Autoquad M4 meets NYLON-quad (RCTimer Spider)

Show us your flight videos and pictures of your aircrafts

Re: Autoquad M4 meets NYLON-quad (RCTimer Spider)

Postby brat002 » Fri Jun 26, 2015 7:35 pm

chschmid wrote:
brat002 wrote:And, may be you can answer, WHY all these people https://www.youtube.com/results?search_ ... overy+naza are flying on flappable TBS frames without serious visible issues? It is just practical interest for me.


That's the problem. The issues are not visible. But the sensors see them. And may go crazy if the issues are too severe.
In Jörgs vid you can hardly see that something is wrong. In the log you can.
https://www.youtube.com/results?search_query=tbs+discovery+naza+crash
Cheers
Christof

Agreed with you, but TBS+Naza are very popular combo for many people. Naza much more resistent to vibrations, that may be on frames with DJI arms.
I expected comparable vibration resistance in AQ, however in AQ implements other behaviour.
brat002
 
Posts: 191
Joined: Tue Apr 21, 2015 10:35 pm

Re: Autoquad M4 meets NYLON-quad (RCTimer Spider)

Postby Patnet » Fri Jul 03, 2015 7:29 pm

Hey Joerg, please let me know if you make any PID changes before you go to Quatos.

My current PIDs are attached (similar frame) if you're interested. My best CTRL_TLT_ANG_D is 5500.
Attachments
revised_params_2.txt
(14.83 KiB) Downloaded 210 times
Patnet
 
Posts: 138
Joined: Thu Jan 29, 2015 10:24 pm

Re: Autoquad M4 meets NYLON-quad (RCTimer Spider)

Postby aBUGSworstnightmare » Sun Jul 05, 2015 7:13 am

I've made some PID tunings yesterday and here is the final parameter set.
PARAMS_spider_20150705_0r1.txt
(2.48 KiB) Downloaded 201 times

I've balanced the props to test if I can reduce the overall vibration level, but due to the bad motors this only showed a minor effect.
Here are some shots from the logs:
imuaccs.jpg
High vibration level is caused by bad motors

motor out.jpg
Motor output with PIDs posted above. The aft motors need to work harder because of the bad weight distribution (M5 =rear left; M6= rear right)

Tested some waypoint flight (mission mode) and the result was pretty good (Mission mode PIDs are all default).

Conclusion of the PID version:
- use good motors and balance your props
- add additional standoffs (37mm when using DJI/DJI clone arms) to the frame to make it stiffer.
- without a Camera in front the weight distribution of the craft is very bad whit the battery in the aft compartment (in-between the frame plates) --> try to mount it on top over the FC (maybe requires a GPS antenna with a longer lead)
- active freewheeling of the ESC32V3 has been disabled

I will proceed with Quatos testings on the craft (without making any of the mentioned changes for sure)!
aBUGSworstnightmare
 
Posts: 1460
Joined: Fri Jun 22, 2012 5:24 pm

Re: Autoquad M4 meets NYLON-quad (RCTimer Spider)

Postby aBUGSworstnightmare » Sun Jul 05, 2015 11:54 am

For QUATOS we need:
- Quatos XML-file of craft
- Thrust curve of motor+prop

O.K. here's are some files:
rctspider_M4CAN_0r1.pdf
RCTSpider - visualization of .XML-File for QUATOS
(42.87 KiB) Downloaded 230 times

Thank's again to Fred for his handy xml-to-vrml-tool!

PARAMS_esc32v3_TGY2215J_GF1045.txt
ESC32V3 calibration parameters for Turnigy 2215J with GemFan 10x45 props
(1.12 KiB) Downloaded 189 times

R2V_esc32v3_Turnigy2215J_1045_3S.jpg
Result of ESC32 calibration

thrutscurve_esc32v3_TGY2215J_GF1045.JPG
Thrust curve measurement result of Turnigy 2215J+GF1045

thrustTorqueCalc_Turnigy2215J_1045_3S.xls.zip
Measurement results for A1/A2 calculation
(6.22 KiB) Downloaded 191 times


Another downside of this setup is it's AUW! It is 1066g with the current components and the Turnigy LiPo. Unfortunatley you will need a QUATOS hobby license for it.
But, let me see how it performs! I have another set of motors and an lighter LiPo in mind which should allow to use it with the free-of-cost QUATOS license.
aBUGSworstnightmare
 
Posts: 1460
Joined: Fri Jun 22, 2012 5:24 pm

Re: Autoquad M4 meets NYLON-quad (RCTimer Spider)

Postby aBUGSworstnightmare » Wed Jul 08, 2015 7:29 pm

Made the maiden with QUATOS today.


Log file analysis and parameters (together with the .xml-files I've used) will come later.
Video shows radio signal loss at the end!
Last edited by aBUGSworstnightmare on Sun Jul 12, 2015 6:25 pm, edited 1 time in total.
aBUGSworstnightmare
 
Posts: 1460
Joined: Fri Jun 22, 2012 5:24 pm

Re: Autoquad M4 meets NYLON-quad (RCTimer Spider)

Postby LPR » Thu Jul 09, 2015 1:44 am

Joerg

It flies like a Quatos quad. Very nice.

I'm wondering with the two F parameters so low when you were using PID FW, did you see a problem with control in wind?

Larry
LPR
 
Posts: 1323
Joined: Tue Jun 26, 2012 1:29 pm
Location: MN, USA

Re: Autoquad M4 meets NYLON-quad (RCTimer Spider)

Postby aBUGSworstnightmare » Thu Jul 09, 2015 6:13 am

LPR wrote:Joerg

It flies like a Quatos quad. Very nice.

