Originally Posted by obelix662000

Miru, thank you for your wonderfull work!
I've built the mod and everything went fine from the very momemt. I made a few flights with no problems. Suddently it stoped working.
1. When the mod is connected (regardless if Tx is on or off) Ar drone cannot pass through normal booting, it just stop with red motor's leds (no motors initial movement).
2. I checked electronics many times with the oscillosope and hyperterminal testing.
Seems Ar drone cannot start when "1" is presented at the RX pin of the drone.

It still works with iOS when the mod is physically disconnected.

What could be the reason? Thanks.

Try to set the drone on a different surface (piece of cardboard without any lettering) for boot and check if things come back to 'normal', because that would indicate a problem with what the bottom camera is looking at. The bootloader does check the USB port (same connector), make sure you don't have anything connected to pin 1, 3 and 5.

Originally Posted by obelix662000

Miru, thank you for your wonderfull work!
I've built the mod and everything went fine from the very momemt. I made a few flights with no problems. Suddently it stoped working.
1. When the mod is connected (regardless if Tx is on or off) Ar drone cannot pass through normal booting, it just stop with red motor's leds (no motors initial movement).
2. I checked electronics many times with the oscillosope and hyperterminal testing.
Seems Ar drone cannot start when "1" is presented at the RX pin of the drone.

It still works with iOS when the mod is physically disconnected.

What could be the reason? Thanks.

Try moving the rudder stick to right(i.e ESTP ) when it's in LAND mode..I think that should reset and you should be able to fly..I had a different problem and it worked.

Folks,

So far mod works great and as i noticed, I felt there are some sudden glitches in the drone behavior to the TX stick movements. And also there is a slight delay in the communication as well..Is this normal ???. I don't feel it's a signal interference..Any suggestion guys..

Thx

Originally Posted by miru

Try to set the drone on a different surface (piece of cardboard without any lettering) for boot and check if things come back to 'normal', because that would indicate a problem with what the bottom camera is looking at. The bootloader does check the USB port (same connector), make sure you don't have anything connected to pin 1, 3 and 5.

Thanks, but problem is more complicated. I double checked everything including soldering new connector. When the mod is attached (which is equivalent to "1" at Rx pin of dron's UART) Ardrone boot sequence is different. I have logs for both normal (no mod attached) and wrong (mod is attached) recorded with serial cable. Unfortuantly I'm on iPad now, so I'll send both logs on Monday. Also (may be its related to the problem) now I cannot update firmware (with lowering version number trick), the update can be loaded but not started. Thanks again.

Originally Posted by miru

Only if you have Drone 2 now.

Still flying drone 1.
Thanks

Originally Posted by LittleG

They are for the drone 1. The cables are IDE cables, I got the idea from miru's
photo of his module for the drone 1 from his earlier posts. I made them for nuroo.

And they work flawlessly. excellent build. Great guy. He even helped me setup transmitter (DX7) on Skype. And made an easy to follow install video to boot.

Have a 2nd drone with miru mod also (thx to LittleG), is there a sketch configurator profile for DX8 like there is for DX6, DX7 or can I just use the DX7 profile with DX8?

Originally Posted by obelix662000

Thanks, but problem is more complicated...., so I'll send both logs on Monday.

1. Now: Now the problem is related not to mirumod functionality but internal Ar drone boot sequence and UART Rx pin.
2. How happened: The problem appeared when AR drone was under RC control and iPad video simultaneously. This was about 10th flight with mirumod (all them were awesome without any problems). The AR drone landed slightly touching the obstacle, I don't remember sequence of powering off iPad, battery and RC. Next time AR drone with mirumod attached could not pass initial startup (red motor's leds).
3. Reasons: The problem may be or may be not caused by mirumod.
4. Ar Drone is operational under iDev.

What is now: When no hardware attached to the connector the boot sequence is normal the log is in the "no_hardware_good.txt" file. When logical "1" is presented at the Rx pin of AR drone's connector boot sequence is incorrect ("hardware_wrong.txt"). Logical "1" at Rx is normal when mirumod is attached and waiting for the end of ardrone init; that is why the whole system does not work. The "hardware_wrong.txt" is recorded with level converter attached, but mirumod is not attached to the level converter for simplicity (from the above its obvious that this is equivalent to the attached mirumod at this stage). Level converter pull Ardrone's Rx pin to 1.8V providing logical "1" at the Rx pin. When this pin is grounded from the HV side the boot is OK. Ar drone's Rx pin itself is not pulled, so it has nearly "0" logical level if no hardware attached.

Both logs are highly reproducible (except some internal addresses).

Connectors, wiring etc. are OK. Checked a million times.


Please advice. Many thanks.

Originally Posted by obelix662000

1. Now: Now the problem is related not to mirumod functionality but internal Ar drone boot sequence and UART Rx pin.
2. How happened: The problem appeared when AR drone was under RC control and iPad video simultaneously. This was about 10th flight with mirumod (all them were awesome without any problems). The AR drone landed slightly touching the obstacle, I don't remember sequence of powering off iPad, battery and RC. Next time AR drone with mirumod attached could not pass initial startup (red motor's leds).
3. Reasons: The problem may be or may be not caused by mirumod.
4. Ar Drone is operational under iDev.

What is now: When no hardware attached to the connector the boot sequence is normal the log is in the "no_hardware_good.txt" file. When logical "1" is presented at the Rx pin of AR drone's connector boot sequence is incorrect ("hardware_wrong.txt"). Logical "1" at Rx is normal when mirumod is attached and waiting for the end of ardrone init; that is why the whole system does not work. The "hardware_wrong.txt" is recorded with level converter attached, but mirumod is not attached to the level converter for simplicity (from the above its obvious that this is equivalent to the attached mirumod at this stage). Level converter pull Ardrone's Rx pin to 1.8V providing logical "1" at the Rx pin. When this pin is grounded from the HV side the boot is OK. Ar drone's Rx pin itself is not pulled, so it has nearly "0" logical level if no hardware attached.

