New board: Sparky

[peabody124]

[Edit: See this page http://buildandcrash.blogspot.com/2013/05/sparky-testing-and-building-no-crashing.html for a more thorough writeup]

So while I was waiting for the freedom parts and pcbs (see previous post, flying now) I was goofing around with a new board(s) design.  I've only put about 6 hours into this so it might be a bit rough still.

Here were the goals:

1) serve in three roles: standard flight controller, gimbal controller (ideally talking to the main FC), and brushless gimbal controller.  Partly I just wanted a really cheap but well performing flight controller for my many quads.

2) Single sided for components. Makes it nice and dead easy for hand assembly and can be sticky taped to frame.

3) two layered - keep it cheap for small quantities

4) CAN connectivity

5) Make the brushless driver modular

6) Use 9150 to provide mag information when used as a compass

7) Since I want it to drive a dual axis BGC it's got 12 channels output :)  Great for octo + gimbal in a standard configuration.

So you can see the brushless drivers (Sparky BGC) can piggy-back on the front or back of Sparky or even be cabled off separately.  It will use CAN to talk to the main flight controller (e.g. another Sparky, Freedom or any other board with a CAN transceiver) to get control signals.  Another Sparky board can also be populated with an MPU3300 and put on the CAN bus to provide an extremely low drift gimbal under strong accelerations.  If that works well it will be easy to extend Sparky and add that to the main board.

Things I'm conflicted about:

1) Size.  It could be made smaller, especially by losing the mounting holes (or two of them).  However I'm thinking to use as a gimbal controller it will be nice to really bolt it onto the gimbal well.  Kenn has also written some vibration analysis software that will be able to run on this to help make sure the gimbal is vibration is clean so you actually wouldn't want to use sticky tape in this case.  Also it's about 300 mil more narrow than the standard footprint - I'm almost wondering if that is worth changing the mounting pattern for.  Possibly I'll just extend the board one way to the standard size with nothing there and it can be cut off ?

2) MS5611.  It's tempting to just add this for indoor altitude hold.  This ties into size - with the current size I can probably squeeze it in and if desired this board can do navigation.  *shrugs*

Right now for single quantities the BOM (with MPU9150) is ~30$ so it's going pretty well as a cheap capable FC.  This is just something I'm goofing around with and i'll probably get 10 cheap PCBs made if anyone else were interested in populating their own.

Schematic: https://dl.dropbox.com/u/6645063/Sparky.pdf - I still need to closely read the F3 datasheet so please point out if I made any goofs.

[ala42]

I am missing the 1k5 pullup resistor on the USB D+ line.

[peabody124]

Ahh thank you!

[ala42]

I also miss it on the quanton board, btw.

[Kenn Sebesta]

Why can't this be done by the STM's internal pull-up? Is this missing on all the boards, ever since OP AHRS?

[peabody124]

We found it's not needed for the F4.  It's removed on Freedom too.

[ala42]

There has to be a 1k5 pullup according to the USB specs. It might be included in the F4 chip. In the F3 datasheet it is explicit listed as required, see the USB characteristics of the DS9118 datasheet on page 98:

>To be compliant with the USB 2.0 full-speed electrical specification, the USBDP (D+) pin should be pulled

up with a 1.5 kOhm resistor to a 3.0-to-3.6 V voltage range.

This looks different in the F4 datasheet on page 113, where RUP is explicit specified for USB use as typ. 1k8, so it seems to be included.

[peabody124]

Ahh thanks for looking that up.  It's nice to have it authoritatively   

[peabody124]

Ok.  It's submitted :D

In the end I said screw it and traded a mounting hole for an MS5611.  It's going to be nice to have a dirty cheap board that can do altitude hold and RTH while flying as well as CC3D.  Not to mention single sided so easy as hell for hand assembly :)

[AlPackin]

very nice James!

