Barometer fix - PETITION

jessestr · Post by **jessestr** » Tue May 29, 2012 9:34 pm

Hi there

I've made a little petition for the people that really need a working barometer and for the people want to support this..
If you really care about it please vote! It will say how much the barometer fix is needed.

I know it's not easy at all to fix this. I am a programmer too but not in this kind of programming. Please support so Alex & other developers will give it more attention and priority than other features. Keep in mind that I've setup a database especially for this.

Thanks!

You can vote here:
http://devbase.be/multiwii.php

We are not making extra pressure. Just try to change priorities!

Katch · Post by **Katch** » Tue May 29, 2012 10:31 pm

I don't think this kind of action is going to encourage anyone to go and re-write baro functions.

Current baro alt hold works so long as your baro is shielded from sunlight and prop wash / wind etc.

The devs are well aware that it can be improved and it will be in time.

You could better spend your time helping to write the wiki pages and documentation or learning arduino IDE rather than setting up petitions...

not signed

jessestr · Post by **jessestr** » Tue May 29, 2012 10:51 pm

I could But no time. That's why I try other ways. Will add some wiki pages too

Hamburger · Post by **Hamburger** » Tue May 29, 2012 10:56 pm

Katch wrote:I don't think this kind of action is going to encourage anyone to go and re-write baro functions.

+1 (though I would not know how to do it anyway)
and another +1 on writing wiki pages or other doc.

Seeing this same attitude expressed in a slightly different way within a couple of days (last time it was about gyro/level mode) makes me wonder if MultiWii has a communication problem? Or has its status progressed so far as to now attract users with different expectations which might be better suited getting a commercial product with proven support and guarantees?

Maybe one of the board suplliers could step in here and sell h/w plus support (email+phone, first+second level, long term versions, etc.)?

refractions · Post by **refractions** » Tue May 29, 2012 11:07 pm

Sorry - not signing this petition.

The guys developing on this forum are doing it completely free of charge and I personally think its wrong pushing them to develop a specific area. The software isnt 100% working in all areas, but its constantly a work in progress.

I would just like to say thankyou to the developers and the hard work they have done so far. Keep it up

kos · Post by **kos** » Tue May 29, 2012 11:18 pm

the search function is more useful than a petition

viewtopic.php?f=17&t=1564&p=13132&hilit=alt+pid#p13248

not arguing this petition.

Katch · Post by **Katch** » Tue May 29, 2012 11:43 pm

Hamburger wrote:Seeing this same attitude expressed in a slightly different way within a couple of days

I was thinking exactly this. Seems to be an influx of pointless posts in what should be our dedicated development area.

Seems to be coming from new users with little experience and poorly set up hardware that immediately decide their experiences are the result of problems with the code rather than searching to see if people have already had similar issue and resolved them.

We see a similar 'you must fix my problem' and 'you must release what I want' attitude over on XDA the android development forum. It has got so bad that many developers no longer release there, some quit developing completely.

Maybe a little gentle moderation is called for and posts like these can be moved away from current dev / bugs. This will tidy up the clutter but not the attitudes.

jessestr · Post by **jessestr** » Wed May 30, 2012 12:21 am

I'm not just saying 'fix this problem" and I don't have poor gear. I know what I'm doing. And I'm on the XDA forums too. Helpingwith AOKP port for the Wildfire S.

Hamburger · Post by **Hamburger** » Wed May 30, 2012 12:53 am

jessestr wrote:I could But no time. That's why I try other ways. Will add some wiki pages too

Ah, sorry, I missed that.
Well, Jessestr, my time is worth as much as yours and more to me. So if you do not have time to participate then I wonder why anyone should spend time helping solve your problems.

Hamburger · Post by **Hamburger** » Wed May 30, 2012 12:59 am

Katch wrote:Maybe a little gentle moderation is called for and posts like these can be moved away from current dev / bugs. This will tidy up the clutter but not the attitudes.

