Electrohub Spider + Flip 1.5 Problems
- Thread starter NHS77
- Start date
I don't think so, no.
Assuming you've not unplugged any of the ESC connectors from the flight controller, that's just really odd behavior. You've had it up in the air before, but this time, it won't fly without flipping.
I looked at the config.h, and it was interesting that there are no special motor mix configurations for a spider quad setup. I almost expected to see a special setup for that. Similarly, I don't see anything too specific for the DX5e TX. This all looks fairly similar to a stock quadX flip 1.5 setup. Maybe just the MINTHROTTLE is different. I edited that online version with a value that you indicated is when the motors all started. You should double check that before using it. What this value does, is ensure that your motors won't stop when you are at an extreme orientation, by ensuring the flight controller won't go below this signal for minimum throttle.
I don't think this is your problem, unless you were doing a full pitch / roll command request on takeoff, which I doubt to be the case!
#define MINTHROTTLE 1250
is what I set it to.
Anyhow, you may want to take the props off, and fully bench test this while connected to the computer and powered by a fully charged lipo. Put tape on the motor shafts so you can see what direction they spin, and how fast. Then run the following tests:
1) Arm, throttle up, and verify all motors are spinning in the proper direction, and spin up at the same time -- just a sanity check on your setup and ESC calibration
2) Arm, Throttle up to just at the point where the motors spin, while quad is on level surface.
- observe what happens when you tilt the quad TOWARDS each motor (e.g. dip the right, left motor down) -- each motor you dip down should speed up, but no other motors should completely stop. This will test that your motors/esc's are plugged into the correct port on the flight controller, and that the board orientation is correct. It will also show that your MINTHROTTLE value is ok.
3) Arm, throttle up to mid throttle, and pitch the quad forward and back ~45, and pause at -45, 0, and 45. Watch the motor PWM level on the GUI to see if the front motors speed up/rear motors slow down when at -45, all motors are about equal at 0, and rear motors speed up/front motors slow down at +45.
4) Repeat #3 for roll.
If you can post your observed results, that might be helpful. Try to post the numbers you see. No need for fancy screenshots or anything.
bgfireguy92
Senior Member
DUDE! I just went into pauls files and you said that you're using a Dx5 correct? Did you load the sketch file that he setup for FT Spider with Dx5?
https://www.dropbox.com/sh/3ptxpwesovd7tx3/gYdZU3lXQh
https://www.dropbox.com/sh/3ptxpwesovd7tx3/gYdZU3lXQh
Last edited:
bgfireguy92
Senior Member
OK so mak my throttle end points going thru multiwii are 1003 at min 2001 max. But as you can see below my define min throttle is 1220. Now Im still learning and will admit Im wrong but mine flys really well. Do I need to change this?
The value of MINTHROTTLE that works for you will be different from what works for me or the OP of this thread. My MINTHROTTLE is 1100 on my tricopter build, because my motors with my ESC's, and RX/TX combination kick in at about 1090, so I added 10 to that to give me 1100...
This value should be whatever prevents the flight board from shutting down a motor at extreme orientation... when it spins motors down low and speeds others up to get to that orientation.
bgfireguy92
Senior Member
Ok gotcha. Like I said always willing to learn or change.
Sorry NHS77 didnt mean to hijack your thread.
Sorry NHS77 didnt mean to hijack your thread.
OK, going through the motions as you suggested I notice the following:
Motors start at around 1265 (all at the same time). Having the frame level at lowest throttle front right and back left motors sit at 1220 rpm while the other two motors go around 1400. No matter how I tilt the frame the two motors stay at 1220 rpm while other two react to the motion.
Going to mid throttle (or above) FR and BL motors increase RPMs but stay lower than the other two. Tilting the frame these two briefly increase (never decrease) RPMs and fall back to whatever RPM they were after half a second or so. The other tow motors behave as expected.
Double checked that my wiring and spinning direction is correct. Everything checks out.
SOMETHING is not right here...
That is very strange. Something is not right, as you say.
In my mind, there's really only one of two possible answers: a software issue or a hardware problem with the flight board. Here's what we "know":
* You've calibrated the ESC's with the calibration build of MW.