[peabody124]

Also leaving htis as a note for myself.  It was pointed out to me that for a BGC we probably will want potentiometer inputs to allow compensating for the changing forward model based on the gimbal state.  Also electronic stops to prevent damage.

[ala42]

At the MPU-9150 you must not connect AUX_CL/AUX_DA (pin6+7) to GND, or you can not talk to the compass. When using just a MPU-6050 this bug does not matter.

Leaving the CPU pin BOOT1 open is a bad idea, as you need it to enter boot mode. Can not tell you for sure right now if it should be low or high, as STM manual names the flag nBOOT1 in the boot mode table for extra confusion, so better test that on a Discovery F3 board. You want to enter system memory boot in case BOOT0 is high to enable the build in boot loader.

[peabody124]

Ah, thanks for those notes.  The CL/DA I should be able to cut on the prototypes and same with the BOOT1 line (crap meant to go back and fix that).  I added it to my change notes for the next revision.

[Kenn Sebesta]

Have you thought about putting some blank pads on the back for the RFMB module and antenna? It could be an easy upgrade for people who want the smal size of Sparky and would like to go all the way in stripping out weight by not having an external receiver.

[peabody124]

It wouldn't work with a 2 layer board.  The pads are free on the MCU but you couldn't route it.

[msev]

Sweet looking board peabody! Whats your source for MPU9150? Digikey isn't my cup of tea :), since it has outrageous shipping rates to my country..

Regarding Kenn's suggestion, are you guys going to add direct support for rfm22b's in a sense that some pads from FC (flyingf3 or others) would connect to it?

[Y Jung]

I have a connection to the INVENSENSE distribution channel, and received some samples of MPU-9150 as well.

 

I can send you if you need. 

[msev]

That would be sweet Y Jung :).

[ala42]

> Digikey isn't my cup of tea :)

As Digikey does not sell Invensense sensors, that is no problem. The prices at Invensense and cdiweb are quite high, even without the high priced shipment.

[peabody124]

Got some test flights out of this:

https://vimeo.com/63619336

and it's working pretty well.  I just ordered a few boards with a revision that should expose an extra serial port to allow GPS + Telemetry (which I hadn't originally intended when I made this).  Still single sided and nice and small.

More complete writeup here http://buildandcrash.blogspot.com/2013/04/triblivion-and-sparky.html

[AlPackin]

wow, very nice!  

[msev]

+1, really cool!

Dne petek, 19. april 2013 19:40:04 UTC+2 je oseba AlPackin napisala:

[peabody124]

Small update.  The brushless gimbal add-on works.  Now I need to get a gimbal :)

http://buildandcrash.blogspot.com/2013/05/brushless-gimbal-driver-add-on-for.html

I need to decide if I want to drop the enable lines and then make a slightly bigger version that can drive three motors for a full gimbal or keep this the same.

[Reddog]

Looks very interesting.

One thing I saw you mention was the antenna tracker on the ground. I too looked into using a brushless motor gimbal for antenna tracking and they don't have enough torque for it.

If you did decide to build an antenna tracker using Tau Labs boards I would really love to be involved if you could handle explaining some possibly mundane things. I am immensely interested in getting involved with coding on Tau Labs hardware.

[Menno]

Why would you want to build an antenna tracker with a brushless motor instead of a servo? I don't see the advantage.

[peabody124]

Eh.  Because I already have that board and it sounds entertaining :)

[Menno]

Ah entertainment and education are 2 of the best motivators.

But can you explain me.

- Do you want to rotate the antenna on the ground with the help a brushless motor (this is what I read in Reddog response, but I might be reading this wrong)

- Do you want to use the data that is being transferred from the brushless gimbal driver to the groundstation.

Op dinsdag 7 mei 2013 15:01:41 UTC+2 schreef peabody124 het volgende:

[José Eduardo S. C. Xavier]

It lookgs great :) I think a 3axis gimbal is a good idea:)

2013/5/7 Menno <menno...@gmail.com>

--
You received this message because you are subscribed to the Google Groups "phoenixpilot" group.
To unsubscribe from this group and stop receiving emails from it, send an email to phoenixpilot...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.
 
