NavX Calibration Advice

[dithier]

Our team is starting to work on having our robot align to the target for shooting in the autonomous period and has decided to use the NavX-MXP. We've read that calibration is very important in order to get absolute heading, but I'm a little confused about calibrating the magnetometer based on our robot setup.

We can't use the default yaw axis orientation because of the orientation of our RoboRio and read on the navX website that we can use the OmniMount feature to adjust for this when needing yaw axis readings: http://www.pdocs.kauailabs.com/navx-...ion/omnimount/. Ours is mounted exactly like the picture on that site shows.

However, the magnetometer calibration page doesn't mention how to adjust calibration if the board is not in the expected orientation on the robot. Do we still follow the calibration method outlined on the website and just make sure that we are using it's "new" x,y,z axis when taking data points or is this not necessary? Lastly, should we be doing this calibration while the board is in the robot and the robot is on?

[cad321]

Originally Posted by jfish872

One word of advice. DON"T USE THE NAVX. The software terrible. The libraries don't work. It is nearly impossible to flash the navx, let alone calibrate it.

To put it simply, don't use the navx.

Why do you say this? Do you have any specific examples as to why you dont recomend the navX? I've heard nothing but positive reviews from all teams I've talked to who use it.

We ourselves had some troubles with it this season but all of it was due to user error and a lack of programming knowledge overall. We will by trying to get ours up and running over the off season so we can use it. In my view, it is a major resource to teams that can help them program more advanced auto routines with greater accuracy.

[WSiggs]

Originally Posted by cad321

Why do you say this? Do you have any specific examples as to why you dont recomend the navX? I've heard nothing but positive reviews from all teams I've talked to who use it.

I agree with this. Everything I have heard about it from other teams is how helpful and useful it is. I have not heard of anything so negative about it that it would make it not worth it to invest in, and as such my team was planning on looking into it for the off-season and next year.

[Caboose]

Originally Posted by jfish872

Our programming team has had nothing but trouble with this product, because of the issues that I originally said. I have altered my post to reflect less biased views of this product.

Did your programming team ask for help when they had issues? I know Scott monitors ChiefDelphi for posts on the navX and helps when asked. It seems to me that your programmers either didn't bother trying to use the navX and just said it didn't work or they tried but did not ask for help when they couldn't make it work and ran into issues.

[jfish872]

Originally Posted by Caboose

Did your programming team ask for help when they had issues? I know Scott monitors ChiefDelphi for posts on the navX and helps when asked. It seems to me that your programmers either didn't bother trying to use the navX and just said it didn't work or they tried but did not ask for help when they couldn't make it work and ran into issues.

I don't know, I don't know if they knew about this site then.

Please just let me delete my original post in peace.

I'm sorry for posting it in the first place, I know that I was unsessissarally harsh about the product and I would like to apologize to everyone who got offended by it.

[jfish872]

Originally Posted by WSiggs

I agree with this. Everything I have heard about it from other teams is how helpful and useful it is. I have not heard of anything so negative about it that it would make it not worth it to invest in, and as such my team was planning on looking into it for the off-season and next year.

I'm sorry for posting such a negative review. All I heard about it was hearsay form our team which was probably colored by the stress that we were going though during the build season. If we were able to get the device to work I think I would have recommended it.

I shouldn't have said anything at all, and I would like to apologize for what I did say.

[frcguy]

Originally Posted by Caboose

Did your programming team ask for help when they had issues? I know Scott monitors ChiefDelphi for posts on the navX and helps when asked. It seems to me that your programmers either didn't bother trying to use the navX and just said it didn't work or they tried but did not ask for help when they couldn't make it work and ran into issues.

Before this gets misconstrued, we did have our issues with the NavX, and posted about them here: http://www.chiefdelphi.com/forums/sh...32#post1536032. We hope to get it working during the offseason.

[slibert]

Originally Posted by jfish872

I'm sorry for posting such a negative review. All I heard about it was hearsay form our team which was probably colored by the stress that we were going though during the build season. If we were able to get the device to work I think I would have recommended it.

I shouldn't have said anything at all, and I would like to apologize for what I did say.

Hi, this is Scott, founder of Kauai Labs, we very much want to work with you to resolve whatever problems you are having while using the navX-MXP - which was created specifically for FRC teams to use and learn about.

We've had a lot of success with hundreds of FRC teams, and navX-MXP was on Einstein at FRC 2016 World Championships, aboard both the champion alliance team captain robot (team 330) and we've also heard it was on runner-up alliance team 2056. While we don't take credit for their success, this does show that some of the best teams in the world are using navX-MXP.

All that said, there can always be problems while integrating something as powerful as the navX-MXP, so we're here to help, and you can contact us directly at support@kauailabs.com. Feel free to ask for help if you need it - our goal is that every team is successful with navX-MXP, and we very much want that to include your team.

[slibert]

Originally Posted by dithier

Our team is starting to work on having our robot align to the target for shooting in the autonomous period and has decided to use the NavX-MXP. We've read that calibration is very important in order to get absolute heading, but I'm a little confused about calibrating the magnetometer based on our robot setup.

We can't use the default yaw axis orientation because of the orientation of our RoboRio and read on the navX website that we can use the OmniMount feature to adjust for this when needing yaw axis readings: http://www.pdocs.kauailabs.com/navx-...ion/omnimount/. Ours is mounted exactly like the picture on that site shows.

However, the magnetometer calibration page doesn't mention how to adjust calibration if the board is not in the expected orientation on the robot. Do we still follow the calibration method outlined on the website and just make sure that we are using it's "new" x,y,z axis when taking data points or is this not necessary? Lastly, should we be doing this calibration while the board is in the robot and the robot is on?

After you've performed the "Omnimount" configuration, and if you want to calibrate the magnetometer, do so using the Magnetometer Calibration tool, using the "new" x,y,z axis orientation - and this should be done while the board is on the robot, and the motors on the robot are *not being energized*. Take care as you perform this calibration to ensure the alignment is as accurate as possible, as this is key to achieving the best results.

Do note that many teams do not find they need the absolute heading data enabled by the magnetometer; instead, most teams use the yaw angle which is based upon fusion of the gyro and accelerometer data, and set the yaw to zero when the robot is in a known orientation relative to the field. So based upon your needs, magnetometer calibration may not be necessary.

[dithier]

Hi Scott,

Thanks so much for the advice-- that clears up my understanding. We're so excited to start using this for autonomous!!