I'm wondering with the two F parameters so low when you were using PID FW, did you see a problem with control in wind?

Larry


Hi Larry,
thanks! Flies pretty crisp on first attempt. I'm even more surprised that I did not need to adjust the throttle factor. Looks like I did a good job with my kitchen scale and the .XML-file 8-)
Well, I decided to go that low on F-params due to the high vibration level of the craft. I did not see and problems in wind conditions so far, but I need to say that PID tuning will require quite some effort on this machine. That's why I decided to keep the parameters 'as good for now' and proceed with QUATOS.
Will now look into the setup to see if I can get rid of 67g --> the new AUW would be 999g --> you can fly it with the FOC (included) QUATOS license.
Joerg

P.S. I will bring this one with me on the 18th! Let's see if we can improve it.
aBUGSworstnightmare
 
Posts: 1460
Joined: Fri Jun 22, 2012 5:24 pm

Re: Autoquad M4 meets NYLON-quad (RCTimer Spider)

Postby aBUGSworstnightmare » Sun Jul 12, 2015 6:24 pm

Took here out today for some flying in windy conditions.


Next steps: IMPROVEMENTS! Need to get rid of the vibrations!

So, what's next:
- change motors
- test 2 prop setups (2x for 3S, 1 for 4S (eclac calculates a top speed of 70km/h for the 4S setup 8-) :lol: )
- change nylon to alu arms
- add GPS ground plane
...
let's see what this will result in ...
aBUGSworstnightmare
 
Posts: 1460
Joined: Fri Jun 22, 2012 5:24 pm

Re: Autoquad M4 meets NYLON-quad (RCTimer Spider)

Postby LPR » Sun Jul 12, 2015 9:22 pm

Joerg

I found that in RTH that my quad will not yaw to the Home position yaw angle if I adjust the yaw and then quickly use RTH.

The more vibration the M4 has coming in the longer you have to wait after you change the yaw angle and then use RTH. With bad vibrations you need to wait 10 seconds to have the yaw angle change to the home positions yaw angle. With very little vibrations you only need less than a second after you change the yaw angle and than switch to RTH.
When I switch to RTH the MR will start moving immediately regardless of vibration levels but not yaw if
there is vibration.

Larry
LPR
 
Posts: 1323
Joined: Tue Jun 26, 2012 1:29 pm
Location: MN, USA

Re: Autoquad M4 meets NYLON-quad (RCTimer Spider)

Postby aBUGSworstnightmare » Sat Aug 08, 2015 1:50 pm

Hey guys,
had some time today to take care of the 'MKII' version of this demo craft.
MKII means:
- better motors
- exchange nylon arms against alu arms
- ...

I will use the T-MOTOR AIRGEAR350 set, so I made measurements of the stock set at 3S and 4S. In addition, I've measured a T-MOTOR CFK 11x3.7 prop on the AIR2213.
AIR2213-thrust curves.jpg
Thrust curves of the AIR2213 with different props/LiPo voltages. The yellow curve is the TGY2215J which I currently use on the craft (for reference only).

In case of efficiency the CFK 11x3.7 and the stock (plastic) T9545 are very (!) close (you need to analyze the log data for this) when assuming an AUW of 1450g (first guess; assuming craft weight of 900g + 550g for battery - needs to be reviewed when real weight is known).

All calibrations/measurements were made on ESC32V3 --> find the results below 8-)
esc32v3_air2213_cfk11x37_3s.txt
ESC32V3 calibration data for AIR2213/920 with T-Motor CFK 11x3.7 @ 3S
(198 Bytes) Downloaded 176 times
thrustTorqueCalc_AIR2213_CFK11x37.xls
Thrust measurement - AIR2213/920 with T-Motor CFK 11x3.7 @ 3S
(15.5 KiB) Downloaded 221 times

------
esc32v3_air2213_t9545_3s.txt
ESC32V3 calibration data for AIR2213/920 with T-Motor T9545 @ 3S
(199 Bytes) Downloaded 164 times
thrustTorqueCalc_AIR2213_T9545_3S.xls
Thrust measurement - AIR2213/920 with T-Motor T9545 @ 3S
(16.5 KiB) Downloaded 230 times

------
esc32v3_air2213_t9545_4s.txt
ESC32V3 calibration data for AIR2213/920 with T-Motor T9545 @ 4S
(201 Bytes) Downloaded 176 times
thrustTorqueCalc_AIR2213_T9545_4S.xls
Thrust measurement -AIR2213/920 with T-Motor T9545 @ 4S
(18 KiB) Downloaded 249 times

------

Updating the .xml file is next.
Some words on the expected AUW!
I plan to use the craft with different LiPos (2x 2700mAH/3S or 5000mAH/4S). The LiPo and prop selection is then related to the style I want to fly: 4S for FAST flying, 3S (maybe with the CFK props) for long crusing.
Here are some ecalc figures to compare:
AIR2213_CFK11x37.jpg
ecalc result for CFK 11x3.7

AIR2213_T9545_3S.jpg
ecalc result for T9545 @ 3S

AIR2213_T9545_4S.jpg
ecalc result for T9545 @ 3S


--- choose your weapons .... (and don't tell me that 4S will be fast 'n long!)
aBUGSworstnightmare
 
Posts: 1460
Joined: Fri Jun 22, 2012 5:24 pm

PreviousNext

Return to AQ Media Showcase

Who is online

Users browsing this forum: No registered users and 2 guests

cron