 

[peabody124]

@Reddog: regarding torque I assume it would be the same as using them for camera gimbals, you need to balance the center of gravity.  The larger motors like the ones I showed have a pretty reasonable amount of torque.  Honestly I'd probably be doing more camera tracking for filming flights though.

@Menno: yeah, I'm thinking of connecting Sparky+GPS to a ground station board via CAN that would inform it of the UAV position.  Since it has a mag that would give sufficient information to point the board at the UAV.

[Reddog]

Why would you use CAN for this instead of simple Serial?

[peabody124]

A few reasons:

1) CAN is a more robust protocol and uses less cables than serial

2) you can add multiple devices to a CAN bus without adding more ports (better for smaller boards)

3) fun :D (someone could reimplement whatever i write serially for fun too)

[peabody124]

More updates and testing on my blog http://buildandcrash.blogspot.com/2013/05/sparky-testing-and-building-no-crashing.html also I have a few spare boards, so if people would like to be testers send me a message.

[AlPackin]

absolutely love it, I'm in :)

[peabody124]

Sweet.  And position hold works :D  it was dusk and windy so no video yet.  I'm glad to get the INS running and happy though.

[Reddog]

I am interested in both Fixed Wing testing, Antenna tracking and OSD testing but am happy to wait if the supply is short and the devs are all over the current supply.

On Friday, February 1, 2013 7:38:24 AM UTC+10, peabody124 wrote:

So while I was waiting for the freedom parts and pcbs (see previous post, flying now) I was goofing around with a new board(s) design.  I've only put about 6 hours into this so it might be a bit rough still.

Here were the goals:

1) serve in three roles: standard flight controller, gimbal controller (ideally talking to the main FC), and brushless gimbal controller.  Partly I just wanted a really cheap but well performing flight controller for my many quads.

2) Single sided for components. Makes it nice and dead easy for hand assembly and can be sticky taped to frame.

3) two layered - keep it cheap for small quantities

4) CAN connectivity

5) Make the brushless driver modular

6) Use 9150 to provide mag information when used as a compass

7) Since I want it to drive a dual axis BGC it's got 12 channels output :)  Great for octo + gimbal in a standard configuration.

So you can see the brushless drivers (Sparky BGC) can piggy-back on the front or back of Sparky or even be cabled off separately.  It will use CAN to talk to the main flight controller (e.g. another Sparky, Freedom or any other board with a CAN transceiver) to get control signals.  Another Sparky board can also be populated with an MPU3300 and put on the CAN bus to provide an extremely low drift gimbal under strong accelerations.  If that works well it will be easy to extend Sparky and add that to the main board.

Things I'm conflicted about:

1) Size.  It could be made smaller, especially by losing the mounting holes (or two of them).  However I'm thinking to use as a gimbal controller it will be nice to really bolt it onto the gimbal well.  Kenn has also written some vibration analysis software that will be able to run on this to help make sure the gimbal is vibration is clean so you actually wouldn't want to use sticky tape in this case.  Also it's about 300 mil more narrow than the standard footprint - I'm almost wondering if that is worth changing the mounting pattern for.  Possibly I'll just extend the board one way to the standard size with nothing there and it can be cut off ?

2) MS5611.  It's tempting to just add this for indoor altitude hold.  This ties into size - with the current size I can probably squeeze it in and if desired this board can do navigation.  *shrugs*

Right now for single quantities the BOM (with MPU9150) is ~30$ so it's going pretty well as a cheap capable FC.  This is just something I'm goofing around with and i'll probably get 10 cheap PCBs made if anyone else were interested in populating their own.

Schematic: https://dl.dropbox.com/u/6645063/Sparky.pdf - I still need to closely read the F3 datasheet so please point out if I made any goofs.

[Alex]

I would like one if you have any spares.

