MultiWii

I installed 200 ohm gate resistors and flyback diodes. Doesent make any difference

Edit:
- I set maxthrottle to 2000 with CLI, default is 1950.
- I did "motor[i] -= (maxMotor - mcfg.maxthrottle)>>2;" in mixer.c, no adverse effects. MicroQuad stays stable at full throttle.

Hi guys.

I got my CJMCU today and just ran cleanflight on it! woohoo!

UART/Acc/Mag OK. Will check motor outputs and RC inputs soon but I just thought I'd let you guys know.

Plüschi wrote:I installed 200 ohm gate resistors and flyback diodes. Doesent make any difference

I guess you are alright regarding the small length and wide arms, tomorrow I'll put it into a mQX frame (geared 8.5mm motors).

Latest Cleanflight binary attached.

PPM and PWM input and PWM output verified on my scope at 16khz. It should fly! Late here now, if anyone flies it let me know how it goes.

Not tested Serial RX, the build has Serial RX support, check the Cleanflight docs for details on how to enable it since it's different from baseflight, see here:

https://github.com/hydra/cleanflight/tree/master/docs

Enjoy!

Ok, I flight-tested my board briefly in the house, made a little video:

https://www.youtube.com/watch?v=KGUfg6Z ... e=youtu.be

The build attached on my previous post was the one used on the board.

I was using stock settings for everything. I will take it somewhere and poke the pids around and post my findings back later.

@dominik quite nice welcome in the CJMCU club. I have seen you also implemented the standard RX functionality now in cleanflight. I have quite some projects and other assignments at the moment and time is limited.

I’m still a bit scared about the brownouts I have on this platform and try to eliminate this first. I may try the bypass for the RX power (and the motors) very likely quite soon. Did some more testing with my NanoHexa (same receiver, 950mAh battery, 8.5mm motors) and this is flying more the 7min. I never have seen brownouts there. Only ones I had a hard landing from low power but at this time I did not check for brownouts and I don’t like to force this Situation again. The CJMCU is always showing brownouts also with a 750mAh battery and 7mm motors after about 3min or even more earlier. Flying is not very smooth yet but this may caused by the missing PID tuning yet but also from signal interruption from the RX.

UPDATE: Get an unintended prove today. There are even no brownouts fro the RX if the NanoHexa does not have enough power to fly anymore. The Hexa was falling down from a not fully loaded battery not sure if STM, sensor and/or motors missing power but the RX was still fine.

mj666 wrote:I may try the bypass for the RX power (and the motors) very likely quite soon.

I would like to know the results. I get no brouwnouts, but to the end of flight the receiver begins flicker its led. No control loss, but still NOT OK.
How about just putting a wire parallel from switch to motor + ?

mj666 wrote: Flying is not very smooth yet

Mine is very very smooth and fine. But it lacks the power to recover from a flip or loop. It just bounces off the ground.

I get no brouwnouts, but to the end of flight the receiver begins flicker its led.

This is the same for me. for the RX31 the flickering means there was the voltage below 3v at least ones.

How about just putting a wire parallel from switch to motor + ?

This is exactly what I want to try with the bypass.

I never get a flip done yet even with my more powerful quadcopters. The problem is not the equipment but my actual skills. At lease there is not so much of a barrier if the equipment is robust and not this expensive.

I did learn to flip with the pockedquad. Its nice to slam the thing into the ground (grass) 10 times and only get a bent prop. My kids didnt like this since they had to retrieve it.

Has anyone tried to use the nrf24l01 chip for this? I have some around, discarded from another project. Getting them to work on tarduino is easy, but on stm32 i dont know my way around. Spec says VCC from 3.6V down to 1.8V.

Edit: got my nrf24l01 receiving something on the STM32F100VB Discovery board over SPI2.

I haven't tried yet, but I was planning to do it, I installed a nrf24l01 into my Devo TX a few days ago...
You can find source code for wltoys V202 compatible RX here:
https://github.com/hackocopter/bradwii- ... /rx_v202.c

Ty Goebish.

I have a Devo10 + 12S, but i did not have the courage to solder in another module yet. If i break the Devo, how will i fly my dsm2/dsmx stuff?

Btw how does this bradwii port feel? Your personal opinion, does that fly comparable to mwii?

Do you know if the pins on the CJMCU are wired to SPI1 or SPI2 ? I want to to make my own protocol, no binding (fix ID), 250kbs, only 6 chan, 3 address, 1 crc. As robust as possible with the 0dbm devices. Not to be better, just to see if i can do it. Transmitter is a PPMSUM module with tarduino 328, 3.3V 8mhz, plugged into the MX22. I got that working with the discovery board.

