Not flying well pops up fast!

n8lbv · Post by **n8lbv** » Fri Sep 19, 2014 5:35 pm

I just built my second quad based on a hobbyking 328P board 10X4.5 props and a 450 size frame.
My first took me about a month to figure out and get working and was a good flying success.

I just finished my second one and have not been able to get it to fly right.
I'm asking here what you might check or do first in this situation.
I am starting out with all default settings, I also tried higher gain values around 6.5 instead of 3.3 as this worked extremely well on my first build.

Problem I'm having now is immediately after bringing it up into a hover it starts to accellerate UPWARD quickly and last try it came right at me.
If I hold it in my had and try.. gains feel good/prooper.. no oscillating or shaking. however just after barely raising it above idle all 4 props speed up A LOT and on their own..
not from a stick command.

Idle is good and at the proper speed, if I raise it just above idle even just a tiny bit it starts to speed way up on it's own.
This is what's causing it to pop up 10 feet off the ground right after I barely lift it off.
Once in the air I can only kill it back to idle I have no or very little vertical control.
Just seems to really want to climb out on it's own.

I've tried and retried reprogramming clearing the eeprom and reflashing multiple times..
After Each I recalibrate mag and accel. (of course)

I've also been sure to reprogram set the ESC endpoints properly.

Frustrated to say the least any ideas please help and thank you.

Arakon · Post by **Arakon** » Fri Sep 19, 2014 5:50 pm

Did you also calibrate the ESCs properly?

n8lbv · Post by **n8lbv** » Fri Sep 19, 2014 6:52 pm

To the best of my knowledge YES.
On my first build I tried two ways.
1 hook all 4 up to RX in parallel and do the traditional full throttle on powerup and drop to minimum on transmitter (endpoints set in esc(s)).
2 Run the Multiwii (2.3) Built in (*cannot fly) esc calibration routine in config.h

Both of which worked for me on my first build.

All 4 props are idling evenly and at first come up even.
In hand hover test feels normal at first but props quickly afterward start all speeding up.

Arakon · Post by **Arakon** » Fri Sep 19, 2014 6:54 pm

What motors and what is the total weight of the quad? Maybe it's seriously overpowered.

Did you balance motors and propellers? You may have strong vibrations.

n8lbv · Post by **n8lbv** » Fri Sep 19, 2014 11:20 pm

Motors: http://www.goodluckbuy.com/emax-multi-c ... hrust.html

Weight 1 pound 15 oz. Just shy of 2 Lbs. including 2200 mAh 3s Lipo.

Balance? no I didn't balance anything.. nor did I on my other quad that flys perfectly.
Same motors same props same weight.
This thing idles very evenly no noticeable vibrations when holding or on a table.
When giving just a little bit of command to lift seems that it on it's own continues to keep lifting harder and pops up about 10 feet cannot hover it have it just wants to keep lifting more & more.
and quite quickly have to save it by killing the power and crashing in the grass.

I've flown many quads with unbalanced chipped bent and very unbalanced props.. and broken cracked parts which would cause severe vibrations.
They can be challenging and usually don't lift evenly but I've never experienced this ridiculous wanting to lift away stuff before. I can't even bunny hop it.

Bbergovoy · Post by **Bbergovoy** » Sat Sep 20, 2014 12:22 am

calibrate the radio? and then recalibrate the model radio the MP... also check out the min throttle settings, and max settings...

n8lbv · Post by **n8lbv** » Sat Sep 20, 2014 2:37 pm

I've calibrated and re calibrated everything.
It doesn't seem to be an issue on throttle from the tx..
If I go just barely above idle the board kicks in with gain or something and the props all start climbing on speed by themselves with no further throttle input..
at idle everything is just fine,, and just above idle fine for a fraction of a second until they all start speeding up on their own.

brewski · Post by **brewski** » Sat Sep 20, 2014 10:28 pm

What are your AUX switch settings? Check that you don't have Baro enabled as it takes over throttle.
If you have TX ranges set correctly then set some expo on throttle in WinGUI or MWConfig to tame down throttle.

capt · Post by **capt** » Sun Sep 21, 2014 2:43 am

Sounds like baro is active.
How about a GUI screen shot after it s armed?

n8lbv · Post by **n8lbv** » Sun Sep 21, 2014 5:39 pm

I will get a screen shot. but no way *should* the baro be active as I am using all defaults after clearing eeprom and all. Should be gyro only.

n8lbv · Post by **n8lbv** » Sun Sep 21, 2014 11:42 pm

I'm not sure what this screenshot is good for but I am thankful you are trying to help!
I'm already aware that I'm keeping things simple as possible and NOT in baro or any other mode except gyro (agile) during initial testing and flying setup.
I'm incredibly frustrated that it's not working and how I have not a clue what is actually causing the quick climb-ups when trying to put it into a simple hover or bunny hop just inches above the ground.