[peabody124]

Ok, I just ordered parts to make another round so if you haven't heard from me you should next week.

[José Eduardo S. C. Xavier]

Can you tell the holes position? Thanks :)

2013/5/20 peabody124 <peabo...@gmail.com>

--

[peabody124]

Three are in the same place as CC3D (30.5mm) and the fourth is 22.9 (0.9 inches) so if you want you can break off the tabbed hole and still have a three position mounting.  I'd probably be more inclined to use tape though at that point (hence the fact there is nothing on the back).

[José Eduardo S. C. Xavier]

I'll do the adapter board anyway so I'll add the holes :)

2013/5/20 peabody124 <peabo...@gmail.com>

[José Eduardo S. C. Xavier]

Have you already decided if you will go with 2 or 3 axis gimbal support? If you have some spare pcbs that can send I can solder it myself :)

2013/5/20 José Eduardo S. C. Xavier <josesc...@gmail.com>

[peabody124]

They aren't exclusive.  The code is written to support 3-axis already and the hardware (Sparky) will support it.  However I need to lay out another board with 3 chips on it.  ReadError offered to throw a few of the gimbal add-on PCBs with one of his orders so I should have a few of those in not too long.

Oh and I got the gimbal working out really well yesterday with an RCTimer gimbal and GoPro.  Now I need to tear it off (always a shame when it's working well) and put it on my larger gimbal and see if that can be replicated.  Video soon!

James

[peabody124]

Here is more testing as a gimbal controller.  It's getting pretty close now.

[José Eduardo S. C. Xavier]

it looks really good :)

2013/5/21 peabody124 <peabo...@gmail.com>

[José Eduardo S. C. Xavier]

Do you know if there are iconic x frame in stock?

2013/5/21 José Eduardo S. C. Xavier <josesc...@gmail.com>

[José Eduardo S. C. Xavier]

offtopic: in my frame and my motor adapter I fix it to the aluminum tube with a hole on it http://dl.dropbox.com/u/6416035/metrocopter/IMG_2638_mini.jpeg and https://dl.dropboxusercontent.com/u/6416035/12040042.jpg

But i look a lot of setups like the gimbal you showed in your last post when tubes are fix in a diferent way, it is only because that way you don't need to make a hole in the tube? or there are anything else like vibrations free etc?

Thanks :)

[AlPackin]

I received my Sparky today and installed it on an old Hover Things 450 that I hadn't flown in a long time.  Its rainy and windy so I just made a short flight but it flew well.  Now I need another GPS and another telemetry transmitter ... another mouth to feed :)

[José Eduardo S. C. Xavier]

great :) I need a GPS to FlyingF3 too, which one should I buy? It'll be use with F3 and Sparky

2013/5/24 AlPackin <faul...@gmail.com>

--

[AlPackin]

I seem to have a problem with U-Center configuring and saving settings on these things.  I have only been successful with this one 

http://witespyquad.gostorego.com/flight-controllers/ublox-6m-gps-w-mounting-backplane.html

but I know others use the one from the "Cheap GPS" or whatever thread.  I have two of them and would love to make them work too

[peabody124]

Ok support for Sparky has been merged into next :)

[José Eduardo S. C. Xavier]

my qual will be ready in a week or two waiting to test the sparky :) just need to figure out how to use PPM in my turnigy 9x. What GPS do you use?

[peabody124]

I'm using this http://www.goodluckbuy.com/crius-cn-06-gps-receiver-v2-0-module.html

[AlPackin]

Is it a stock 9X? 

I added the FrSky RF module to mine and use the FrSky PPM receivers.  It's easy to do.  The stock RF module" unplugs" from the back of the transmitter, you just snip the antenna wire that is connected to it.  Plug in the FrSky module and you are done.  I opened the transmitter and physically removed the stock antenna so I wouldn't have two but you don't have to do that.