I'd prefer to continue without moderation if at all possible.
A little discipline on our part would help though. Instead of jumping in and collecting links and info on the person's individual problem we might just point out it being an individual problem, wrong sub-forum and refrain from further adding info to that particular thread.
I know I sound harsh again but to me it is a trade off thing and I'd rather see us spend time on improving the software instead of aiding the user shouting out the loudest (and most inappropriate, maybe) for individual help.

timecop · Post by **timecop** » Wed May 30, 2012 3:42 am

jessestr trolling without understanding as usual.
The same kind of people that bitch about not using a $20 MS5611 pressure sensor because it looks "less noisy".
Properly protect the sensor from wind/sunlight and althold works as good as it ever will without integrating accelerometer into it.

ziojos · Post by **ziojos** » Wed May 30, 2012 7:50 am

dongs wrote:.......
Properly protect the sensor from wind/sunlight and althold works as good as it ever will without integrating accelerometer into it.

Absolutely false, the speed evaluation (Dterm or Vel PTerm) cannot be accurate without zAcc evaluation and fusion with baro data, the dinamic of altitude control may be less good

(eg in wind gusts).

Ciao

timecop · Post by **timecop** » Wed May 30, 2012 7:56 am

No shit, and what do you think I just wrote in the post you just quoted?

ziojos · Post by **ziojos** » Wed May 30, 2012 8:06 am

Sorry

I missunderstand

, I'm Agree with You

fiendie · Post by **fiendie** » Wed May 30, 2012 8:19 am

refractions wrote:The guys developing on this forum are doing it completely free of charge and I personally think its wrong pushing them to develop a specific area. The software isnt 100% working in all areas, but its constantly a work in progress.

That is the sentiment you get with a lot of open source projects. While I agree that starting a petition is a bit weird I don't deem it inappropriate to point out if an advertised feature doesn't work as expected, or not at all for that matter.
I think the problem is that it's difficult for people to get an impression of the current state of a feature. Is it stable/safe to use, Is it experimental/buggy? While it's always fun to work on new features you have to live with people complaining if you don't focus on getting current features to work properly

That said, I don't have any problems with the current ALT-Hold implementation. Put some low density foam over the Baro, tweak the PID settings a bit and it works pretty good.

ziojos · Post by **ziojos** » Wed May 30, 2012 8:31 am

This is a test of GPS Position Hold from take Off to Landing + Altitude Hold (from 1:00 to 1:35) actiated with Throttle stick in middle position
No Roll,Pitch and Yaw during the test the wind was quite strong
SW is MWii 1.9 + JJ-copter Customization (GPS+Altitude)+JJ-Easy Fly

http://www.youtube.com/watch?v=Tv_MPCMMjC0

Altitude use Thrusted ACC (Alt + Vel PID)
GPS is quite the same I used on Aeroquad one year ago (no altitude hold) http://www.youtube.com/watch?v=YPM7lS5PmNM

Ciao

Pyrofer · Post by **Pyrofer** » Wed May 30, 2012 9:05 am

I think Alt hold works quite well.
After watching the sensor when stationary on my desk in the GUI I think the code does an amazing job of holding altitude from that input!

Moaning at open source developers gets you nowhere. That said, if something is wanted by a lot of people there should be a way to communicate that to the people working on it rather than having them guess or just do whatever takes their fancy.

Maybe a multi option Poll setup by the devs listing possible routes for future development would help gauge what people want and stop the need for posts demanding something is fixed or implemented.

Something simple along the lines of ,
What would you like to see next in multiwii development,
1>improved alt hold
2>GPS waypoints
3>moon on a stick
4>exploding lipo issue fixed

etc.

btw, what is the best way to PID tune the existing ALT hold? I get a lot of yoyo action although its quite slow.

Hamburger · Post by **Hamburger** » Wed May 30, 2012 9:28 am

we have a wishlist thread in this forum already.
You want more, paying for it might get you what you want, maybe not.

Pyrofer · Post by **Pyrofer** » Wed May 30, 2012 9:40 am

If there is already a wishlist thread then this one should simply have been a link to that and thread closed.
Was the comment about paying directed at me? If so why?
And, I found the PID tuning tips for alt hold linked to in the project documentation thread.