Yes, i know your way is better.

The pads on one of the arm are wired to SPI2, looks like SPI1 (PA5, PA6, PA7) is not broken out on the board.

Plüschi wrote:Btw how does this bradwii port feel? Your personal opinion, does that fly comparable to mwii?

I haven't touched it for a while, it works but doesn't fly near as well as multiwii/baseflight.
If you've got a spare hubsan x4 or v202 or jxd FC give it a try

Latest cleanflight build attached.

https://github.com/hydra/cleanflight/commits/master

mj666 wrote:@dominik quite nice welcome in the CJMCU club. I have seen you also implemented the standard RX functionality now in cleanflight. I have quite some projects and other assignments at the moment and time is limited.

In git serial rx is enabled too if you fancy building your own .hex.

This evening I was working on WS2812 LED strip support for it but it's complicated by the use of TIM3 for the motors. At one point i got a signal coming out of PA8 on my scope so i connected up my LEDs to it, they all lit up and then after that i've not had it working again, very odd indeed!

Anyone know if these boards are still available? The link at the start of this thread seems to go to a different board now that's got a '328 AVR on it instead of a STM32 and includes extra material between the arms

I did a search on goodluckbuy but couldn't find this board on there anymore.

Thanks, no longer has CJMCU in the description anywhere and that's what I was searching for.

Just wanted to report that I hooked up a Graupner GR-12SH+ HoTT Micro Receiver via SUMD (CH6 RX output to PA3 CJMCU input) and enabled RX_SERIAL in cleanflight and it works just great! Much better battery life than the comparatively giant FrSky D4R-II that I was initially using to test it with.

I might have a crack at getting HoTT telemetry working on the CJMCU board but I think there's not enough flash space to do it.

Note that my Gr-12SH+ came with firmware 1.00 which had SUMH which did not work, the RX had to be upgraded to V1.01 to get SUMD output - see my youtube channel for the video for the upgrade procedure.

I also picked up 3 other different kind of 1S batteries from Hobby King, I will report back with links and flight times once I've flown with them, but I need some battery connectors first...

@dominicclifton: "Copterwerkstatt" once submitted a code for SUMH https://github.com/Crashpilot1000/TestC ... pnersumh.c . Since I have no Graupner gear I can not test it but he said it works.. hope to be helpful.
Cheers Rob

Here's the link to the firmware upgrade procedure for the Graupner GR-12SH+

https://www.youtube.com/watch?v=oyYisiNSiDY

I think all of the CJMCU boards have 128kb flash (I have 2 and they both have 128kb flash). I.e. you need to use the STM Flash Demo loader and choose 128kb in one of the first dialog which is coming up. Also the Baseflight app will flash 128kb. For sure the simulated EEPROM needs to be at the end of the 128K area.

My CJMCU is grounded since a while. One motor stoped working and during replacement I found the Motor is still functional. Next step was replacing the FET but this also not fixed the problem? Need to do more investigation. I have not yet build the second board i have.

If you use motors with too high KV the fets will burn.

I havent seen a single fet burned and i have 4 brushed quads. Open you eyes.
One of the "savy" guys on rcgroups did mention it long time ago. This "upscaling" the aussie does aint going nowhere we want to go.

Ok, latest CJMCU binary for cleanflight now has SUMH support. I will build a 128k CJMCU variant at some point to, just to test to see if my one has 128k flash or not, with 128k it'll hopefully be possible to have led strip support and HoTT telemetry!

Here's the link to the latest binaries:

https://github.com/hydra/cleanflight/tree/master/obj

Problem with led strip is that there's no TIM3 output broken out on the board, only TIM1 & TIM2 which are already used.

My CJMCU is working again. Also installed the bypass from the switch to the power pin of the receiver. After a few initial tests this looks keep the voltage of the receiver more stable. Need to do some more testing. Anyhow I have seen brownouts also ones at my NanoHexa. But here I drained the battery too much.

Hi
I am a beginner in quadcopter.
Recently I bought this board.
Is it possible to use the Prolific PL2303 USB to Serial adapter cable (with 3.3v OUTPUT) for flashing STM32 ?