I just read something about using a HK 6 channel RX to get PPM from a stock 9X but I don't know anything about that.

http://multicopterworld.com/2011/11/turnigy-9x-ppm/

[José Eduardo S. C. Xavier]

Is it a stock 9X? 

yes

 will see the links

2013/5/24 AlPackin <faul...@gmail.com>

--

[AlPackin]

I looked around quite a bit ... it does not seem that there is a PPM RX solution for a stock 9X.  Most seem to go the FrSky route like I did.

[JamesL]

my sparky arrived too.. one thing i found after plugged in is the board warms up a bit more than other boards.. this might be due to the compact design. one concern i have before flying it is the possible temperature drift on all the sensors.. and especially 5611

[peabody124]

I have noticed that too but haven't had any issues so far.  It gets really running directly off 12V which is what I do on my brushless board.  If you need it we have a third order polynomial correction available for the gyros.  The baro should correct for temperature although I think that value is meant to reflect the ambient temperature more.  To be honest, I've been flying with an issue on my Freedom board that makes it suck half an amp and it gets quite warm and seems to work, so I'm not terribly concerned.  However, please let me know if you see signs it is an issue.

[peabody124]

Took my Sparky + SparkyBGC out to the beach when friend was sailing and got some great video

https://vimeo.com/66987484

[Jörg Rohde]

Are there Sparky Boards aviable?
I have to say I´m in Germany.
jörg

[peabody124]

I'm working on another batch and will be emailing people in the order they requested.

[José Eduardo S. C. Xavier]

if you have a pcb board of the gimbal control that you can send together with the sparky i can solder it myself :)

--

[JamesL]

i guess we need to compile the fw myself since the .tlfw is not there yet on 25's

[pug398]

peabody in reading about sparky there is currently no baro correct? My robotics kids are going to build quad project but they need a barometer. The current one has my quanton on it and I like the kids and all, but just not that much :)

 

On Thursday, January 31, 2013 4:38:24 PM UTC-5, peabody124 wrote:

So while I was waiting for the freedom parts and pcbs (see previous post, flying now) I was goofing around with a new board(s) design.  I've only put about 6 hours into this so it might be a bit rough still.

Here were the goals:

1) serve in three roles: standard flight controller, gimbal controller (ideally talking to the main FC), and brushless gimbal controller.  Partly I just wanted a really cheap but well performing flight controller for my many quads.

2) Single sided for components. Makes it nice and dead easy for hand assembly and can be sticky taped to frame.

3) two layered - keep it cheap for small quantities

4) CAN connectivity

5) Make the brushless driver modular

6) Use 9150 to provide mag information when used as a compass

7) Since I want it to drive a dual axis BGC it's got 12 channels output :)  Great for octo + gimbal in a standard configuration.

So you can see the brushless drivers (Sparky BGC) can piggy-back on the front or back of Sparky or even be cabled off separately.  It will use CAN to talk to the main flight controller (e.g. another Sparky, Freedom or any other board with a CAN transceiver) to get control signals.  Another Sparky board can also be populated with an MPU3300 and put on the CAN bus to provide an extremely low drift gimbal under strong accelerations.  If that works well it will be easy to extend Sparky and add that to the main board.

Things I'm conflicted about:

1) Size.  It could be made smaller, especially by losing the mounting holes (or two of them).  However I'm thinking to use as a gimbal controller it will be nice to really bolt it onto the gimbal well.  Kenn has also written some vibration analysis software that will be able to run on this to help make sure the gimbal is vibration is clean so you actually wouldn't want to use sticky tape in this case.  Also it's about 300 mil more narrow than the standard footprint - I'm almost wondering if that is worth changing the mounting pattern for.  Possibly I'll just extend the board one way to the standard size with nothing there and it can be cut off ?

2) MS5611.  It's tempting to just add this for indoor altitude hold.  This ties into size - with the current size I can probably squeeze it in and if desired this board can do navigation.  *shrugs*

Right now for single quantities the BOM (with MPU9150) is ~30$ so it's going pretty well as a cheap capable FC.  This is just something I'm goofing around with and i'll probably get 10 cheap PCBs made if anyone else were interested in populating their own.

