hi

im just adding gps into the mix to see if my cuad is capable of decent position hold,

i have him sat in the window, with 7 sats and 3d fix, hdop goes between 150 and 250.

looking on the map, it shows the drone as moving a lot, between what i would imagine is at least a 20m radius from the house, and it shows 45 meters to waypoint, which i guess is the home position,


i have the feeling something is up and that its not going to work , any ideas?

Views: 153

Reply to This

Replies to This Discussion

Why do you think your GPS will work inside?  I'm curious why this is such a recurring theme here.

Have you never seen mention of a "clear view of the sky"?  I'd love to figure out if people just aren't reading the docs, don't understand how GPS works, or if there's a translation issue somewhere.

If you want good GPS performance you need a clear view of the sky.  That means 180 degrees of sky visible in all directions.  That's a top mounted GPS out in a field, which is how most people should be operating.

Anything that reduces this will reduce your GPS performance and accuracy.  Even being near buildings will block some portion of sky and weaken your performance.

Being inside, even near a window, only gives you a tiny, tiny fraction of a clear view of the sky, so why would you think the GPS would work well?

Further, the reason why hdop is jumping all over the place is ghosting signals from reflections. The latest 2.71 code uses the velocity info from the GPS module, whihc is innacurate due to reflections from nearby walls, buildings etc.

In general GPS is less accurate inside or near manmade structures. Some of that discussion is in this thread as well http://diydrones.ning.com/forum/topics/new-3dr-ublox-gps-causes-int... 

I feel like i shoudl defend myself against accusations of being a complete tool ,

I do know how gps works, Perhaps i was mistaken in thinking that the GPS on the cuad would report the same accuracy and other gps units, I wouldnt expect the same, but a roughly similar accuracy no?

in U center the ublox 5 reports 6 meters, in the same spot and the cuad which reports 44m, that cant be right, , the GARMIN I have reports 7m,

i understand that in the window it can only see a few satalites, but it reports 7, and it gets 4 from inside the house, although i suppose not very well,

I havent had a chance to fly it yet, but i have low hopes for its success,

Can i easily implement the ublox 5?

Yes, ublox 5 in general is compatible.  The ardupilot firmware has several GPS "drivers" it loads via an auto detection routine at startup boot of the APM. It might require a specific firmware be on your ublox5. The driver simply reads the native GPS data format and converts it to the variable locations used with the APM functions. Think of it as an adapter.

in U center the ublox 5 reports 6 meters, in the same spot and the cuad which reports 44m, that cant be right, , the GARMIN I have reports 7m,

Point being, there's more to it than just the GPS position and that's what I was attempting to explain and you largely ignored it. There is also velocity calcs that come into play. The point is, the final position the APM reports is not just based on the raw GPS position as you seem to imply. Thus switching the GPS module may not change anything at all. See this thread http://diydrones.ning.com/forum/topics/lea-6-gps-causing-pitch-osci... The thread I linked is about the latest and greatest ublox GPS also not seemlingly providing a steady position as you are expecting so again, your assumption the GPS module is faulty is not valid. Because the APM uses more than one sensor to determine all aspects of it's physical position, GPS is not an absolute. Likely, the commercial GPS systems to you are comparing to have significant software filtering functions to make you think the module always reports the same postion. The Raw data tells a very different story.

This thread is about a user making the ublox 5 work with APM 2.0 and arducopter 2.7.1. He does get it to work so it is compatible.

http://diydrones.com/forum/topics/apm2-0-with-ublox-gps-not-detected

 

Nobody is trying to call you a tool. Believe it or not, I'm trying to educate you and anyone else who has the same question when they read this thread. That's what forums are all about. Somtimes the answer isn't directed at you in a personal way, it's about giving info so others who may not know have a more complete answer. Please try to have a little bit thicker skin, look at the forum objectivly and realise nobody here tries to insult.

Also, again, reading this thread, I could see how someone who hasn't been around could take things as an insult. If you've been around here for a while, you'll see the occasional disagreement or even what appears to be someone talking down to someone. Please just take this as normal, it's what happens in the disfunctional social system that is a forum. It's a place where we have a discussion, sometime based on fact, but a lot of opinion comes intop play when we don't have the entire scenario in front of us, we have no choice but to make assumptions and sometimes it's wrong. When people post a question that doesn't give us every fact (location, firmware, complete hardware specs including radio system, outside conditions, and a whole host of other factors) it's hard to not make an assumption and 90% of the time it's wrong. But what happens is the original poster clarrifies the exact situation, asks a more pointed question and then we give a far better answer. It can be insulting at times, but sometimes, that's just the technical side and you must get past it and realize we all are learning. I learn just from answering questions. I have 3 APMs, a quad, a hexa, and 2 planes. I still haven't setup every feature even if I have the hardware such as optical flow. It's all in experimentation phase. Thus, I'm trying to help set your expectations that you need some tuning and testing on your aircraft with this open source project and hardware before absolute accuracy you are expecting is going to work.

Yes, the hardware and software is capable and there are tons of success stories. No it doesn't yet work out of the box on random hardware. That's the problem, there are a couple million combinations of hardware you can build a machine from, no garanteed settings to work for any of them exist. There are guidlines and how to's, but that's not the same as a commercial, take it out of the box and fly it solution.

Everything I said was dead serious, not to be taken as an insult in any way.

I really want to know why people repeatedly post "GPS not working, Help!" threads, and after we spend several posts trying to troubleshoot their problems they reply with "Hey, I went outside and it works perfectly, I don't understand why it won't work in my basement though?"

This has happened many times here.  Commercial steel and concrete buildings, tin roofs, basements, metal buildings, etc., etc..

I'm just curious where people get the expectation that GPS works in highly shielded areas and how they miss the "clear view of the sky" part.

RSS

Social Networking

Contests

Season Two of the Trust Time Trial (T3) Contest has now begun. The fourth round is an accuracy round for multicopters, which requires contestants to fly a cube. The deadline is April 14th.

A list of all T3 contests is here

Advertisement

© 2013   Created by Chris Anderson.   Powered by

Badges  |  Report an Issue  |  Terms of Service