If the signal outputs and power are 5v or 3.3v the adapter should work. You only need to connect the proper pins (GND, 5V or 3.3V, RX, TX). RX and TX should be cross wired between adapter and the CJMCU board. After initial flash you need to bridge the BOOT0 pins. if the firmware is initially installed you can use the Basefight Chrome App to flash new versions if need. No soldering at the BOOT0 pin is required anymore.

mj666 wrote:If the signal outputs and power are 5v or 3.3v the adapter should work. You only need to connect the proper pins (GND, 5V or 3.3V, RX, TX). RX and TX should be cross wired between adapter and the CJMCU board. After initial flash you need to bridge the BOOT0 pins. if the firmware is initially installed you can use the Basefight Chrome App to flash new versions if need. No soldering at the BOOT0 pin is required anymore.

Thank you mj666.
After a initial flash, two-pin "boot 0" to be connected & Then be separated ?

is there a working baseflight hex for this ?

ezio wrote:is there a working baseflight hex for this ?

It looks like it's included in Baseflight.

@idreams_ir
for running the firmware Boot0 pads need to be bridged all the time. You only need to open the bridge again if the firmware is broken and you are not able to start bootloader mode via the GUI.

@eizo
For Baseflight the is no actual hex available. If I’m back on my computer (sometimes during the weekend) i can build a recent hex file for the CJMCU target. The normal Baseflight hex will not work.
For Cleanflight there is an hex on Github: https://github.com/cleanflight/cleanfli ... master/obj.

Edit: Here is the latest HEX build from actual source. This is not flight tested. There is a flight tested HEX somewhere earlier in this thread.

Hello!
I installed baseflight on my microquad, after initialization, LED2 blinked a few times and then everything crashes. There's no connection to the GUI, I tried installing different versions of baseflight and cleanflight both through FlashLoaderDemo and through the GUI installer, but the results are the same every time. Do you have any ideas, could you please help?
Thanks!

mj666 wrote:@idreams_ir
for running the firmware Boot0 pads need to be bridged all the time. You only need to open the bridge again if the firmware is broken and you are not able to start bootloader mode via the GUI.

@eizo
For Baseflight the is no actual hex available. If I’m back on my computer (sometimes during the weekend) i can build a recent hex file for the CJMCU target. The normal Baseflight hex will not work.
For Cleanflight there is an hex on Github: https://github.com/cleanflight/cleanfli ... master/obj.

Edit: Here is the latest HEX build from actual source. This is not flight tested. There is a flight tested HEX somewhere earlier in this thread.

Thank you.
I will try it when my board arrives

@pJK
Normal Baseflight or Cleanflight HEX file will definitely not work. Always be sure you use a specific CJMCU taget build. After initial flashing you need to close the connection of the Boot0 pads to run the firmware. Be sure if you are using the flashloader demo to switch to 128k flash at the very beginning otherwise you will only flash the first 64k of the firmware.

I found out the serial output format of the Deltang Rx31 is mostly compatible with the Spektrum serial protocol (10bit, 22ms) which is used by Spektrum satellite receivers. The differences will not affect the reception since the different bytes in the protocol will not be used in any of the firmware implementations (MultiWii, Baseflight, Cleanflight). So changed my Rx31 to serial output and reconnected it accordingly. Its working well with the serial output now and you avoid the conversion to PPM since the signal will arrive as it was send by the transmitter. I will now start to change this for all my Deltang receiver based builds if possible. I also have upgraded the CJMCU to the last recent firmware. Here are the actual settings I use:

mj666 wrote:My CJMCU is grounded since a while. One motor stoped working and during replacement I found the Motor is still functional. Next step was replacing the FET but this also not fixed the problem? Need to do more investigation. I have not yet build the second board i have.

Hi,

I've got this exact same problem on mine... what was the fix?

(I haven't yet replaced the FET)

Was bad soldering on the resistor near the fet (cracked). Likely the fet wasn't bad either.

I've been playing around with the CJMCU just learning a bit about Cleanflight and PIDS. But really I'm quite the newb.

I have a few comments and questions. On the latest version (0.63) of the Cleanflight Configurator running on a Chromebook with CJMCU running 1.8.1, the autotune feature isn't available in modes. I looked through the target.h file in the fleanflight source and I see that it doesn't have #define AUTOTUNE. Is that the cause of the lack of Autotune? (I haven't gotten around to actually compiling source and uploading it yet, I'd more likely mess it up than anything).

I'm also curious if the CJMCU needs to have the #define USE_SPI for the SPI pads on the arm of the quad. I'm hoping that I can add a barometer to this board and I was wondering if I would then also need to define the type of barometer I'd be using, which SPI pads it's on, in order to get it to work with the hardware.

