Hello, I started a project 2 years ago that led me down one hell of a learning curve. I documented most of it on the EMLID forum but I have decided to move most of the information here as I continue the project, now that it is very F9P based.
This first post is a condensed version of the original thread, its going to be long, but more action packed and I will be updating my progress here.
The original project that started this was to help another friend who used Trimble RTK, trying to get a personal cellular Base Station for far less than $25K. The objective was to do exactly what @torriem finally figured out how to do. Before this I thought GPS guidance was a valuable but overpriced tool that I would never use much and knew zero really about it.
I was roped into this project by my friend because I had SCADA network experience, and do industrial instrumentation control and PLC programming. So he thought I might be useful 
. It turned into one massive hobby.
I eventually ended up figuring out how to get his cellular base completed, but also ended up with a couple Emlid M2’s in the process.
Look at the happiness on this mans face, he owns his own base station. Not paying $1200/year/reciever will make you this happy too!
A old farmer sold me a used box of “not sure if it works, pretty sure this one does not work” Ez Guide 150 equipment for $250. I custom installed this system on my John Deere 3038e. After trying use WAAS to cut my lawn I found though it could drive straighter than I could, with only 72" equipment it was pretty miserable to use. Even on a good day, 7 Sats, I needed about a foot of overlap. The old L1 receiver would loose signal pretty regular too.
But reading its user manual found it could take NMEA GPGGA, and GPVTG only at 38400bps and 5Hz, no extra sentences, no extra GN talker codes, just like an old Trimble 132 receiver output.
The EMLID output NMEA 
, but two problems. The first the NMEA all had GN talker code, and it output all the NMEA sentences possible. The 150 rejected this input outright.
So I searched for a workaround and found “shipmodul”. It was able to sort NMEA sentences and change talker codes. All for the low low price of $210. And it worked!
NMEA YEAH!
************* If you have the F9P you can just configure it to change the talker codes, and output just GPGGA and GPVTG 5hz. **************
Now I was able to cut grass, using dual band rtk, off a base station in my own yard. The old Ez 150 smartened right up and drives extremely straight and reliable after tuning. It still is a little wobbly at the start of the line, but I think that is due to its very much older gyroscope / compass. I notice it most going straight north on that 0-359deg crossover.
The EMLID M2’s fix fast in seconds and work extremely well beside tall trees even off a cellular base 30km away. But at EMLIDs heart is the Ublox F9P.
So after lending this converter to a farmer that I help he was hooked, The Stieger still uses a ez150, “but will it work with the Ez250 on the seeder?” No :(… the 250 will only will allow guidance from TSIP data.
So could I even get TSIP to work with the 150? After much searching I found right on Trimbles website the 1999 TSIP manual, and a suite of information and programs for testing TSIP. (A major reason i like Trimble more than DEERE GPS, no secrets) I also invested in a RS232 serial tap and monitor program so that I could see the active data.
Much to my surprise TSIP for the 150 consisted mostly of 2 NMEA repeated sentences inside some HEX data structure. Using the Ez 250 to send data to the 150 I was able to see that this was not that much data. but how do I change it?
So at this time I find @Torriems much matured 1008 injector, and the man codes with notes I can understand. I decide I should give C a go. Also at this time I decide it would be cool to have a colour screen for guidance and then i started looking at pricing… Even Ez250 pricing 
.
The friend who I helped set up the cellular base, I visit his farm and he is moving the Ez Pilot from the seeder to the swather steering column. He looks miserable and tells me, “I am done with this re-installation every season!” but with more swearing. Then he tells me “Have you seen how good AOG is looking lately?” So I did look, watched some of @BrianTee_Admin youtube, and now I am here HAHA.
Back to TSIP. So after seeing 1008 injeckt, and many, many, many, hours of you tube videos on how to code Arduino. I finally did manage to code my first program that takes NMEA swaps talker codes, recalculates the checksum and encapsulates it in TSIP and the EZ150 allows me to start an A+ line! but not tested in the field yet, but I am pretty sure its good to go. There was some other messages being sent periodically from two way calls and responses. But just the repeating data seems to work.
So onto the fully accepted challenge of getting it to work with the Ez 250. I soon find out the data structure is bizarrely different than NMEA, now using ECEF data. Many more calls and responses, configuration handshakes, a larger group of repeating messages and only a couple pieces in ASCii characters.
But experiencing the rejuvenated performance of the Ez 150 with the Emlid M2 external receiver has me convinced that this is old but definitely not obsolete equipment. It can still be useful and relevant for a very, very, long and productive life. Until it finally dies of old age, on its own terms, on the AB line of its choosing. Planned obsolescence, is killing this planet faster than anything, causing more unneeded carbon emissions and pollution than anything, just go for a walk at the Landfill.
The data of position and velocity expressed in numerical terms can only be expressed in the one true language of math, no matter how you dress it up. The F9P looks to have almost all the data required to complete this project. But you have to parse the original data, swap data type, swap endianess (sometimes twice), calculate new checksums, check validity, encode the new structure, and output it all in real time. It sounds bad but it takes quite a few hours to move just a few bytes of data the first time.
As well as new coding I shrunk my original converter.
So as of recently I have been able to move almost one full sentence through. Also working on a snappy looking 3D printed housing so I can move it from machine to machine.
I will be doing my updates in this forum from now on.
more power) setup that goes between the 252 and 250 that just injects the new RTK position information and lets any old 252 deal with the rest of the less important info. I think this would be viable enough to do and might be a good way to move forward on many other and newer receivers. Costs $60 DUE + $250 Ublox vs $3500 unlock?