MultiWii

Lecostarius

Wow - I did not know there exists a Spanish subforum

Lecostarius

Hi, this is related to the fact that the angles measure two rotations where one is executed after the other. The order counts: if you first do the pitch rotation, and then the roll you get a different result than if you first rotate around roll and then pitch. Therefore, the formulas for the inversi...

Lecostarius

I am not completely sure, but I guess you should post any Naze32 related question in the 32bit threads and not here. Chances are smaller that ppl in the 8bit section have much to say about the Naze32.

Lecostarius

David,

2nd version is right. The >> operator divides by two as often as its argument says, so by changing >>3 into >>2 we divide by 4, not by 8 any more. To compensate that, we omit the *2.

So it is alt.EstAlt = (alt.EstAlt * 3 + BaroAlt ) >> 2; just as Dennis wrote.

--Leco

Lecostarius

From your description I assume you might have looked at a version of the following: m I checked the code in 2.4 in IMU.cpp and it does not look as if the description in that document is accurate at all. In particular, the relative importance of accelerometer versus barometer is fixed at 1.5%/98.5% (...

Lecostarius

@ravid can you send me the URL of the description how to tune the ALT hold PID? I tried that myself today. All I can say is that D is required, which is probably due to a huge amount of smoothing that is applied to the Baro readings, and hence a large time delay. For me, large values of D (30 to 40)...

Lecostarius

David, I am not the best person to comment on the tuning process for Alt Hold PID since I did not have to tune it, it worked out of the box. I hope someone reads this who can help... anyway, if I find the time I will try myself the tuning and see how it behaves. If I do so, I will keep you updated. ...

Lecostarius

@Plüschi: thanks for the suggestion. It is much better readable like this. @ravid: I have a CRIUS AIOP and ALT HOLD is working. It is stupid, of course - I fly in hilly terrain, and it holds altitude over sea level, which means that it does not hold altitude above ground if going uphill or downhill....

Lecostarius

Here is the MW2.4 code: if(calibratingB > 0) { logBaroGroundPressureSum = log(baroPressureSum); baroGroundTemperatureScale = ((int32_t)baroTemperature + 27315) * (2 * 29.271267f); // 2 * is included here => no need for * 2 on BaroAlt in additional LPF calibratingB--; } // baroGroundPressureSum is no...

Lecostarius

Definitely!

Can you please post here once you have published the material?

Best, Leco

Lecostarius

Many thanks for sharing your observations! I guess the code with the alt estimation is correct as it is currently in 2.4. If you modified the 6 by a 7, what happens is that the relative impact of the new noisy barometer reading is reduced (a bit), and in general, your results are too big by a factor...

Lecostarius

After a quick scan through 2.4 code, I would say it is no problem to run the data logger in 2.4. The main issue is that the current design puts the logging commands all over the code, to make sure they do not spiky-fy the cycle time too much. Writing lengthy datagrams to the flash is time consuming....

Lecostarius

Thanks a lot! I think it will be a Taranis, then. @FengShuiDrone: you are using two RX because each one of them does not have enough channels, right? Other than that, I can not see how you can make more independent functions available than the number of TX channels you have. But even so, with the ma...

Lecostarius

Question: is it worth it to switch from 2.3 to 2.4? If I would move to 2.4, I would certainly port my data logger to 2.4. But so far, I have not seen a compelling reason. Some of the discussion in the 2.4 forum even seemed to indicate that 2.4 is less stable than 2.3 was. I think you have tried it, ...

Lecostarius

Hi all, I would like to control a gimbal. I have a DX6i which has the 4 analog stick channels, plus two switchable channels (on/off, basically).This is clearly insufficient so I need to upgrade. But to what? I am looking for a FC that has 6 (more would be better but 6 will do) analog channels: the 4...

Lecostarius

I dont know this sensor. Is it a current sensor, or a voltage sensor, or both? You say "current" all the time and then talk about voltage, so this part is unclear to me. I have current sensors which go both directions (positive and negative current), and therefore, do not read zero when th...

Lecostarius

Serial.cpp contains a nice interrupt-driven serial library. It offers SerialWrite(uint8_t port, uint8_t c). I added SerialPrint(int port,char *message) to it as this: void SerialPrint(uint8_t port, char *s) { while (*s) SerialWrite(port, *s++); } #include "Serial.h" and you are ready to go...

Lecostarius

I wish I had bought mine at the same supplier...
By the way, can you give me the coordinates of your supplier? I think I will become his customer, too

(you can PM me if you dont want to post advertisements into the forum)

Cheers, Leco

Lecostarius

Hi all, thank you for the great work! -- By the way, where should any bug reports for 2.4 go? Incidentally, I have a bug report . Inflight calibration with the CALIB switch does not work as designed. The fix is as follows: lines 1016..1018 of MultiWii.cpp currently look like this (in 2.4, but the bu...

Lecostarius

Hi all, after my copter got more and more "wobbly" in the air, I discovered the reason: one of the arms - made of a CFK tube - was still very rigid against bending, but no longer against torsion. I could easily turn it by 10-20 degrees back and forth. This was not a problem with the fixati...

Lecostarius

At least with Crius their engineers build quality products & know how to lay out boards as per manufacturer chip design parameters. Crius AIOP is the benchmark for the APM/ 8 bit Megapirates branch because of the reliability of these boards which appear to be copied from APM2.5 but with excepti...

Lecostarius

I have the same with my TX. And no, it does not cause issues - there is a deadband implemented.

Cheers, Leco

Lecostarius

I found a clue to what stops my copter from booting in the cold: I think it is the accelerometer. I connected a cold, non-booting copter to MultiWiiConf and found that the ACC reading indicated that the device was tilted 90 degrees (which it wasn't). The ACC reading would also not change over time, ...