Here is a screenshot while

: armed motors turning

armed motors turning.

waltr · Post by **waltr** » Mon Sep 22, 2014 1:21 am

Very odd. The only ting I can think of is set it up as for "tuning on a string" so that it is captive but free to pitch/roll/level. This way you can have the USB connected and watch what is going on in the MWconfig.exe. The screen shot looks good.

Since you have loaded the "CAL NO FLY" sketch and then the re-loaded with this commented out the code should be ok.

Vibration tends to decrease power so doubt this is the issue.
About 900gram AUW and 3S. What motor/ESC/props?

Even overpowered quads can hover and hold altitude.
Maybe some Expo on the throttle?

n8lbv · Post by **n8lbv** » Mon Sep 22, 2014 2:40 am

Motors: EMAX MT2213-935KV see link above if needed.
Esc is Quad Brain 25Amp per.
I've even double checked the programming on each ESC section for mid-timing and Lipo battery type.
Props are 1045s

capt · Post by **capt** » Mon Sep 22, 2014 6:45 am

The GUI looks OK.
I'd look again that ESC calibration procedure, seems to be a lot "issues" in the reviews. Do you have another ESC to test with and rule it out?

brewski · Post by **brewski** » Mon Sep 22, 2014 6:55 am

If you connect ESC direct to throttle on RX does motor increase speed in linear fashion as TX throttle moved from zero to max?

n8lbv · Post by **n8lbv** » Mon Sep 22, 2014 7:20 am

Yes connecting throttle output signal from RX direct to esc does result in nice linear and consistent response as expected.

I'm seeing a "weird" problem with sticking throttle in the multiwii software/firmware..
When I first power up the board I see "1500" in the throttle RX channel when I'm actually sending "1000" from the TX and RX to the flight board.
During this time I cannot arm the motors.
I have to bring the stick up past half way then back down.. now it is updating and I see it come down to "1000" with stick all the way down..
Now I can arm the motors.

I also see another related problem that only happens if I jerk the stick very quickly up and then back down. (HAPPENS WHEN ARMED WITH MOTORS TURNING OR NOT!)
The RX shown value will "stick" at a higher value and not update even though I am 100% positive the RX is is getting minimum throttle from the TX
which is back all the way down but the motors will continue to turn faster than idle with the stick at minimum throttle.
Moving the stick up a little causes this stuck condition. to reset and show me that it is now RXing the proper value again from the RX and it's updating again.

This is a new issue in itself and seems to be pointing to a firmware bug as I am 100%positive I do not have a problem in my TX or RX.
No sticking when direct to the ESC or to a servo.

Here's a video to it happening.

http://youtu.be/uh5TZ6K_E54

n8lbv · Post by **n8lbv** » Mon Sep 22, 2014 7:41 am

Here's another video of the same thing but instead of looking at the quad copter, it's the gui view.
I am always moving hte throttle all the way back down to zero which is "1000"
But you will see it sticking in the software at various values way above "1000" every time I move the stick quickly up and back down.
Have to move stick slow to get minimum throttle to come back again.

Half-way though the video I realize I did not have the quad copter or motor power display so I disconnected and reconnected to get it back.
That in itself is probably a bug but not critical.
I also was talking but the video capture software will not get audio from the microphone so sorry it came out silent..
I was explaining what I was doing on the controls as I was doing it.

Re-take> better video with voice/description of what's going on: http://youtu.be/MsaABFhlrkM

http://youtu.be/jvjaQib3xwc

brewski · Post by **brewski** » Mon Sep 22, 2014 8:27 am

Something is really screwed with FC or firmware running on it.
I had this same 328P based Crius v2.5 clone which came with MW2.2. I upgraded it to V2.3 & it flew perfectly. I only replaced it with Crius AIOP V2 AtMega2560 so I could run EOS Bandi MW2.3 navi b7 for advanced GPS .
Try this. Run EEProm Clear in Arduino IDE then upload clean download of V2.3. Configure for your board which should be Crius SE_V2_0.
Next ensure that you are running the MWConfig that came with V2.3 download (you can also try latest WinGUI)

n8lbv · Post by **n8lbv** » Mon Sep 22, 2014 8:59 am

I did try clearing the EERPOM twice now.
Should I try it a third time?
I did my first build on this same board and it came with 2.2 and I upgraded to 2.3
Worked perfect on another quad copter over a month ago same board.
Will try as you ask... it can't hurt.