Both logs are highly reproducible (except some internal addresses).

Connectors, wiring etc. are OK. Checked a million times.


Please advice. Many thanks.

Very interesting, but you are making a mistake not connecting the Arduino to the LLC. I think that the Arduino bootloader drives the Arduino TX pin low for a short while after a poweron RESET. This must make the drone bootloader happy. I never had any problems there with drone 2, maybe you have a different bootloader?
I do remember from the very early stages of the project, that the drone bootloader puts out a short pulse on the RX line and checks if it can do it, if so it would mean the line is in a high impedance state (aka nothing connected) and go through the normal bootsequence. Somehow this is a burried subject and once people get around it they never come back and investigate, since it might have changed anyway.

Originally Posted by miru

Very interesting, but you are making a mistake not connecting the Arduino to the LLC. I think that the Arduino bootloader drives the Arduino TX pin low for a short while after a poweron RESET. This must make the drone bootloader happy. I never had any problems there with drone 2, maybe you have a different bootloader?
I do remember from the very early stages of the project, that the drone bootloader puts out a short pulse on the RX line and checks if it can do it, if so it would mean the line is in a high impedance state (aka nothing connected) and go through the normal bootsequence. Somehow this is a burried subject and once people get around it they never come back and investigate, since it might have changed anyway.

Thank you. I did happen when all system was connected, also for the first tens of try Arduino was connected, for simplicity I've disconnected Arduino because I did not notice any Rx lowering with the oscilloscope (also from your script it initiates UART almost immediately after power-up), may be this time is just too short to see. The whole system was OK for several flights, it happened suddenly so I do think the bootloader is OK. High impedance state is not necessarily, it can be "0" also. I'm thinking to modify your script to force "0" for longer period. May be your suggestion is correct and my drone is at the border uncertain state between tested "rx-yes" and "rx-no" while booting, few more milliseconds can make sense.

Just wanted to give you a heads up on what is coming down the next week or so...

• some cleanups in the Arduino code
• rx_read() is checking the UP time of a signal against the current time and faults the channel if it is too long.
• limit of <rol>,<ptc>,<gaz>,<yaw> values to +-1.000
• added NSQUIET back in for boot.
• new config option 'record_on_usb' for D2 (ignored on D1)
• D2, auto start/stop video recording on LAND->FMx/FMX->LAND if 'record_on_usb' is ON AND there is a stick with space for a video.
• Auto pilot program is not part of the compagnion program any more because the whole program became too big to fit as a 'piggyback' on the Arduino proram. This means, you have to put the auto pilot program on an USB stick where the companion program can get it or you have to 'ftp' it to the drone (D1)
  .

You're awesome miru, thanks. Hopefully my new motor will be in by the time this is out. My brother in law crashed my drone fairly badly. :F

Hi,

i build the mod for my drone (version 1) too.
Since I have no receiver with 5 channels or more, but one withe a SUM-Signal Output,
i have attached a 8 Leg Bug (ATTiny45) on the bottom side of the Arduino. This Controller
converts the PPM Datastream from the Receiver to 5 independent Channels, that i wired
to the correspondending Pins of the Arduino.
Big thank to miru for his awesome work !

Nos!

Originally Posted by miru

Just wanted to give you a heads up on what is coming down the next week or so...

• some cleanups in the Arduino code
• rx_read() is checking the UP time of a signal against the current time and faults the channel if it is too long.
• limit of <rol>,<ptc>,<gaz>,<yaw> values to +-1.000
• added NSQUIET back in for boot.
• new config option 'record_on_usb' for D2 (ignored on D1)
• D2, auto start/stop video recording on LAND->FMx/FMX->LAND if 'record_on_usb' is ON AND there is a stick with space for a video.
• Auto pilot program is not part of the compagnion program any more because the whole program became too big to fit as a 'piggyback' on the Arduino proram. This means, you have to put the auto pilot program on an USB stick where the companion program can get it or you have to 'ftp' it to the drone (D1)
  .

Yeeeahhhhh,

Since I've never been able to test your "failsafe" function, could you please explain it again ?

When the sketch detect a faulty channel, what's happening ? does it immediately land ? does it "fix" a hoovering position before ? does it hoover for a time ?

Sev

Originally Posted by miru

• Auto pilot program is not part of the compagnion program any more because the whole program became too big to fit as a 'piggyback' on the Arduino proram. This means, you have to put the auto pilot program on an USB stick where the companion program can get it or you have to 'ftp' it to the drone (D1)
  .

Glad to see the mod is evolving. Quick question - will there be any visual feedback if the auto pilot program is missing or it's version on the USB/drone is not compatible with the sketch?

Originally Posted by Nosferatu

Hi,

i build the mod for my drone (version 1) too.
Since I have no receiver with 5 channels or more, but one withe a SUM-Signal Output,
i have attached a 8 Leg Bug (ATTiny45) on the bottom side of the Arduino. This Controller
converts the PPM Datastream from the Receiver to 5 independent Channels, that i wired
to the correspondending Pins of the Arduino.
Big thank to miru for his awesome work !

Nos!

Nos, what kind of a receiver do you have, I can't find a DSM2 receiver with a sum signal output. It would make the wiring so much easier...
BTW, you can use a Johnson counter to decode a sum PWM signal.