Espritmodel.com Telemetry Radio
Reply
Thread Tools
Old Oct 30, 2012, 02:17 PM
hacker
Canada, BC, North Vancouver
Joined Dec 2010
934 Posts
Quote:
Originally Posted by LittleMo View Post
I really wanted to try my simonk'd quad in the high winds but I broke a prop on the workbench this morning (watch out when resetting your multiwii with the TX on; yes there was blood) and my spares are in the mail.
Can you explain what happened here? Is this something intentional on the MultiWii board?

Quote:
Originally Posted by Razor_Racer View Post
How is the analog comparator set up to detect zero crossing?
It's set up to compare between the "MITTEL" and the "NULL_A", B, or C. There's lots of stuff on the web about this. Basically, it's comparing the average voltage of all phases (divided down to not exceed 5V on the ATmega pins) with the voltage of the phase not being driven high in the current commutation step.

Quote:
Originally Posted by fantasy1988 View Post
hello simon
you can write any program atmega 8 run forward, run reverse and brake. I want to use it for car
thank you
You just need to set RC_PULS_REVERSE = 1 and MOTOR_BRAKE = 1 in the source and rebulid. You will probably also want to adjust the brake speed and force. There's no brake at lower than full reverse PWM input, though I do want to add this -- it just brakes at neutral for now. Most other controllers will brake until they see the motor stop, which requires merging the run and brake loops. Let me know which ESC you are using, and I can build a .hex for you for testing if you can't get it to build.

Quote:
Originally Posted by sharonx View Post
The problem is this : when I try lo lift off the quad is very unstable and tries to fly forward and right and possibly flip over.
When I tried the motors without props it seems that they are not reacting in real time but with big delay.
You've confirmed the motor order is correct? Can you show a video of this, maybe with only one ESC plugged in? You said DJI, but is it a Naza or something else?
simonk is offline Find More Posts by simonk
Reply With Quote
Sign up now
to remove ads between posts
Old Oct 30, 2012, 03:20 PM
Fly like never crash as always
tascheri's Avatar
Rio de Janeiro
Joined Mar 2007
275 Posts
I would love one for an F-20... - I have been fooling w/ AVR studio but no success... Im a PIC guy LOL...

Quote:
Originally Posted by simonk View Post
You just need to set RC_PULS_REVERSE = 1 and MOTOR_BRAKE = 1 in the source and rebulid. You will probably also want to adjust the brake speed and force. There's no brake at lower than full reverse PWM input, though I do want to add this -- it just brakes at neutral for now. Most other controllers will brake until they see the motor stop, which requires merging the run and brake loops. Let me know which ESC you are using, and I can build a .hex for you for testing if you can't get it to build.
tascheri is offline Find More Posts by tascheri
Reply With Quote
Old Oct 30, 2012, 03:46 PM
Registered User
LittleMo's Avatar
United States, VA
Joined Dec 2006
1,488 Posts
Quote:
Originally Posted by simonk View Post
Can you explain what happened here? Is this something intentional on the MultiWii board?
I honestly don't know and it has never happened again despite resetting many times since then. I even tried to recreate the situation (with no props this time ) and nothing happened.

I had just flashed a new MultiWii firmware, my TX was on with throttle all the way down, I reached over and pushed the reset on my pro-mini board then I heard a motor spool up and I just managed to pull my arm back before the quad flipped and shrapnel went everywhere. I would guess it was running about 50% throttle (could have been more, I don't know) and it was just one motor. It shut off by itself once MultiWii was booted.

Some guesses: Maybe MultiWii corrupted the PWM state as it was booting. Maybe for some reason MultiWii went into failsafe mode as it booted (this will crank up the throttle even if idle; not a good design IMHO). Or, less likely I think, maybe the ESC went bonkers during the reset because I do notice lots of odd beeps and such when doing a reset.

