to test it harrowing fields with his 325HP Stieger Panther III 4WD articulated tractor. Nothing like free Beta Testing!

If you look really hard you can see the Tallysman helical Antenna at the front of the cab. After installation, calibration and syncing with his phone to use a cellular base. I left to go Combining at another farm.

He started sending me pictures of how straight it was driving. The land is quite hilly and treed in places, they have had longstanding issues using GPS with WAAS corrections due to this, getting only 4 to 6 gps satellites max with many dropouts. Also the geostationary corrections satellites come in at a very shallow angle and are easily blocked on the north side of hills.

As you can see the M2 does not have a signal issue whatsoever, this screen capture was taken right beside a row of grain bins in a treed yard. 38 satellites not bad. But the next day I got a phone call with issues connecting the M2 to their phones hotpots. Needing to get the work finished they tried running without corrections and the results were very surprising. I told them next morning I would come update firmware to try to fix the problem.

The uncorrected M2 was running about 14cm Pass to Pass, better than WAAS that runs 20cm Pass to Pass with corrections my mind was blown. Also they did not have to remark or nudge the line even without the corrections. With WAAS you are doing this constantly each pass. Both the farmer and his Father were very impressed by the Emlid systems performance. I think the Emlid really has a quality algorithm for computing position even before corrections are applied.

Agricultural GPS is always stated in Pass to Pass Error. It is basically double the error of any single measurement. If your first swath is out by an inch one way, when you turn around for the next swath it could be out that same distance the other way in that same spot. This gives you “Misses” or “Strips” left behind. RTK basically fully eliminates this. There are various levels of very expensive pay to play correction services available to farmers.

This is an example of WAAS drift during Combining. Within about a minute I drifted over about 1.5feet. Usually with WAAS, if it is running perfectly I give it one foot to wander back and forth in the header. Unfortunately we cannot fix measurement issues with post processing everything is done on the fly and quality live position information is required.

Here are two more photos of the WAAS wave. It is really not efficient. If you loose a foot on forty passes, you have to make an extra pass. A combine usually harvests at 2.6MPH so extra passes are costly. Pretty soon after many acres you have wasted the time, energy and resources of 4 to 8% of your total costs using WAAS vs RTK. If you are driving by hand you might as well throw your time and money in a fire. Misses, overlap and over application add up quite quickly in time and extra expense in any area agricultural coverage work.

Thank you for those who have read to the end, I am very excited to see if i can integrate Emlid into more machinery.

tatiana.andreeva · September 25, 2020, 8:56am

Hi @PotatoFarmer,

Congrats on finishing this!

Is it the same iPhone connection issue you mentioned in another thread? With an iOS hotspot, I’d recommend following these steps while connecting to avoid any difficulties:

Add the iOS hotspot to the saved networks using “Connecting to a hidden network” option
Turn Reach off
Enable Wi-Fi on iPhone and make sure it’s not connected to any Wi-Fi network
Open Personal Hotspot page on iPhone and allow others to join
Turn Reach on
Keep the Personal Hotspot page opened when Reach is loading

After that, Reach should automatically connect to iOS hotspot.