Hamburger · Post by **Hamburger** » Wed May 30, 2012 10:33 am

Pyrofer wrote:If there is already a wishlist thread then this one should simply have been a link to that and thread closed.

agree. We re-used it to come to terms on how to deal with overly enthusiastic individual demands.

Was the comment about paying directed at me? If so why?

no, not at you. It is about the nature of a wishlist and how far that goes. A wishlist is for things to wish for, not an order sheet of any kind. For orders one usually has to pay someone somehow something. Sorry if that was misleading.

Pyrofer · Post by **Pyrofer** » Wed May 30, 2012 10:46 am

I have seen other open source projects use a "bounty" system where people donate money towards features they want.
Any dev completing the bounty gets the cash saved.
Each bounty has a set of criteria required to be met to "complete" it.

Maybe this is a way forward? It seems to be a great way to encourage development, reward hard work and allow people to shape the way things go with money rather than skill.

Not saying its right for Multiwii, just putting it out there for people to think on.

timecop · Post by **timecop** » Wed May 30, 2012 10:47 am

Any new advanced development should probably occur on appropriate platform.
Demanding professional level features from an obsolete 8bit MCU seems pointless.

Bledi · Post by **Bledi** » Wed May 30, 2012 10:47 am

I don't sign it for me baro is working very good. Just have a look at the tuning procedure

Katch · Post by **Katch** » Wed May 30, 2012 10:55 am

link to wishlist

viewtopic.php?f=7&t=1417

think we are done here.

jessestr · Post by **jessestr** » Wed May 30, 2012 11:35 am

Seemed to be lots of reaction on this and I hope people will still read this.
I have almost no time at all.. Most people don't and I know it's an open source project. I tried to do good in some way but apparently not the right way. After all I should better help developing MultiWii rather than make an online petition.
It was never my idea to lay pressure on anyone. Just change a priority. I've been playing with the barometer PID's and it works great now. Still got to make it wind/sun-free but it already works.

I think some problems should be mentioned on once place where people can always go to check. I know the forum is full of those problems but there is always a different bunch of solutions. Just saying that. Like someone mentioned here before that mulitwii might have a connection problem and I think it's true. There's no real topic with any latest changes (yes on googlecode you can see the latest changes of the devs) or bugs page.

Anyways, I'll close the petition & start helping the devs

Jesse

fiendie · Post by **fiendie** » Wed May 30, 2012 11:41 am

dongs wrote:Any new advanced development should probably occur on appropriate platform.
Demanding professional level features from an obsolete 8bit MCU seems pointless.

Do you *ever* leave out an opportunity to peddle your own stuff?

Hamburger wrote:agree. We re-used it to come to terms on how to deal with overly enthusiastic individual demands.
...
It is about the nature of a wishlist and how far that goes. A wishlist is for things to wish for, not an order sheet of any kind. For orders one usually has to pay someone somehow something. Sorry if that was misleading.

This is usually just a cop-out. We're not talking about support for some outlandish hardware, we're talking about an advertised feature. If the argument holds water then consider it, if the person raising the issue just doesn't know what he or she is talking about educate them. But bringing up money and suggesting that you would divert resources to the highest bidder should not be in the spirit of this project...

timecop · Post by **timecop** » Wed May 30, 2012 11:58 am

fiendie wrote:Do you *ever* leave out an opportunity to peddle your own stuff?

What stuff? That mwc should be on a 32bit platform in 2012? If that isn't obvious to you (as a developer) then maybe you can continue counting processor clocks inside pinchange interrupts to try and make RX input less jittery. FYI, I have never linked to "my stuff" in any of my posts, other than links to source code (and this IS developers forum after all, right?)

jessestr · Post by **jessestr** » Wed May 30, 2012 1:03 pm

Is there any official wiki besides the google wiki of MultiWii? Don't really like the google code wiki's ..

warthox · Post by **warthox** » Wed May 30, 2012 3:27 pm

jessestr wrote:Still got to make it wind/sun-free but it already works.