Edit:
You know, I hadn't noticed but it hit the arm so hard that it almost cut a zip tie completely off. Maybe it was in fact running full throttle. I don't know, I didn't have time to think.
LittleMo is offline Find More Posts by LittleMo
Last edited by LittleMo; Oct 30, 2012 at 04:15 PM.
Reply With Quote
Old Oct 30, 2012, 06:53 PM
Registered User
Joined Nov 2008
23 Posts
Hi,

I think, all FC can produce noise on PWM outputs during initialisation. The similar issue was reported by Heiko on OpenPilot:
http://code.google.com/p/wii-esc/iss...ail?id=8&can=1

regards,
ziss_dm
ziss_dm is offline Find More Posts by ziss_dm
Reply With Quote
Old Oct 30, 2012, 07:18 PM
hacker
Canada, BC, North Vancouver
Joined Dec 2010
934 Posts
Quote:
Originally Posted by LittleMo View Post
I had just flashed a new MultiWii firmware, my TX was on with throttle all the way down, I reached over and pushed the reset on my pro-mini board then I heard a motor spool up and I just managed to pull my arm back before the quad flipped and shrapnel went everywhere. I would guess it was running about 50% throttle (could have been more, I don't know) and it was just one motor. It shut off by itself once MultiWii was booted.
Ok, I think I know what happened.

If the flight board MCU goes tristate on reset while a motor PWM output was high, the pin will continue to float high until air resistance or a very weak resistor eventually pulls it down to where it becomes detected as low. If it looked high for up to 2400 microseconds, and the ESC is armed, it will follow this request until the RC timeout happens. 2012-09-30 changes this timeout to be about 128ms, but does so by forcing a low rc_timeout on every successfully-received pulse once running, which is after a higher rc_timeout is allowed to build when stopped to verify a stable input before starting, armed or not. So, it should normally stop within 128ms, but not if there is only one pulse after a series of idle pulses and the motor is not spinning. In this case, it will work like older versions with a ~1 second timeout. Will verify and push an update if so.
simonk is offline Find More Posts by simonk
Reply With Quote
Old Oct 30, 2012, 10:51 PM
Registered User
larryf8's Avatar
Philippines, NCR, QC
Joined Dec 2011
109 Posts
Quote:
Originally Posted by larryf8 View Post
Hi guys, i used to flash my Fms 20A esc with simonk firmware using Bsnfet file. Now i recently purchased the new Fms 35A esc which have different pad lay-outs now, than the previous older 20A and 30A.
Question: does anyone have any experience on this esc's regarding pad identification and file to flash it with.

Appreciate any help to point me in the information highway.

best regards,
larry
No answer yet guys? previous fms 20 and 30A escs have UBEC and now the 35A is SBEC...i've identified the pads but not sure what hex file to use. Will i use BS_Nfet hex or BS.hex?

????simonk may have the answer.

thanks,
larry
larryf8 is offline Find More Posts by larryf8
Reply With Quote
Old Oct 30, 2012, 10:57 PM
Too Many Thread Subscriptions
TheFernMan's Avatar
United States, FL, Miami
Joined Jul 2012
2,083 Posts
Quote:
Originally Posted by larryf8 View Post
No answer yet guys? previous fms 20 and 30A escs have UBEC and now the 35A is SBEC...i've identified the pads but not sure what hex file to use. Will i use BS_Nfet hex or BS.hex?

????simonk may have the answer.

thanks,
larry
You might have to cut the heat shrink and post pictures of it. Might be a different hex file all together
TheFernMan is offline Find More Posts by TheFernMan
Reply With Quote
Old Oct 30, 2012, 11:31 PM
Registered User
Australia, SA, Renmark West
Joined Sep 2011
399 Posts
Quote:
Originally Posted by simonk View Post
.......


The sync loss avoidance does turn off power but only for about 12 degrees, which is a very short time at normal hover speeds. It should only ever happen at full or nearly full throttle, where the motor is saturating and we are otherwise unable to see the zero crossing, and it is difficult to see even on the oscilloscope. I suspect something else is going on, like something else resonating. Can you maybe get a video of the issue?



..
Sorry, no video but further testing was still showing problems with uncommanded power reductions in the MT4008-12s, accompanied by a sort of "wok wok wok" noise, with a frequency of 10 hz or so. That made me suspect an interaction between the WK-M and the flashed F30A.

I replaced the F30As with the original set of TigerMotor T40As and immediately hit a snag - during the first flight test (and I ALWAYS flight test by hovering at 1 meter over grass for 5 minutes) a motor stalled on yaw command and the octo went in on its side (no damage). Since then I've found that the default settings of the TigerMotor ESC's are not suitable for the TigerMotor MT4008-12, the timing has to be set "high", which I've done. The motors now run OK and I think airframe oscillation can be eliminated as a possibility. but now I'm missing the immediate response of the SK firmware.

One problem is that in bench running it's not possible to manually throttle up as fast as the Wookong does when it calls for differential rpm on the motors so the bench runs have shown no issues - I've noticed that an abrupt yaw command from the pilot (or yaw correction from the FC) is the manouvre most likely to show up a throttling issue.