* You've modified the config.h with a valid MIDRC and MINTHROTTLE for your setup (DX5e, ESC's, motors)
* You've erased the EEPROM, and reflashed with the correct MW23 build for the DX5e
* You've been able to fly about a minute with just a persistent forward drift in the past
* No hardware changes, have verified motor directions and commands to appropriate motors
* Two motors/ESC pairs are behaving as expected on bench, two are not
What I know of motors/ESC's are: ESC"s and Motors have a voltage level based feedback to each other. ESC"s and flight controllers/RX's are generally unidirectional PWM signal based (RX/Flight board -> ESC). So, what you're seeing on the MultiWiiConf (PWM rates being sent to the ESC's/motors) is based only on sensor feedback to the software running on the board and your RC input commands.
I didn't think it was necessary, but perhaps even with MW2.3, despite being in Acro/Rate mode, you need to recalibrate the ACC after a EEPROM erase. You can use the stick commands to do so (Low throttle, full left yaw + full pitch back):
https://multiwii.googlecode.com/svn...tiWii-StickConfiguration-23_v0-5772156649.pdf
Do that while the quad is level and perfectly still. In fact, you might even recalibrate the gyros if you want before re-running those tests above.
If that doesn't work to resolve it, here's another shot in the dark (trying to remove, somewhat by brute force, some of the software variables):
I've taken the liberty of merging your MW23 DX5e Quad config with MW2.4:
https://dl.dropboxusercontent.com/u/25601637/MultiWii24_FTSpiderQuadDX5e.zip
I'm not sure why this would make a difference, as it's more of a shot in the dark, but I'm at a loss for other software options.
Just check that the MIDRC and MINTHROTTLE is ok for you. I reset what you had for MINTHROTTLE with your latest test result from above (bumping it up a bit more).
One other thing: with this build, you WILL need to use Arduino 1.6.1 (the latest) SDK from http://arduino.cc/en/Main/Software
It will result in a smaller binary file/hex file that's uploaded to the board, due to some fairly major libraries changes... so, you're not only updating the MultiWii codebase, but also the shared libraries / build environment...
Definitely do the EEPROM erase (from the new Arduino IDE) before flashing MW2.4, and recalibrate the ACC afterwards. You may notice after this, that in MultiWiiConf, it'll look like you have a compass... you still don't, but the quad might rotate around the compass because it's trying to calculate it in software.
In my mind, there's really only one of two possible answers: a software issue or a hardware problem with the flight board. Here's what we "know":
* You've calibrated the ESC's with the calibration build of MW.
* You've modified the config.h with a valid MIDRC and MINTHROTTLE for your setup (DX5e, ESC's, motors)
* You've erased the EEPROM, and reflashed with the correct MW23 build for the DX5e
* You've been able to fly about a minute with just a persistent forward drift in the past
* No hardware changes, have verified motor directions and commands to appropriate motors
* Two motors/ESC pairs are behaving as expected on bench, two are not
What I know of motors/ESC's are: ESC"s and Motors have a voltage level based feedback to each other. ESC"s and flight controllers/RX's are generally unidirectional PWM signal based (RX/Flight board -> ESC). So, what you're seeing on the MultiWiiConf (PWM rates being sent to the ESC's/motors) is based only on sensor feedback to the software running on the board and your RC input commands.
I didn't think it was necessary, but perhaps even with MW2.3, despite being in Acro/Rate mode, you need to recalibrate the ACC after a EEPROM erase. You can use the stick commands to do so (Low throttle, full left yaw + full pitch back):
https://multiwii.googlecode.com/svn...tiWii-StickConfiguration-23_v0-5772156649.pdf
Do that while the quad is level and perfectly still. In fact, you might even recalibrate the gyros if you want before re-running those tests above.
If that doesn't work to resolve it, here's another shot in the dark (trying to remove, somewhat by brute force, some of the software variables):
I've taken the liberty of merging your MW23 DX5e Quad config with MW2.4:
https://dl.dropboxusercontent.com/u/25601637/MultiWii24_FTSpiderQuadDX5e.zip
I'm not sure why this would make a difference, as it's more of a shot in the dark, but I'm at a loss for other software options.
