Thread Tools
May 23, 2021, 12:34 AM
OlliW
Thread OP
Quote:
Originally Posted by Sebbbb
That's annealed nylon with 15% carbon fiber, it's stiffer than it looks
I have heard that so often ... and later folks are complaining about not so good gimbal operation ... stiffer than it looks and stiff are not the same

maybe have a look at my 3D printed gimbals of the last 1/2 year, and the arm designs ... it's for a reason

I guess I would make them at least 4 times more material LOL

just saying

btw, the carbon fiber thing is IMHO pretty much totally worthless ... I think there are sufficiently many tests on yt which allows to conclude this
Sign up now
to remove ads between posts
May 23, 2021, 08:22 AM
Registered User
Quote:
Originally Posted by OlliW
maybe have a look at my 3D printed gimbals of the last 1/2 year, and the arm designs ... it's for a reason

I guess I would make them at least 4 times more material LOL

just saying
Will do, where do you have these? I searched the wiki and didn't find them.

Anyway I'm a fan of trying things myself first and learning from my mistakes What will happen with too thin parts, will it show more vibrations? Wil I see a bad behavior only during flight or already on the ground?

Quote:
Originally Posted by OlliW
btw, the carbon fiber thing is IMHO pretty much totally worthless ... I think there are sufficiently many tests on yt which allows to conclude this
What material would you use? I printed my drone with this and wasn't able to destroy it yet, other than previous versions where I tried PLA, PETG and PP...
Even if these carbon fibers don't improve strength, I read they reduce warping which already helps. At least I hope
May 24, 2021, 05:13 AM
OlliW
Thread OP
I just have smartphone available currently, so somewhat limited
it is not in the wiki but threads, look at attachments to this and the tstorm thread, easiest way to find pics

i m not very experienced in 3d print yet, spent too little time on it yet, I tried peg but wasnt successful yet, just did pla so far. however from basic principles and from yt videos one can not expect massive differences due to material... at the end it needs to melt at few 100 c degrees, right, ergo you re never getting properties of materials melting at few 1000 c degrees lol.
i would be interested to learn from your 3d experience, would you post an link to the exact material?
May 24, 2021, 08:33 AM
Registered User
Just was outside trying it out, but still saw some vibration on the video.
Thing is, I don't know what to expect from a well set up gimbal yet, how good it counteracts especially vibrations...

So I'll first try printing the arms thicker, maybe that helps

Quote:
Originally Posted by OlliW
i would be interested to learn from your 3d experience, would you post an link to the exact material?
I used this here:

https://fiberlogy.com/de/filamente/nylon-pa12cf/

and a 0.6mm nozzle.

Sebastian
May 29, 2021, 10:54 AM
Registered User
Hello,

I've been very unhappy about the problem of wireless control for newer GoPro cameras. I'm using this for triggering photos when executing STorM32 scripts. So far I've been using wifi with ESP8266 boards, but this was not good. Big energy consumption, RF interference, and did not work for GoPro 8 and 9. They have moved from wifi to Bluetooth Low Energy (they still have wifi but it must be activated by using BLE). But if you have working BLE, wifi is not necessary. There is a very nice repo with a python BLE API:

https://github.com/KonradIT/gopro-ble-py

but I could not find an implementation for ESP32 (which is perfect to this aim, small and cheap). So I've adapted a NimBLE client and... it works! It's even better than the previous wifi implementation, as there is no need to hard-code the cam address or name.

It works as follows. First one needs to start pairing with the gopro cam app on the camera. Then the board with the uploaded program (Arduino for ESP32 with installed NimBLE library) needs to be powered up. The cam and board are then paired (the cam exits the pairing mode). And now when the pin 23 gets the high voltage (powered from the IR output), a photo is taken. Lots of parameters can be set up, I'm posting a development version. But it works, soon I'll test it in flight. The problem is that GoPro 9 is somehow bigger than the older cams and I need to rebuild the gimbal.



