Poll NEW! QuadCopter platform - QuadPowered

[S11D336B]

Originally Posted by bob.titus

Yes, you are absolutely right, this is the problem and it has not been sorted yet. I spent no less than 50 hours a week for 2 weeks to try and sort this out, but the coding of the Arduino is very new to me, and the forum could not come with solutions, so I stayed pretty stuck.

That really won't be an issue with this board. I have an algorithm that will automatically sequence which looks at the order the PWM is sent and sample in that order. Also, I can always just use one core per channel to get real time PWM . Will be cool.

[bob.titus]

Ok, great, let me know what's going on, I'm staying tuned :-)

[RCvertt]

Originally Posted by S11D336B

...Since I am developing something, I can pretty much add whatever features I want to it. What features would you guys like to see on this product?...

I'm a fan of accelerometers which it sounds like you may have plans for in the future or at least something similar.

Originally Posted by S11D336B

The circuit board, components and electrical design will be available (open source) But the software will not be at this point...

Sounds very reasonable to me.

May I ask what language the source is likely to be in if released?

[S11D336B]

Originally Posted by RCvertt

I'm a fan of accelerometers which it sounds like you may have plans for in the future or at least something similar.

Will this be open source?

The circuit board, components and electrical design will be available (open source) But the software will not be at this point. I may change that later. The reason for this is I want to avoid the problems that many are experiencing currently with other open source based projects. Having a project open source increases the complexity and chance for failure significantly. I know 100% that if I write the code myself and precompile it that it will work when it gets to the user. Essentially I am going for a plug and go type deployment. I say I may change that later because you might be able to convince me to release the code openly and sell the kits rather than selling the code and giving away kits . Remember this is supposed to be a revenue generating deal for me. I'm not looking for much just enough to fund some of my other projects.

I will say that the platform I am working with is incredible. Suffice it to say it has 16 parallel processing cores capable of running at 80Mhz each. It has a total of 512Kb onboard memory.

End the end we are talking about an upgradable board and software capable of driving a quadcopter efficiently and flexibly for around $200. I think it's a steal compared to some of the other offers out there in the $300+ range.

[S11D336B]

Originally Posted by RCvertt

May I ask what language the source is likely to be in if released?

SPIN is the language I will be writing it in. SPIN is kind of like C and was designed to be easy to learn and easy to read. Should be a snap for you guys who are working on other C based micro controllers.

[S11D336B]

Originally Posted by RCvertt

I'm a fan of accelerometers which it sounds like you may have plans for in the future or at least something similar.

I think accelerometers are very nifty little devices. But as I said earlier it's important to be careful in their use. Subtle motions such as decent, ascent or rapid changes in attitude can cause problems. Kalman filters can filter out a lot of that in the short term, but long term motions such moderate speed decents from an altitude can induce incorrectable biases into the system that ultimately result in instability. I will place headers on the board for one incase I find it absolutely necessary for implementation of a feature, but I'd prefer not to use one for the above reasons and to avoid the additional $30 one would cost for each board. I think you'll find that a decent gyro can have reasonable accuracy for up to 2 minutes. Measuring drift via GPS should enable future software releases to have position hold with decent accuracy without the use of accelerometers.

[bob.titus]

Originally Posted by S11D336B

I think accelerometers are very nifty little devices. But as I said earlier it's important to be careful in their use. Subtle motions such as decent, ascent or rapid changes in attitude can cause problems. Kalman filters can filter out a lot of that in the short term, but long term motions such moderate speed decents from an altitude can induce incorrectable biases into the system that ultimately result in instability. I will place headers on the board for one incase I find it absolutely necessary for implementation of a feature, but I'd prefer not to use one for the above reasons and to avoid the additional $30 one would cost for each board. I think you'll find that a decent gyro can have reasonable accuracy for up to 2 minutes. Measuring drift via GPS should enable future software releases to have position hold with decent accuracy without the use of accelerometers.

All this sounds very reasonable asRCvertt said earlier
But about GPS assisted position hold, don't GPS need to be in constant motion to be able to function well?