sorry, but u cant complain about a not working baro when u dont even do the basics :roll:

general:

everybody should keep in mind that the mwc project is opensource and started in the intention to be a super simple multicopterplatform. therefor its amazing what happened the last 2 years.
at beginning maybe not even alex thought that the mwc will grow to a fully equipped multicoptersystem like it is now.
of course you can perfect everything everytime. such a software project will never be finished.
and of course u can also say what u would like to see to be improved or added. and there are 2 options for it:
- put your wishes in the wishlist and hope that there will be somebody who spends his freetime for your wish and works on it for no charge or anything,
or
- do it yourself. its opensource, everybody can work on it.

sometimes i think there are some people which dont see that this is not a commercial product where somebody gains alot of money of it and is paid for working on it, but its a social project where many people work on for no charge to have fun and give others the possibility to also have fun with the result of their work.

to baro:

of course videos like eg the indoor alt hold demo of the naza are impressive. but keep in mind that this is a commercial system where all people which work on it are charged for their work and they do it as fulltimejob.

for me the baro is working good. you will see it in a video which ill upload the next days where the copter hovers in gps position hold and alt hold for more than 10min without any input from me. and this with complete default settings.

timecop · Post by **timecop** » Wed May 30, 2012 3:44 pm

of course videos like eg the indoor alt hold demo of the naza are impressive. but keep in mind that this is a commercial system where all people which work on it are charged for their work and they do it as fulltimejob.

They're only impressive if you don't realize how useless those examples are in real life.

jessestr · Post by **jessestr** » Wed May 30, 2012 3:54 pm

Again sorry to everyone

!

I almost got vacation so I can spend some time on looking to the MultiWii code and the wishlist. I hope I can start developing something because it's really something I want. I am a programmer in gamedev & webdev, never programmed flightcontrollers so will be hard to get in but I'll try and I hope I can deliver some new features. Also the MultiWii is documented well but not the way it should be (my opinion).

Would you like to have a real wiki especially for MultiWii? Else I can setup my server for a wiki

. Server won't go offline, it's in a datacenter so 24/7 online. Good idea?

jessestr · Post by **jessestr** » Wed May 30, 2012 3:55 pm

@Dongs

It's get useful in here:
https://vimeo.com/35432485

I know you can't compare the quality. But it is useful.

timecop · Post by **timecop** » Wed May 30, 2012 4:04 pm

????????????????????????????????????????????
Baro was not used for even a second in that video.

copterrichie · Post by **copterrichie** » Wed May 30, 2012 4:18 pm

Personally, I have viewed a barometer on a multi-rotor copter useless because of the nature of the airflow and pressure created by the spinning props. However because of the lack of a better altitude measurement system, the barometer is all there is. A GPS altitude measurements are just not accurate enough. I am going to order a barometer on its own breakout board and play around with mounting locations to see if that will help.

ronco · Post by **ronco** » Wed May 30, 2012 4:25 pm

well i dont know it for sure but this video dosent feels real ..if you look at 0:20 the copter is tilted to the right but dont moves .. and if the snowflakes are real you can see that there is no wind at all .. maybe it is strorng slow motion..

i think people who dont want or cant develop on multiwii but like to have good results may help the most if they test new features and report their findings as good as thay can..

regards felix

jessestr · Post by **jessestr** » Wed May 30, 2012 4:48 pm

Ronco. I've noticed that too. But yes I think it's very strong slowmo. Those guys fly for National Geographic too so would be weird if it was fake.
It's done very well in my opinion. The gear is way more expensive than MultiWii but I've seen good results with MultiWii too.

To be honest I was a little bit scared to touch the PID rates at the beginning.. never really tried them out because there is not really a 'rule' on how to tune them. Every copter is different and I think that beginners with PID's like me don't try enough to get better results. Since I've started editing them it feels so much better in the air.

Another but more expensive way is to send sonar-like signals to a base station with measures the distance to the multicopter which then calculates the height. But would be too expensive. I don't know much about sensors that can read lengths or heights.. But isn't there any laser like sensor that can measure distances?