There is a new version of ArduCopter, 4.1.0 beta 3 with quite important changes, especially concerning Auto flight mode with Spline waypoints, and DSHOT telemetry. @OlliW, could you please publish the BetaPilot based on it? Maybe you could describe what needs to be changed and how to do it? I'll look at this using Git, but it would be a good starting point.

EDIT: some more remarks about the BLE solution. The problem is that the connections stop working after around 5 minuts. Not a problem for me as my panorama flights are usually 3 minutes, but probably there is a timeout or the connections should be reestablished. But there are also more advantages in comparison to wifi:

-the camera doesn't need to be switched off to connect again
-during some time (8 hours?), there is even no need to switch on the camera, BLE does it on its own

EDIT2: I post an updated version. The previous one was beeping instead of taking photos. All is running fine now. It did not want to work with a certain ESP32 board instance (while working fine with other boards of this type).
Last edited by maciek252; Jun 20, 2021 at 07:32 AM.
Jun 01, 2021, 01:40 PM
Registered User
Hi! I wonder is it possible to change the roll degrees in the Initialization sequence at startup?

I have modify a gimbal and now it can only roll to around 24,5 degrees,and the Initialization sequence seems to
want's go further. I fly gentle (only taking photo) so I think it can work with max 24,5 degrees.

*Edit: I went through the "Configure gimbal tool" again and now it works!
I secured the roll and pitch movement before the 45 degrees tilt procedure
maybe that's what made it work now .
Last edited by sjak; Jun 05, 2021 at 07:39 AM.
Jun 14, 2021, 10:45 AM
Registered User
I've had a lot of success working with the storm32 but am having a weird issue with a new one (v1.3x). The pitch axis works like a charm, but the roll continuously oscillates slowly on startup as if it doesn't see data from the IMU correctly (see attached screenshot). It will do this indefinitely if I force it into normal mode after it fails to initialize at startup.
Any ideas?

EDIT: Found a solder bridge shorting one of the FETs, fixed and working

-Ian
Last edited by IanBurwell; Jun 14, 2021 at 01:02 PM.
Jun 16, 2021, 01:54 AM
Registered User
So, there is an ArduCopter 4.1.0 beta4. Important for me as it adds the support for one more Cuav Nora's IMU and fixes an bug preventing auto missions with spline waypoints. Probably this bug made my three trips to the flying place completely unproductive, as I was struggling with the "terrain data failsafe" error instead of exploring STorM32 cooperation with the BetaPilot...

But I think it is very likely that most of the changes do not interfere with the BetaPilot's stuff and the new versions should be merged automatically by git? Maybe reorganizing the git repository would help to automate this process? How about subtrees for including external repos (ArduPilot in this case)?

https://medium.com/@porteneuve/maste...s-943d29a798ec
Last edited by maciek252; Jun 16, 2021 at 04:32 AM.
Jun 16, 2021, 08:53 AM
OlliW
Thread OP
unfortunately, also beta4 seems to have issues, one of them affecting H7 and thus me LOL

but you are right, it should be not difficult to go from beta2 to beta4, so I will try it the next days ...

concerning more automated approaches, it's always a balance between "time to set it up and have it working" and "time saved by having it done". If there would be many users of BetaPilot some automatism might be worthwhile, but given that there are less than a handfull I guess it would just kill more time tnan it saves
Jun 16, 2021, 09:36 AM
Registered User
shuricus's Avatar
Hi Olli!

I have a good news and a strange news!
After a couple of weeks of troubleshooting that weird gimbal shaking issue I've mentioned earlier, I found a very simple solution - I've disconnected Storm32 uart cable and all shakes are gone. Gimbal is working so cool as it has never did before. I was really very surprised, but it looks like the only reason for such a behavior is V1 option set for Storm32Link Configuration. With Off or yaw drift comp. - no shakes are visible.

Here are two videos to compare one with uart connection and one is without. I didn't fly yet with Storm32Link Configuration set to Off, but I'm sure it will be good because the shakes you see at the video is very well visible by the eye too. It's moving both axes for about 1mm but mostly by the roll axe.