[peabody124]

Sorry sent this as a direct email

Oh Jenkins might not be including it.  I'll talk to Lilvinz.

Here is a recent firmware: https://dl.dropboxusercontent.com/u/6645063/fw_sparky.tlfw .  I actually was just testing it on PH and RTH and had good results even in 15 mph wind :)

Pug, no the final version does have a baro.  The full sensor suite :).  The updated pdf is in the repo:

https://github.com/TauLabs/TauLabs/blob/next/hardware/Production/Sparky/Sparky.pdf

[pug398]

In that case if you have any extra or do another batch put me down for one (or possibly 2). Seems to be just the thing needed. TauLabs has shown awesome stability and altitude hold in the classroom. Especially on a 225mm micro which is usually a bugger to control.

[peabody124]

Here is a video of return to home testing

https://vimeo.com/67166522

[José Eduardo S. C. Xavier]

it looks great but it easy to see how it perform when you post the plots :p

2013/5/28 peabody124 <peabo...@gmail.com>

--

[JamesL]

just a clarifiction on the ports on sparky

flexi... serial comm... 

RcvrPort.......from radio receiver to 0/vcc/3.3/rx.... 

1-10.... pwm out

USB_HIDPort... tx rx above pwm7, another serial comm port?

USB_VCPPort... the two pin CAN port? 

[peabody124]

Second serial port is now working on this branch https://github.com/peabody124/TauLabs/tree/sparky_second_serial

new clean up steps required before merging.

Also to answer your other question yeah 10 outputs, but the last three can be ADC inputs.  This PR https://github.com/TauLabs/TauLabs/pull/568 adds that

Two pin is CAN port.  I'm working on that still.

[peabody124]

Well CAN communications is working now and I used it to get the gimbal to point at a target while flying

https://vimeo.com/67528638

next step will be someone running around with the android app being tracked

[JamesL]

speaking of andorid gcs... it doesnt seem to like the latest firmware.

bt is set 115.2k .. but there is a jar error on android gcs after connected

[Alex]

Wow.  That is awesome. I just got back to see my board has arrived.  It looks amazing and tiny too.  If you plan on doing more boards, please put me down for another one to use as a gimbal controller as well.

Thanks!

[peabody124]

I hope you like it!

So here are some updated packages

OSX: https://dl.dropboxusercontent.com/u/6645063/TauLabs-20130603-0965da02.dmg

Windows: https://dl.dropboxusercontent.com/u/6645063/TauLabs-20130603-0965da02-install.exe (you can use the version from Jenkins too but the setup wizard automatic uploader won't work)

Android: https://dl.dropboxusercontent.com/u/6645063/android-20130603-0965da02.apk

This version adds the Setup Wizard which was written by m_thread.  Please use caution since this is the first version testing this, and of course please say if there are any bugs.

The firmware also now supports the second serial port.

Also for some reason last night the android app wouldn't bind my bluetooth but I don't think it was the software itself.  Let me know.

[ethaller]

Could I also get some Sparky boards? I must say, I'm from Austria.
Would it be possible to get the Sparky design- files (is it designed with Altium?) as a starting point for own board development?

[F1Driver]

Hi Everyone,

If you have any more Sparky boards I'd be honored to put one to good use to replace my coptercontroll that I am struggling with gyro drift. If sparky isn't available can anyone suggest anywhere I can get something else similar. (I'm in the UK)

Cheers,

F1 Driver

[peabody124]

There are a few people on my list currently and I'm a bit backed up but working through it so I'll add you guys and ask you when I have them made.  I'd also highly recommend Quanton which is manufactured commercially and thus available immediately https://groups.google.com/forum/#!topic/phoenixpilot/Se8_LD6-c1g%5B101-125-false%5D .  It has a more powerful chip and more ports too so for people wanting  more capabilities it is a very good option.