Lecostarius

Hi Ammarali,

I just checked and the log file for a 65 second flight has 10 MByte (in human-readable ASCII) or 520 kByte as Matlab .mat file, so I can post neither of them here. PM me your email and I will send it to you.

BTW, which hardware do you have?

Cheers, Leco

Lecostarius

Maybe you want to take a look into the datalogging for Crius AIOP v2 that I posted elsewhere in this forum. I set out to do exactly the same that you did - analyze in detail the control loop behaviour and how the PID works. I am now logging all the relevant sensor values every MultiWii cycle - so es...

Lecostarius

I think the easiest way to control a flying MultiWii drone via a computer is by Bluetooth. Basically, attach a bluetooth class 1 (for range) dongle to your copter, and then connect to this using simple serial protocol (any terminal program will do that). That enables you to give commands (using the ...

Lecostarius

Everyone, I stumbled over the initialization of the compare registers for timers 3 and 4 for the Atmel Mega controllers. This is used to set the PWM rate. In Output.cpp there is the following line: ICR3 |= 0x3FFF; // TOP to 16383; If ICR3 is set to 16383, the PWM frequency will be half of (16 MHz / ...

Lecostarius

Well, I tried to optimize them. It is embarassing to say, but I do not notice -probably due to a lack of flying experience- a huge difference in performance between P values as long as they are in the range between 2 and 10 or so. Less than 2 or more than 10 is pretty bad though. The response gets &...

Lecostarius

I did a logging of the Gyro values inflight and found that for the default values of D, the motor PWM signals are sometimes dominated by D response, probably due to noise with the Gyros (probably vibrations, I guess). That D induces control signal is very spiky. Therefore I reduced D to very little ...

Lecostarius

A bit unrelated but I did not want to open a new thread for that. It took me a while to figure out how the three LEDs on the CRIUS AIOP v2 board are connected, and how I can program them. Thought I can save someone else the time to look it up. Here it is: red LED (LED A) - Arduino digital pin 13 - A...

Lecostarius

I still do not know the reason, but it is not the Transmitter. I put the copter outside and kept the transmitter indoors in the warm, and the problem appeared again. So it is most unlikely unrelated to the radio.
If I eventually find the reason for that problem I will post it here.

-- Leco

Lecostarius

All, not power-up related, but in case you want to use the OLED and control the row and column when printing characters to the screen and want to use the 8x5 font that is supplied with the MWii code, the function that sets the cursor position is not right. It looks like this (in LCD.cpp): void i2c_O...

Lecostarius

Thanks for the input, I did not think about that. Way to try without oscilloscope would be to keep the TX outside for 30 min and the copter indoors, and then try to fire it up with the cold TX. I will try and report the result.

Cheers, Leco

Lecostarius

Hi everyone, I encounter the problem that my copters are not booting up properly. I have the assumption that it is related to temperature - they are always booting without trouble when it is warm (indoors), and often have trouble when I try to fly outside when it is cold. When I bring them back into...

Lecostarius

Hello Brewski, it is based on the MultiWii code from the official svn repository, last August. So most likely its a standard 2.3. The details about the svn version that I used originally, and the directory in the svn that I cloned, are in the documentation (cannot check it out from here, sorry). Che...

Lecostarius

Everyone, just to avoid confusion: this is *not* any flavor of official release. It is just something I wrote for myself and what I considered might be useful for others too, so I am sharing it. Beginners should definitely stay away from it and rather go for one of the well-maintained and thorougly ...

Lecostarius

Everyone,

and here is the final part, the MultiWii code itself.

Cheers, Leco

Lecostarius

Everyone,

attached is the code of MultiWiiConf.

Cheers,

-- Leco

Lecostarius

Everyone, I put together a MultiWii version that supports logging to the dataflash chip that is part of the Crius AIOP2 V2 board. It is based on a great low-level library made by Jordi Munoz and Jose Julio. Many different parameters can be logged in-flight, and the log can be downloaded via an exten...

Lecostarius

Hawk,

did anybody reply to your request how to add / submit code to the main project?

Lecostarius

I tried Sonar myself. It took me a long time to make it work. I could not get the cheap boards to work at all (they only worked reliably as long as the copter was not running and producing noise and turbulences and vibration), and even for the expensive MaxBotix Sonar, I had to use the one with the ...

Lecostarius

Hi all, I am using ESCs made by YGE, which have a special calibration procedure that the current MultiWii code can not generate. Therefore, I created some code which calibrates all of them at once. I post it below in the hope that it is useful to others that use YGE ESCs. The code is also useful to ...

Lecostarius

I just purchased a set of the new "cam-carb-light" Aeronaut propellers (first time for me to use them). I had experimented with several manufacturers of propellers, and so far I am using APC propellers; in particular the APC are sturdy enough to withstand an occasional crash and very durab...

Lecostarius

Hi Leo, thanks a lot for the beautiful update of MultiWiiConf! I am working on some additions to MultiWii and MultiWiiConf, and I like your improved look so much that I have now refactored my code to reflect all your modifications. If I ever manage to feed back my additions to the community, your im...

Lecostarius

Great that there is some interest. I am happy to share what I have. Sadly it is ongoing work, so not finished yet. Reading the logged values is done with MultiWiiConf - there is a new register tab called "Dataflash", and you just connect your copter after flying (and logging ), and hit the...

Lecostarius

Hi all, I am currently writing a Dataflash logging facility for MultiWii, using the Crius AIOP 2.0 built-in dataflash (AT45DB161 from adesto). First basic things are already working fine, including communication with MultiWiiConf - however, I encounter an unexpected difficulty when trying to create ...

Search found 46 matches