And while I have a whole lot more questions, does anyone know if it's possible to get VBAT to work to monitor voltages, or is that not possible because of how the vbat is wired up to VDD?

I also believe that SPI1 (PA5, PA6, PA7) is located on the board, but labelled (on the bottom, next to PA8, PA4, and PA11) as SO, SCK, and SI of course. Though I haven't confirmed this, but it's my guess.

I'm hoping to add a tiny GPS as well at some point along witht the barometer to have a very basic, brushed quad that can do some RTH and GPS Hold.
I also see that the I2C pads are located on the bottom of the board, though I didn't notice this for the longest time. But they're the pads near the FTDI connectors with the two pullup resistors. (I think)

Has anyone else had problems with the CJMCU yawing under throttle? My setup is a little heavy (fpv), and once I get to around 55% throttle, it begins to yaw on me.

Hi,

I try to connect a gy-68(bmp180) Barometer.
Only connecting it to i2c didnt work.

I took a quick look at the Source but my coding Knowledge is limited.

i added to the CJMCU target.h

#define BARO
#define USE_BARO_BMP085

and commented out
#define SERIAL_RX

because otherwise i got an error because of flash size and i anyway use PWM RX.

But i get Reference errors for some Baro Methods.
Someone can help me? Is there much to change to add the Baro or maybe is it impossible because of Flash size or something else?

thx for any answer

Is anybody else having no luck getting these board to work? I flashed cleanflight and baseflight to them with no problem but the motors just wont spin. I dont have a scope but I can see that the transistor output is sitting at +5 along with the vcc that is connected directly to the positive side of the motors. I use the GUIs to try and spin the motors but it has no effect. Using the throttle of the controller doesn't work either even though it registers in the GUI. I bought another CJMCU from a different source (goodluckbuy.com) thinking I just got a defective unit but it had the same behavior. Anybody got any suggestions? I see a few negative posts here...do some of the CJMCU's just not fly?

I'm having the same problem. I am using this board flashed with Cleanflight 1.9 and a FrSky VD5M. I am using the Hubsan 7mm brushed motors which spin OK if connected directly to the LiPo. I can't get the motors to spin no matter what. They don't spin using my transmitter or the in motor test. My throttle goes around 988 at minimum and 2010 at max. All other channels are okay. The green LED2 on the board blinks and LED3 is on. LED1 only turns on when I switch to a profile that has Horizon mode activated but the motors still won't spin. I can see the motor sliders increasing in the GUI when I increase throttle but the motors don't spin. I don't think it is a hardware fault since it is unlikely that all 4 FETs are bad. I measured the voltage on the FET pins and it is constant around 0.64V between the middle min and one of the pins on the side, and around 4V between the 2 side pins. (i know this is a bad explanation but I don't know the FET pinout).
Please help! I can post my CLI dump if required.

Exactly the same here, received a board today too. Everything works fine in the GUI except for the motors.
Edit: Okay, I can't f*n believe it. ALL FOUR FETs are fried out of the box.
Just replaced one, boom, it works.
Edit 2: All four replaced, all four working now.

I noticed that there was some weird sort of residue around the FETs on my board. Could you please share the part number Arakon, looks like I'm going to have to replace the FETs to and send a mail to GLB. It is still somewhat understandable if there is 1 bad FET but all 4? This is not good.

EDIT: I just did some searching and I believe that these are 2TY S8550 PNP transistors. Is this correct?

Yes, and those are not just defective, they're the wrong part for the job.. only rated for 500 mA and probably the wrong pinout.
The residue is some sort of flux.

Replace them with IRLML2502. Cheap to get off ebay and handle up to 4.2A.

Thanks, I will order some. I think the pinout is OK but I'm not sure about the ratings. I will ask some people with working boards what part no. they have.

Arakon wrote:Yes, and those are not just defective, they're the wrong part for the job.. only rated for 500 mA and probably the wrong pinout.
The residue is some sort of flux.

Replace them with IRLML2502. Cheap to get off ebay and handle up to 4.2A.

Thanks for the info, I have 2TY FET as well on two boards from goodluckybuy, no motor spinning either.
Just ordered IRLML2502 from eBay, fun soldering ahead..

Meh. One motor died on me while trying to finally fly it. On top of that, the voltage connection to the spektrum satellite I am using is apparently so dirty that if you throttle up, the Sat loses connection and the board goes into failsafe.
Now to wait for a set of motors again, and figure out how to get a filtered voltage.
Edit: Okay, using a regulated 3.3v pin instead of VCC directly works for the sat at least.