The motor is clearly very fussy about timing and I have to say that the SK Flashed F30As didn't actually get to a stall like the TigerMotor ESC did so the technique of waiting 12 degrees did work to that extent but it may be happening repeatedly, causing the motor to lose power just when more is needed. Sudden uncommanded power reductions like that are very dicy on a multicopter. I'm wondering if I can recompile the hex for the F30As with higher default timing - or am I barking up the wrong tree here?

Edit:

I've had more sudden power losses during yaw manouvres using the TigerMotor ESCs so I really don't know what's wrong now - I'd love to be able to use the iPower 2814Qs which are completely untemperamental but they just aren't powerful enough at 7kg AUW
RedSky is offline Find More Posts by RedSky
Last edited by RedSky; Oct 31, 2012 at 02:11 AM.
Reply With Quote
Old Oct 30, 2012, 11:44 PM
Registered User
Joined Oct 2012
224 Posts
hi simonk

I use brushless esc hoppyking blueseries 30A (usd code bs_nfet.hex of multicopter)





can you help me change the program like this

1 neutral position pulse 1.5ms

2 forward pulse 1.5ms to 2.0ms (pwm 1.5min and 2.0 max ) , reverse pulse 1.5 ms to 1.0 ms (pwm 1.5ms min and 1.0ms max )

3 brake ...when running forward, move the joystick in the opposite direction, then esc brake motor. continue to move joystick in the opposite direction, the motor runs reverser.


and complete file hex

If the request is complex, you can change the program in your own opinion
thank you very much and look forward to your return soon
fantasy1988 is offline Find More Posts by fantasy1988
Last edited by fantasy1988; Oct 31, 2012 at 12:01 AM.
Reply With Quote
Old Oct 31, 2012, 12:07 AM
James not bond
jy0933's Avatar
United States, MN, Minneapolis
Joined May 2012
2,074 Posts
can some1 sum up what is the possible problem when having "error during reading high fuse" at the beginning

I got a friend trying to flash red brick 30A.. but keep getting the error... double check all the pin outs... there must be something wrong.. but just couldn't get it thru..

thx

the pic of esc..
jy0933 is offline Find More Posts by jy0933
Reply With Quote
Old Oct 31, 2012, 01:06 AM
Registered User
larryf8's Avatar
Philippines, NCR, QC
Joined Dec 2011
109 Posts
Quote:
Originally Posted by TheFernMan View Post
You might have to cut the heat shrink and post pictures of it. Might be a different hex file all together
Thanks fernman, will do that...now where is that camera with macro. obviously a cam pic might not show what you need. I'll take pics of front and back showing the layout and the mosfets at the back.

Gracias,
larry
larryf8 is offline Find More Posts by larryf8
Reply With Quote
Old Oct 31, 2012, 01:36 AM
Master of Flash
LazyZero's Avatar
Alsdorf/Eifel, Germany
Joined Jun 2005
1,554 Posts
Quote:
Originally Posted by jy0933 View Post
can some1 sum up what is the possible problem when having "error during reading high fuse" at the beginning

I got a friend trying to flash red brick 30A.. but keep getting the error... double check all the pin outs... there must be something wrong.. but just couldn't get it thru..

thx

the pic of esc..
I know that RedBrick 25A has locked reset pin, which makes flashing with normal means impossible, since some months.
The spread sheet form Tom doesn't indicate that, maybe this should be added like for the HK SS

Best Christian
LazyZero is offline Find More Posts by LazyZero
Reply With Quote
Old Oct 31, 2012, 04:02 AM
hacker
Canada, BC, North Vancouver
Joined Dec 2010
934 Posts
Quote:
Originally Posted by simonk View Post
Ok, I think I know what happened. Will verify and push an update if so.
Verified. Fixed and pushed to github (for now).

Quote:
Originally Posted by RedSky View Post
Sorry, no video but further testing was still showing problems with uncommanded power reductions in the MT4008-12s, accompanied by a sort of "wok wok wok" noise, with a frequency of 10 hz or so.
...
I've had more sudden power losses during yaw manouvres using the TigerMotor ESCs so I really don't know what's wrong now
I am sure they are just difficult to drive, and as usual, expensive, too. It's no surprise that the T-Motor ESCs have problems. You may have success with slightly advanced timing or if you remove the sense capacitors from the F-30A boards, since that lags the timing which will be more of a problem on motors with higher pole counts. I have some debug code that flashes an LED on demag_timeout. We could do the same on one of the MOSI/MISO/SCK pins or something if you'd like to see it visually. Or it could count them and beep the count once the motor stops. This would probably let us know if it's oscillating instead of happening only very rarely.