I also attach log files of some different flights with shaking gimbal. I've set ZFLAG to 8 at one of the flight, but that didn't help. I think it was Log #1 or 2, but I'm not sure. https://drive.google.com/file/d/12Yc...ew?usp=sharing

might be not finished rendering
Storm32link shake (0 min 25 sec)


Gimb test (0 min 37 sec)




Regards.
S.
Last edited by shuricus; Jun 16, 2021 at 10:29 AM.
Jun 17, 2021, 03:40 AM
Registered User
Quote:
Originally Posted by OlliW
unfortunately, also beta4 seems to have issues, one of them affecting H7 and thus me LOL
I fly Nora Cuav (H7) and it works, what could go wrong? Where you can check these issues? (other than in a flight)

Quote:
Originally Posted by OlliW
concerning more automated approaches, it's always a balance between "time to set it up and have it working" and "time saved by having it done". If there would be many users of BetaPilot some automatism might be worthwhile
Technically, I think that if the updated versions are to be prepared at all, the number of users does not matter that much

I've given a first try to subtrees and it seems to work:

-create an empty git repository
-add an ardupilot's remote:
git remote add -f ardupilot_orig https://github.com/ArduPilot/ardupilot/
-create the subtree using this remote:
git subtree add --prefix ardupilot_orig ardupilot_orig master --squash
-now I can pull the selected versions of the original code, I think it was beta4 (I've found the commits id in the repository log, there are no separate branches nor tags for betas):
git subtree pull --prefix ardupilot_orig ardupilot_orig 1a4f67955df0488266961463926ef86133c2952f --squash
-I've modified the file git add ardupilot_orig/ArduCopter/GCS_Mavlink.cpp
added to commit, commited, pushed (into my repo)
-and now I have switched the ardupilot_orig into another commit, corresponding to beta3
git subtree pull --prefix ardupilot_orig ardupilot_orig 37070bfe68ed06fead0978e6b2b8e567679deb2b --squash

and the locally modified file is still here, while the ArduCopter's code has been updated accordingly. A merge commit/message to be filled has appeared, but everything was done automatically as there was no conflict.
Here one can create a branch, a tag or generate required binaries.

I'll try to do all this once more Maybe I've done some more pulls etc
Last edited by maciek252; Jun 17, 2021 at 09:02 AM.
Jun 18, 2021, 05:00 AM
Registered User
shuricus's Avatar
Hi Olli!

May I ask do you have any ideas about my shaking problem?
Jun 18, 2021, 05:15 AM
OlliW
Thread OP
I'm sorry, I'm not neglecting you, I'm just terribly busy currently with all sort of things and also another project, I hope to come back to STorM32 on weekend
sorry, sir
Jun 18, 2021, 05:19 AM
Registered User
shuricus's Avatar
Sure! Great!
Thank you very much!
You usually answer very quickly, that's why I was a bit confused of the silence ))))
Last edited by shuricus; Jun 18, 2021 at 02:34 PM.
Jun 26, 2021, 09:00 AM
Registered User

Issue with 2.59a and motors switching off at low voltage


Hello,

I recently switched from LiPo to LiIon and noticed that the motors switch off at low voltage. I have a 4S setup and a BEC in front of the STorM32, which regulates to normally 12V. But as LiIon can go down further, the voltage behind it may be only around 10V.

Although I set Setup -> Low Voltage Limit to" off", when the voltage drops in a range of 10.8 and 10.0V, the motors switch from ACTIVE ACTIVE ACTIVE to OFF OFF OFF.

I just tried with a power supply:

Voltage at startup was 11.71V. When regulating the voltage down, at around 10.8V motors switch off.
When I reset the controller with a voltage around 10.0 and 10.8 V, the motors stay off.
When I reset the controller with a voltage around 9.8V, everything works fine again..

Also the connected camera (Firefly Split 4K) beeps as if it gets signals, but I don't understand why..

I use STORM32 MICRO V3.33I / STORM32 MICRO MOTOR TRIPLE V0.30 boards.

Any ideas?


Quick Reply
Message:

Thread Tools