Alexinparis · Post by **Alexinparis** » Wed May 30, 2012 6:39 pm

warthox wrote:everybody should keep in mind that the mwc project is opensource and started in the intention to be a super simple multicopterplatform. therefor its amazing what happened the last 2 years.
at beginning maybe not even alex thought that the mwc will grow to a fully equipped multicoptersystem like it is now.
of course you can perfect everything everytime. such a software project will never be finished.
and of course u can also say what u would like to see to be improved or added. and there are 2 options for it:
- put your wishes in the wishlist and hope that there will be somebody who spends his freetime for your wish and works on it for no charge or anything,
or
- do it yourself. its opensource, everybody can work on it.

sometimes i think there are some people which dont see that this is not a commercial product where somebody gains alot of money of it and is paid for working on it, but its a social project where many people work on for no charge to have fun and give others the possibility to also have fun with the result of their work.

well said

It's exactly my view.

Hamburger · Post by **Hamburger** » Wed May 30, 2012 8:02 pm

fiendie wrote:
Hamburger wrote:agree. We re-used it to come to terms on how to deal with overly enthusiastic individual demands.
...
It is about the nature of a wishlist and how far that goes. A wishlist is for things to wish for, not an order sheet of any kind. For orders one usually has to pay someone somehow something. Sorry if that was misleading.

This is usually just a cop-out. We're not talking about support for some outlandish hardware, we're talking about an advertised feature. If the argument holds water then consider it, if the person raising the issue just doesn't know what he or she is talking about educate them. But bringing up money and suggesting that you would divert resources to the highest bidder should not be in the spirit of this project...

For the record
I was tryinv to explain the line between a wishlist and an order form.
Iam not advocating payed development for mwii. And I do not want decisions for/against features based on highest bids.

jessestr · Post by **jessestr** » Wed May 30, 2012 8:11 pm

I agree now..

Crashpilot1000 · Post by **Crashpilot1000** » Thu May 31, 2012 5:55 pm

I did something!
http://fpv-community.de/showthread.php? ... post150774
I will write up a full ist of findings and changes i made later. The Baro PID controller is not altered and the filtering is also unchanged.
You can try this code. Or just look at my changes in MAIN (variables, changed locking new hight), IMU/ getEstimatedAltitude(), and a byte boolean "newbaroalt" in "sensors" set to 1 every time a new "BaroAlt" was gathered.

So long

Crashpilot

http://www.youtube.com/watch?v=tYam_n6IBaw

jessestr · Post by **jessestr** » Thu May 31, 2012 6:00 pm

Nice.
Did you do any testflights?

Jesse

Crashpilot1000 · Post by **Crashpilot1000** » Thu May 31, 2012 6:22 pm

Yes, and it works better now. But with my BMP085 and the still missing ACC Z Vector implementation - a naza style alt hold is not possible.
Due to the fact that i can not program an own PID controller or kalmanfilter the changes a just minor but i think effective. So its very close to the original code. The possibility of getting a wild, amok copter with this is very low.

So long

Crashpilot

jessestr · Post by **jessestr** » Thu May 31, 2012 6:47 pm

Will test it out when I got my new Y6 build. Thanks for helping!

realpastaman · Post by **realpastaman** » Mon Jun 04, 2012 12:37 pm

Just doing some work with the MPX2053 that looks less noisy to me and is already temperature compensated.

http://www.multiwii.com/forum/viewtopic.php?f=16&t=1769

Just wondered if anyone has used it ?

Mystic3D · Post by **Mystic3D** » Mon Apr 01, 2013 5:37 pm

My findings are if you have BMP085 you are going to get marginal results as the hardware outputs such an erratic signal in MWGui. So hardware limitation..

The MS5611 yields very good results as far as outputting a consistent signal.

felixrising · Post by **felixrising** » Mon Apr 01, 2013 11:27 pm

Mystic3D. This is a very old thread, the last post was more than 9 months ago and the recent 2.2 changes included some improved altitude hold code. This thread should be marked closed.

Mystic3D · Post by **Mystic3D** » Tue Apr 02, 2013 1:52 pm

felixrising wrote:Mystic3D. This is a very old thread, the last post was more than 9 months ago and the recent 2.2 changes included some improved altitude hold code. This thread should be marked closed.

Agreed, my results are with 2.2.
Point is also need good hardware.