Quote:
Originally Posted by tascheri View Post
I would love one for an F-20... - I have been fooling w/ AVR studio but no success... Im a PIC guy LOL...
Quote:
Originally Posted by fantasy1988 View Post
can you help me change the program like this

1 neutral position pulse 1.5ms

2 forward pulse 1.5ms to 2.0ms (pwm 1.5min and 2.0 max ) , reverse pulse 1.5 ms to 1.0 ms (pwm 1.5ms min and 1.0ms max )

3 brake ...when running forward, move the joystick in the opposite direction, then esc brake motor. continue to move joystick in the opposite direction, the motor runs reverser.
Ok, attached is modified source code as an example and built .hex files for all targets. For bs_nfet, you could probably also enable COMP_PWM to get natural drag regeneration when input is lower than speed. A similar build worked in my RC-10 a while ago, but this is untested. I would still like to make neutral not brake and instead have thumb brake work on demand (do RC car remotes still have thumb brakes? ).
simonk is offline Find More Posts by simonk
Reply With Quote
Old Oct 31, 2012, 05:23 AM
Registered User
Australia, SA, Renmark West
Joined Sep 2011
399 Posts
[QUOTE=simonk;23145102]Verified. Fixed and pushed to github (for now).



I am sure they are just difficult to drive, and as usual, expensive, too. It's no surprise that the T-Motor ESCs have problems. You may have success with slightly advanced timing or if you remove the sense capacitors from the F-30A boards, since that lags the timing which will be more of a problem on motors with higher pole counts. I have some debug code that flashes an LED on demag_timeout. We could do the same on one of the MOSI/MISO/SCK pins or something if you'd like to see it visually. Or it could count them and beep the count once the motor stops. This would probably let us know if it's oscillating instead of happening only very rarely.



....QUOTE]

I managed to get a 5 minute flight in, with the octo bucking and swerving every time I tried to yaw it amd randomly at other times too and after it landed I felt all the motors and ESCs for heat build up - all the ESCs were cool The Tigermotor ESCs run a lot cooler than the F30As but nos 4 and 7 motors were almost too hot to touch - all the others were barely warm, even though it was in an air temperature of 35C. At least I have an idea where the problem is, even if i don't know what it is. I'll check the throttle ranges and progamming on those two ESCs as a first move.

Anyway, as I still have problems on the Tiger ESCs, I don't think any significance can be read into stability issues on the flashed F30As so their ability to handle the MT4008 is s still an open.question.
RedSky is offline Find More Posts by RedSky
Reply With Quote
Old Oct 31, 2012, 09:26 AM
Registered User
larryf8's Avatar
Philippines, NCR, QC
Joined Dec 2011
109 Posts
Fms 35A esc

Hi guys/simonk, here are pictures of the newer fms 35A esc with an atmel chip:




..and the hengli 48-22 690kv pancakes that it will power up.


Please advise best hex file to use to flash the esc's. Previously, on my Y6 and another quad, the Bs_nfet file was used on the fms 20A and 30A which shared the same pad arrangement. And also it was with UBEC. Now since the 30A was discontinued by fms, they replaced it with 35A and it's now SBEC and additional mosfets.

Any help here???
best regards,
larry
larryf8 is offline Find More Posts by larryf8
Reply With Quote
Reply


Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Discussion Rctimer 30A esc cutting out with KK board Efliernz Multirotor Talk 2 May 26, 2011 02:20 PM
Sold FS: Tricopter Parts- HK401b, Pentium 30A ESC, rctimer 750kv motors pwu_1 Aircraft - General - Miscellaneous (FS/W) 2 Jan 18, 2011 09:58 PM
Sold RCTimer 2826-6 + ESC 30A New in packaging will trade as well Pics Added! planebldr Aircraft - Electric - Power Systems (FS/W) 4 May 25, 2010 09:54 PM
Discussion RCTimer 30A Programmable ESC hunter7193 Power Systems 8 Jan 02, 2010 02:27 PM
For Sale Several New ESC's - 10a, 20a, 30a - Reduced Callen9345 Aircraft - Electric - Power Systems (FS/W) 0 Apr 20, 2009 02:31 PM