[lazer155]

No gps can funcition well while it is stationary as long as it has contact with at least 3 satillites.

[S11D336B]

Originally Posted by lazer155

No gps can funcition well while it is stationary as long as it has contact with at least 3 satillites.

For position hold, GPS can give position and altitude data when it is stationary. You do need some motion to get the vector for track and speed, but thats the beauty of the thing. Generate a vector with a certain track and speed until you get back to the hold location and then try to keep no vector until outside of an acceptable error range. But this kind of functionality is a little ways out. I'm really trying to focus on getting the hardware and basic software released at this point. Then I can focus on implementing new features like GPS position hold and altitude hold. As long as I provide the headers in the board for the necessary hardware (which I will) it's as simple as reflashing the software and plugging in/soldering the correct hardware to the board. The GPS board I was looking at using has 56 GPS channels and an omni directional antenna so we don't loose sattelites if we bank too much. I'm going to try and make the interface universally compatible so it will work with different GPS boards. I'll need to do some reasearch on this, but this functionality will come later.

[S11D336B]

I started redisigning the board yesterday so I can include all the features I promised as well as updatable software and upgradable components. I think I'm just going to group upgrades into modules which I'll label with obvious names

Barometer Module (For altitude hold)
GPS Module (For Position hold)
Others to come with your input

Say how many of you have 3 position switches on your transmitters? I was thinking of doing this: up is standard RC mode, middle is altitude hold, and down is GPS position hold with altitude hold.

Oh yeah, attached is a teaser from the new board design

[bob.titus]

Originally Posted by S11D336B

For position hold, GPS can give position and altitude data when it is stationary. You do need some motion to get the vector for track and speed, but thats the beauty of the thing. Generate a vector with a certain track and speed until you get back to the hold location and then try to keep no vector until outside of an acceptable error range. But this kind of functionality is a little ways out. I'm really trying to focus on getting the hardware and basic software released at this point. Then I can focus on implementing new features like GPS position hold and altitude hold. As long as I provide the headers in the board for the necessary hardware (which I will) it's as simple as reflashing the software and plugging in/soldering the correct hardware to the board. The GPS board I was looking at using has 56 GPS channels and an omni directional antenna so we don't loose sattelites if we bank too much. I'm going to try and make the interface universally compatible so it will work with different GPS boards. I'll need to do some reasearch on this, but this functionality will come later.

would any type of GPS work? Like for example this one (that I happen to own) :
http://www.sparkfun.com/commerce/pro...roducts_id=465

[bob.titus]

Originally Posted by S11D336B

I started redisigning the board yesterday so I can include all the features I promised as well as updatable software and upgradable components. I think I'm just going to group upgrades into modules which I'll label with obvious names

Barometer Module (For altitude hold)
GPS Module (For Position hold)
Others to come with your input

Say how many of you have 3 position switches on your transmitters? I was thinking of doing this: up is standard RC mode, middle is altitude hold, and down is GPS position hold.

Oh yeah, attached is a teaser from the new board design

I do, 2 of them

[S11D336B]

Originally Posted by bob.titus

would any type of GPS work? Like for example this one (that I happen to own) :
http://www.sparkfun.com/commerce/pro...roducts_id=465

So in order for it to be universally compatible 2 things need to happen. The modules need to use the same communications protocol and have the same command set. e.g. the same input always retrieves me the same data from each module. I'll have to do more research on it, worst case only 50% of modules out there work with it. But that's still a huge selection.

[bjornez]

I'm interested!

would it be possible to ad a slow powerdown incase of radiosignal loss?
I'm looking for a good FPV quadro and it's not very funny when a moment of singal loss makes it fall from the sky...

[S11D336B]

Originally Posted by bjornez

I'm interested!

would it be possible to ad a slow powerdown incase of radiosignal loss?
I'm looking for a good FPV quadro and it's not very funny when a moment of singal loss makes it fall from the sky...

Yes! This is definitely possible! Great idea and simple to implement. I'll add it to the list of features on the first page. Also new board picture. I'm almost done with it. Just need to add a few more components and tripple check everything.