AlPackin also got it flying tuned up and doing position hold yesterday https://groups.google.com/d/msg/phoenixpilot/nE9H-naLfUc/NnVer0GIw7AJ

[rip]

i hope iḿ aleready on the list .

[peabody124]

Updated software:

OSX: https://dl.dropboxusercontent.com/u/6645063/TauLabs-20130605-7e9b9df5.dmg

Windows: https://dl.dropboxusercontent.com/u/6645063/TauLabs-20130605-7e9b9df5-install.exe

APK is still: https://dl.dropboxusercontent.com/u/6645063/android-20130603-0965da02.apk

[JamesL]

not sure what has been changed. but right now in UAVOBJ window, the save/sent/request/load are all grey and disabled

[peabody124]

That's only until you click the name of a UAOV.  They they become ungray (since you can do things).

[JamesL]

it gave warning

"QAccessibleWidget::rect: This implementation des not support subelements! (ID 30 unknown for QWidget)

[peabody124]

Can you create an issue with a screenshot?  Also list your operating system, etc. It sounds like there is a regression in the UAVObjectBrwoser.

[JamesL]

on windows 7 .. problem "kind of" diagnosed. it has to do with windows UAC.

the problem occurred when GCS is installed in path  "program file". 

re-installed to non-system folder solved it.

But this still happens to windoes 8.(from jenkins and this.. even it is not in system folder)

[peabody124]

I think you better open an issue about this so we can collect all the information.  We might have to get you to try some old versions as well to see if this is a regression

[JamesL]

will do

[JamesL]

issue about GCS created..

about sparky... everything is getting correct now. 

one thing that i concern about is gyro temperature drift. before the board reaches a stable temperature. one of the axis would drift around 1-3 deg/s.. and the drift only is correctly offset after the board reaches stable temperature. In real flight condition, there should not be much temperature during one flight. It just would be difficult to determine whether or not it is ready.

another i have encountered is motor output issue. i have filed it to git and i will do some further test to help with debugging

On Friday, June 7, 2013 1:50:54 PM UTC-5, peabody124 wrote:

[JamesL]

I was thinking, for FPV would it be the best to use INSIndoor for better alt hold and INS for RTL and fail safe?

[peabody124]

For your gyro temp concerns, you can perform a temperature calibration in one of the attitude tabs.  I normally set it for 40 degrees and put it in the freezer first and then use a hair dryer to get a good temperature range.

AltHold doesn't get affected by the indoor/outdoor distinction.  It runs its own EKF I wrote that is separate from the other systems.  For RTH you definitely do not want indoor mode as it ignores GPS.  Note that right now failsafe still just turns off the motors until we are more confident in RTH and have lots of testing.

[pug398]

Flying sparky around on a mini with spektrum sat rx. Glads thats working again. Looking good :) 

[JamesL]

does it need to be as critical to AQ or just generally leave it somewhere least interfered?

[JamesL]

i guess if i want to test navigation on ground would be using complementary and INSNAV?

[AlPackin]

hey Pug what size is your mini? 5" props?

[pug398]

Hey Ed. Actually its a 240mm so guess it might be called a micro :) 5" props, 2100kv motors, 10A esc and 3s 850ma battery.

http://www.rcgroups.com/forums/showpost.php?p=25247601&postcount=1847

[AlPackin]

nice, I fly the BlueSky RC micro about the same size with a very similar build.

[pug398]

had it on a bluesky micro at first but the pids were so much different than CC+OP that I thought there was a problem. Changing sparky to this frame gave same response so I ended up with gyros set for 2000 range to reduce sensitivity and pids at 10/60/42. Not bad but attitude change still decays slightly so next testing will bump the integral up some more and perhaps bump up derivative as there is just a slight wobble in fast decent.

[AlPackin]

yeah, I had a hard time getting it tuned.  I set it aside when I got into PH testing with quanton.