This *is* the board I have and what I am currently set for:
#define HK_MultiWii_328P // Also labeled "Hobbybro" on the back. ITG3205 + BMA180 + BMP085 + NMC5583L + DSM2 Connector (Spektrum Satellite).

waltr · Post by **waltr** » Mon Sep 22, 2014 2:59 pm

DSM2 Connector (Spektrum Satellite).

This may be where the problem is. Is there a good ground connection between the FC and RX?
I guess you have used this RX on other copters? Yes or no? If no then try swapping for a different RX.

capt · Post by **capt** » Mon Sep 22, 2014 3:43 pm

Load version MWC 2.2 and try again, if good then load back 2.3
This will also clear the EPROMM.

I had an issue where my motors armed but would not spin up until half throttle on 2.3 and a Flyduino mega board. I had to reload 2.2 and the back to 2.3 and all is good. Might be a bug?

n8lbv · Post by **n8lbv** » Mon Sep 22, 2014 7:45 pm

Pulling out my hair!
I really really want to fly this thing but instead am dealing with really stupid stuff!

1. completely cleared the EERPOM
2. checked good solid ground connection between RX and FC no problems here.
3. voltages good stable.
4. RX in proper working order tests good direct on esc.
5. Do not have a spare esc but this problem is not the esc it appears to be software in the FC.

6. Downloaded fresh untouched Multiwii 2.2
7. edit only the one line to define my board (HK 328P one).
8. Edited the line that calls for internal pullup resistors as per HK instructions for this board.

9. Flashed board with 2.2

10. calibrated mag
11. Calibrated ACC

Still have exact same behavior with throttle "sticking" as shown in the videos.

My head hurts.

http://youtu.be/c7AihnLbw1E

brewski · Post by **brewski** » Mon Sep 22, 2014 9:13 pm

You said you had same FC in first quad. Try this FC with V2.3 in your second build. Also as Waltr pointed out it could be an issue with PPM so try direct connections instead to see if this resolves throttle issue.

n8lbv · Post by **n8lbv** » Mon Sep 22, 2014 9:25 pm

The working board is 3 days away so I am unable to get my hands on it any time soon.
You lost me on the PPM item.
I'm not using a sat. RX it's a direct PPM connection to a conventional RX.
Thanks.

brewski · Post by **brewski** » Mon Sep 22, 2014 11:21 pm

n8lbv wrote:The working board is 3 days away so I am unable to get my hands on it any time soon.
You lost me on the PPM item.
I'm not using a sat. RX it's a direct PPM connection to a conventional RX.
Thanks.

Disconnect the PPM & hardwire flight & AUX functions from RX to FC.

n8lbv · Post by **n8lbv** » Tue Sep 23, 2014 12:05 am

And then what?

Also I'm sure hwat you mean but think you mean disconnect everything from the FC except throttle.

OK I did that only thing between the RX and FC is a 3 wire servo cable.

And throttle value still sticking in this configuration.

brewski · Post by **brewski** » Tue Sep 23, 2014 12:13 am

n8lbv wrote:And then what?

Also I'm sure hwat you mean but think you mean disconnect everything from the FC except throttle.

OK I did that only thing between the RX and FC is a 3 wire servo cable.

And throttle value still sticking in this configuration.

What I am saying is don't use PPM. comment out PPM in Config.h & connect throttle, aileron, rudder etc by direct cables. Your throttle issue can be due to PPM setup & doing parallel instead of serial PPM connection will prove this.

n8lbv · Post by **n8lbv** » Tue Sep 23, 2014 11:32 am

Thanks!
Bit of a newb here and needed that clarification.

n8lbv · Post by **n8lbv** » Tue Sep 23, 2014 11:36 am

And sorry I'll need further clarification:
Do you mean to specifically comment *these* lines out?