Just check that the MIDRC and MINTHROTTLE is ok for you. I reset what you had for MINTHROTTLE with your latest test result from above (bumping it up a bit more).
One other thing: with this build, you WILL need to use Arduino 1.6.1 (the latest) SDK from http://arduino.cc/en/Main/Software
It will result in a smaller binary file/hex file that's uploaded to the board, due to some fairly major libraries changes... so, you're not only updating the MultiWii codebase, but also the shared libraries / build environment...
Definitely do the EEPROM erase (from the new Arduino IDE) before flashing MW2.4, and recalibrate the ACC afterwards. You may notice after this, that in MultiWiiConf, it'll look like you have a compass... you still don't, but the quad might rotate around the compass because it's trying to calculate it in software.
Last edited:
AdamBomb42x
New member
I had the same setup as you and I think your have the yaw rate too high, set it to .15 and lower the roll and pitch P rates to 3.0. I had made the mistake of setting the yaw rate to .5 and it behaved the same as your second video. I also had a wobble on descent which lowering the P for pitch and roll took care of. I'm a noob too, so take what I say with that in mind, but I too had an electrohub spider quad with the RTFQ bundle. I had mine flying great right before I was an idiot and lost it.
Thanks for the new MW2.4 Sketch!
Did everything as you suggested. Same result. I did a screen recording of Multiwii Conf:
https://youtu.be/OYg-7xlbId4
I started at lowest throttle. Notice how two motors sit at 1265 while the other two start to spin up slowly to above 1500 without any stick input. I then started tilting and rolling the frame. Front right and back left motors stay at 1265 while the other two react to the movement. Lastly I went to mid throttle and moved the frame again. FR and BL motors react a little but always much less than the other two... I'm not even going to attempt flying again at this point.
One more thing though: I suspected the ESC's being a problem so I went through the standard ESC calibration with each ESC hooked up individually to the throttle channel of the RX, powering up with max throttle on the TX, etc. That seemed to work fine and the interesting thing is that this way the motors kicked in after raising the the throttle just a tiny bit. They also spun much slower at the lowest throttle and I was able to increase RPM's all the way to max stick. In comparison, when the esc are hooked up to the board min throttle is much faster and I need to raise the stick to about 1/4 throw. Same at the top end, which I hit at about 3/4 stick and at least from observation and the sound of the motors they don't spin as fast at max throttle.
I'm tempted to just get another board...
You should send that video and a link to this thread to Paul: http://www.readytoflyquads.com/contacts
He should see this. Maybe he will have an idea.
That is very strange indeed.
He should see this. Maybe he will have an idea.
That is very strange indeed.
Alright, heard back from Paul - he suspects an ESC calibration issue... will look into that.
In the meantime: Having loads of fun with my first build (excl. the fact it didn't fly) I had already ordered parts for a second build incl. a Flip32+ board. So I decided to slap that on the same frame and did a basic setup in Cleanflight. I also got a DX6i to be able to dial in subtrims.
Result: She flies! https://www.youtube.com/watch?v=do4DbLPThMk
Still got the vibe and still got the forward drift, but now I can start working on improving that.
The only thing I'm wondering: I'm having quite some problems to maintain a solid hover. While that is certainly in part my poor skills I wonder whether I should dial in an expo for throttle on the TX or in Cleanflight or maybe even both? (As it is I had 0.50 on the throttle in Cleanflight but nothing on the TX)
In the meantime: Having loads of fun with my first build (excl. the fact it didn't fly) I had already ordered parts for a second build incl. a Flip32+ board. So I decided to slap that on the same frame and did a basic setup in Cleanflight. I also got a DX6i to be able to dial in subtrims.
Result: She flies! https://www.youtube.com/watch?v=do4DbLPThMk
Still got the vibe and still got the forward drift, but now I can start working on improving that.
The only thing I'm wondering: I'm having quite some problems to maintain a solid hover. While that is certainly in part my poor skills I wonder whether I should dial in an expo for throttle on the TX or in Cleanflight or maybe even both? (As it is I had 0.50 on the throttle in Cleanflight but nothing on the TX)
Similar threads
