Maybe my own wording was harsh, but the fact that you are continually getting an “unhealthy” message means that it’s not working as well as you likely think it is, and you may be on the ragged edge of failure despite your anecdotal (and empirical) evidence.
I have dealt with countless users trying to make their use cases unique, but it’s all the same hardware, and it’s all deterministic behavior. Something is objectively wrong with your configuration, and we should work to sort it out.
I always recommend to NEVER use u-Center with these modules except to update the firmware. Doing so usually causes more harm than good, and the ArduPilot driver does an EXCELLENT job of curating settings appropriately. Unfortunately, it can’t account for all possible initial states, and the one your modules is in at present is not ideal.
I do see that you have GPS_INJECT_TO set to 0, when it defaults to 127. I recommend leaving it at 127, as no (ArduPilot moving baseline) documentation ever directs otherwise, as counterintuitive as that may be.
This is indeed how it works. I have set up over half a dozen of these configurations in person without issue, and the only thing I use u-Center for is updating firmware as needed. And if you search my username and “F9P” on this forum, it will show up an embarrassing number of times as I tried and usually succeeded to get user vehicles working.
I guess I do use u-Center for other things, but that is beyond the scope of this conversation (and bolsters my opinion that most should stay away from it).
EDIT: And now that I think of it, I had half a dozen of these things on the bench all at once for a consulting job, so that’s a very conservative estimate!
At the time I had a partially faulty GPS card. It would not communicate correctly and it was getting very hot. I had to replace it and my problems went away.
Yuri, no offense taken from your comments. I appreciate and respect your direct nature. I think we are just viewing the (empirical) evidence from different perspectives. The grass getting cut without having a heat stoke is mine.
That said, I hear and heed your warning that I “may be on the ragged edge of failure”.
I changed the GPS_INJECT_TO 127 as suggested. I previously had it set to 127, and only changed it in the last few days trying to pinpoint my problem.
Changing it didn’t seem to have an effect, at least in the HUD GPS status. I didn’t take it that you suspected it would though.
I then turned my attention to getting the requested .bin file.
When I attempted to download the .bin file I was getting a message to disarm the vehicle. If I ignored this message I got error when trying to load the logs. When I tried to disarm the vehicle, it would not disarm. I started working on sorting that out so I could get the .bin file. I changed the Arming_REQUIRE from 0 to 1.
When I returned to the DATA screen to my surprise (Likely not yours) the GPS signals were both showing RTK Fixed. I set the Arming_REQUIRE back to 0 just
to confirm the results, and again I lost the RTK Fixed status on GPS2.
Seems I now have a working system with setting GPS_AUTO_CONFIG = 1, and no more pesky "Unhealthy GPS Signal” on the HUD.
I greatly appreciate the help from you and Steve. Without both of your countless post and guidance to myself and others, I would have never had a chance at getting it working to the degree that it is.
I was just rereading about your fire and thinking to myself “This here is the best argument for not having rover Return To Base if there is a malfunction”
I noticed in the pic of your layout on the mower, you had a RFD 900x-us long range radio! I recently acquired a pair tx/rec radios. I was wondering how they worked for you, also if they are plug and play or if there is a lot of setup involved?
Hey Rusty, sorry for the late reply. Being winter I haven’t been very active in the forum and with my mower.
I have had very good luck with the radios. I have several missions that take the mower 1/2 mile or more from my base station and rarely do I loose communication, and only when there is an obstruction in the way.
Adding LIDAR for crash avoidance and remote start in the coming weeks.
Caleb, I’m very busy also . Thanks for your reply. I bought the radios a year ago as a replacement for a couple of the sik radios. They actually work really well. even 1/4 mile and down a hill ,out of line of sight. I guess I will use them till they quit working. I will probably use them on a base station some day.Thanks again.
Rustey
Hello everybody, I have recently fell into autonomous mower rabbit hole and ended up here thanks to Kenny trussell and Yuri’s Youtube videos. I’m looking for advice on steering. I see that most people are using servos, and some people have used linear actuators. Servos seem to be a lot more responsive and designed for this sort of thing. However what I’m unsure about is how servos will handle a mower with a return to neutral spring on the hydros. Will servos overheat if they have to hold a constant spring tension indefinitely? I tested the spring force with a luggage scale and its between 4 and 5lbs. I don’t want to remove the springs because i want this mower to be able to transition between autonomous mode and human operator easily and the springs are not easy to install and reinstall. I know linear actuators wont have this problem but they seem to have incompatible duty cycles and are very slow. Another idea was to use a planetary-BLDC servo like you’d find on a robot arm considering those have to hold up a heavy arm indefinitely. Reading through this forum it looks like you guys know what you are doing and are super helpful. I’d be super appreciative for some expert advice. This is for an old walker model T.
Here is a link to a writeup I put together on a way I came up with to be able to switch between auto control and having a manual drive ability with the original drive handles. The hydro control valve damper springs generally have to be removed to use typical servos. My approach was to remove the linkage between the manual drive handles, install a new linkage connecting a control servo to each hydro valve. The servos were then driven by the Ardupilot system. I then installed hall sensors on the handles to generate a voltage depending on handle position and voltage to RC PWM signal converter that could be switched in to control the servos. It’s basically a drive by wire system, so even when using the handles, the servos are moving the hydro valves.