/* The following lines apply only for specific receiver with only one PPM sum signal, on digital PIN 2
Select the right line depending on your radio brand. Feel free to modify the order in your PPM order is different */
//#define SERIAL_SUM_PPM PITCH,YAW,THROTTLE,ROLL,AUX1,AUX2,AUX3,AUX4,8,9,10,11 //For Graupner/Spektrum
//#define SERIAL_SUM_PPM ROLL,PITCH,THROTTLE,YAW,AUX1,AUX2,AUX3,AUX4,8,9,10,11 //For Robe/Hitec/Futaba
//#define SERIAL_SUM_PPM ROLL,PITCH,YAW,THROTTLE,AUX1,AUX2,AUX3,AUX4,8,9,10,11 //For Multiplex
//#define SERIAL_SUM_PPM PITCH,ROLL,THROTTLE,YAW,AUX1,AUX2,AUX3,AUX4,8,9,10,11 //For some Hitec/Sanwa/Others

// Uncommenting following line allow to connect PPM_SUM receiver to standard THROTTLE PIN on MEGA boards (eg. A8 in CRIUS AIO)
//#define PPM_ON_THROTTLE

brewski · Post by **brewski** » Tue Sep 23, 2014 12:14 pm

Make sure all serial sum PPM lines inc Throttle have // in front & connect cables to all ports from RX to FC individually.

n8lbv · Post by **n8lbv** » Wed Sep 24, 2014 12:41 pm

What you see above is every line I have that involves "PPM" and they are already all commented out.

My hardware setup right now is simply a single 3wire throttle servo cable between the RX and the FC.
FC is getting 5V supply from that cable.
RX + FC is getting good clean 5V from ESC.

I am still showing the throttle "sticking" problem as seen in the video.

brewski · Post by **brewski** » Wed Sep 24, 2014 8:38 pm

If you only have the one servo cable between FC & RX how are you controlling other flight functions without using PPM?

n8lbv · Post by **n8lbv** » Thu Sep 25, 2014 3:40 am

Seriously dude?
I'm NOT! :-)

I'm simply testing the throttle channel at this point and if it still sticks in the software (multiwiiconf).

And it does.
It's on a test bench.
This thing has not been in the air since I started this thread.
And I'm seriously bumming because of that.

brewski · Post by **brewski** » Thu Sep 25, 2014 5:11 am

n8lbv wrote:Seriously dude?
I'm NOT!

I'm simply testing the throttle channel at this point and if it still sticks in the software (multiwiiconf).

And it does.
It's on a test bench.
This thing has not been in the air since I started this thread.
And I'm seriously bumming because of that.

OK got ya! Myself & others automatically assumed you were using PPM with only one servo cable between FC & RX.
You say you are sending 1000 at minimum throttle & it is showing as 1500 (half throttle) in WinGUI. Are you running correct WinGUI for firmware version installed in FC?
If not you can get crazy results & even corrupt the FC.

Noctaro · Post by **Noctaro** » Thu Sep 25, 2014 9:07 am

Hi,
have you tried to connect roll or yaw as throttle and watched if behaviour remains? (i think we need to know if it depends to the signal port of your FC or RC)
Just do it without props

Greetz Noc

n8lbv · Post by **n8lbv** » Thu Sep 25, 2014 2:39 pm

Awesome thanks for helping out guys!
I really appreciate the help!
Now to answer you:

On bootup it shows 1500 *until* I move throttle stick up and back down then it shows "1000" (minthrottle) at minimum stick.
I detail this very well i nthe video above where I'm talking while showing how on bootup it starts at 1500 (apparently ignoring any throttle input) until the throttle is changed.
Soon as you move the throttle at all it 'wakes up' and begins showing live data from the throttle input.
In answer to the other question I can certainly try what you ask but I have determined via other testing that the RX or the TX are not doing the sticking, it's happening only on the FC.
My RX truly is sending it's minimum signal to the throttle input on the FC when it sticks or when I bootup and it's showing 1500.
I also hooked a digital servo and one of the ESC inputs directly to the RX for testing (two separate tests) and there is no sticking at all whatsoever when going back to min throttle.
There also is no problem with the TX or RX sending anything but minimum throttle at all times when the stick is at minimum or when anything is booted/powered on.
I am confident in that and I know why you are asking me to try a different channel/signal into the throttle input on the FC.
Just a basic and good sanity check.
I will do this.

waltr · Post by **waltr** » Thu Sep 25, 2014 3:53 pm

You seem to have checked everything and it is looking to be the FC board. Swap that next.

n8lbv · Post by **n8lbv** » Thu Sep 25, 2014 8:39 pm

Yeah bummer eh?
It seem so much like a software issue (sticking in software) and not like a hardware issue.
But yeah I'll have to try another board soon.

:-(

waltr · Post by **waltr** » Fri Sep 26, 2014 3:58 am

I have used MW v2.3 on four different FC boards and never saw anything like this.

There was only time I believe I had a code issue, the quad would be flying fine then just freak out. Dis-arm, re-arm and it would fly again for 2 minutes then freak out again. I run the clear_EEPROM sketch and then re-loaded MW and never saw the problem again on that board. I chalked this up to a corrupted bit in the FLASH or EEPROM.

What you have could be a bad FLASH memory or a bad Atmega timer, bad interupt logic (used to time the PWM pulses from the RX) or a number of other hardware failures (within the Atmega processor). This will look like bad code but its really bad hardware.
If the new FC board fixes the problem then it must be hardware.

brewski · Post by **brewski** » Fri Sep 26, 2014 11:49 am

waltr wrote:I have used MW v2.3 on four different FC boards and never saw anything like this.

There was only time I believe I had a code issue, the quad would be flying fine then just freak out. Dis-arm, re-arm and it would fly again for 2 minutes then freak out again. I run the clear_EEPROM sketch and then re-loaded MW and never saw the problem again on that board. I chalked this up to a corrupted bit in the FLASH or EEPROM.

What you have could be a bad FLASH memory or a bad Atmega timer, bad interupt logic (used to time the PWM pulses from the RX) or a number of other hardware failures (within the Atmega processor). This will look like bad code but its really bad hardware.
If the new FC board fixes the problem then it must be hardware.

Even more so with the cheap HK boards & clones..as they say you get what you pay for!

brandon3055 · Post by **brandon3055** » Sat Sep 27, 2014 2:08 am

Just a thought from a noob but maby its overpowered? I ask because i am currently having that problem with my quad and the symptoms sound similar. When i increase throttle just above idle it jumps into the air and ether wants to rise like crazy or if i reduce throttle drop like a rock. I tried a adding a 1kg of weight to it which increased its total weight to 2.5gk (5.5lb) and it flew beautifully. So maby try adding a little (or a lot) extra weight and see how it effects it.

Edit: turns out I just needed some expo on the throttle (added through the GUI not the transmitter) but I'm not shure if that would help in your case

n8lbv · Post by **n8lbv** » Sat Sep 27, 2014 7:20 am

Nah, not overpowered the other one I built and sent away was in the same weight area and had the same board motors esc and props and it flys like a champ.

This throttle value sticking in the FC (software) issue needs to get resolved first because it simply should not be doing that.
Getting this taken care may also fix the other problem.

n8lbv · Post by **n8lbv** » Sun Sep 28, 2014 6:29 am

Okay so I disconnect my FLYSKY RX and put on a generic Orangerx spektrum compatible RX an the sticking throttle problem is gone.
*headscratch*.
So there's something the FC does not liek about the FLysky RX *sigh* but what?
The Flysky recievers work great and I want ot use it because it's 9CH and I can use the extra channels for some really fun stuff on the quad.

I'm also positive the sticking throttle problem is on in the RX or Flysky TX module.

This is really weird.

weird weird weird freaking wierd.

Anyhow I'm on to something.

n8lbv · Post by **n8lbv** » Sun Sep 28, 2014 6:36 am

It's actually one of these: http://www.xheli.com/79p-r8b-receiver.html

I've had not problems at all with these on my CP helicopters, and if I hook up a servo (or my quad esc directly to the throttle output channel)
There is smooth completely control min to max with NO sticking of any kind.

capt · Post by **capt** » Sun Sep 28, 2014 8:13 pm

The wonderful world of electronics. Some things just don't play well together, cheap can be hit or miss.

n8lbv · Post by **n8lbv** » Sun Sep 28, 2014 9:19 pm

Yep, but I need to figure out why..
Doesn't make sense yet.
Those flysky radios are actually decent full range radios I've had nothing but excellent performance and long range from them.

It's like it's sending a lower than 'normal' minthrottle number or *something* that's upsetting the FC.
Again no sense here yet.

I'll have to try things.

n8lbv · Post by **n8lbv** » Mon Sep 29, 2014 3:06 am

Once I figured out that using a different RX "fixes" the problem I decided to start another thread over here to try and deal with the
FlySky RX over here: viewtopic.php?f=15&t=5635

My mind is blown I'm stumped and confused and frustrated as to why that RX would not work properly.
Meanwhile maybe I'll try and fly this thing on the working RX and see if that also takes care of the original problem.

Will report back here with results soon.