Most multiwii issues I can figure out on my own but this one has really got me.
After a lovely flight around last week I powered down everything, removed batts and put the copter on the shelf.
Today I received a pile of new props and decided to take a set out for a test but when I was doing my pre flight checks the copter wouldn't do what it should.
Hooked her up to the GUI and found the gyro yaw value was locked over to 511 and i2c errors were mounting quickly.
If I left it running eventually all the sensors would lock up and the board would need a hard reset.
I have done absolutely nothing to the setup between the last flight and now which is why I find this so very odd.
I removed the controller from the copter so it had no plugs plgged in and it showed the exact same symptoms-within ~1min of power of one value would lock hard up/down followed by the rest soon after with i2c errors.
I then uploaded the newest DEV version but the same result-As far as i can tell it is hardware related.
I checked the board and removed the odd piece of stray dry grass although I doubt that would have any major effect.
The board in question is a hobbyking 328p (cruis se clone basicily) with I2c GPS and bluetooth. I also changed the code to remove the GPS and unplugged it, no change.
Any idea what has gone wrong?
Senors locking up and i2c errors
-
mon_lolo_fr
- Posts: 40
- Joined: Tue Nov 15, 2011 9:50 am
Re: Senors locking up and i2c errors
Hi,
I'm not having this exact type of board but flying with a 328P and MPU6050 witho0ut anything else through the I2C bus.
Last time I had a strange behaviour like yours was the day my ITG3200 fried on the board while flying
As it seems that you've got an all integrated card it'll be a little tricky to find out what's going on the I2C bus.
In fact, try to fly only with the gyro (enable only the gyro code through the configuration files (config.h) and see through the GUI what's happening then.
Laurent
I'm not having this exact type of board but flying with a 328P and MPU6050 witho0ut anything else through the I2C bus.
Last time I had a strange behaviour like yours was the day my ITG3200 fried on the board while flying
As it seems that you've got an all integrated card it'll be a little tricky to find out what's going on the I2C bus.
In fact, try to fly only with the gyro (enable only the gyro code through the configuration files (config.h) and see through the GUI what's happening then.
Laurent
-
NikTheGreek
- Posts: 348
- Joined: Thu Dec 08, 2011 4:17 pm
- Location: Greece
- Contact:
Re: Senors locking up and i2c errors
Try one sensor each time.
Comment all in defines except one every time ...this way you will see which one creating the problem.
Comment all in defines except one every time ...this way you will see which one creating the problem.
Re: Senors locking up and i2c errors
Ok tried out that method of deduction but the results weren't that conclusive....
Just the gyro - during 5 min sitting on the bench about 1700 I2c errors happened and you could see spikes in the graph whenever they happened.
Jusr Acc - About 20 I2c errors in 10 min then it locked up completely giving 100% reading on the Z axis and other weird values on y and x, roughly 2 min latter the values went back to normal but the board was still locked up and the sensor didn't respond to any movment.
Acc and gyro - After 2-3 min the whole thing went crazy again, same issue that was first observed the other day.
Lesson learnt, pay $27 for a board, get a cheap board. May just pop the 328 chip off and use it in another project.
Will stick to known brands of boards now/I'm looking into other solutions.
Naza is far to expensive, copter control is always out of stock, I don't enjoy KK that much any other options/what is a good multiwii board to look into? is 32bit that much more amazing?
Just the gyro - during 5 min sitting on the bench about 1700 I2c errors happened and you could see spikes in the graph whenever they happened.
Jusr Acc - About 20 I2c errors in 10 min then it locked up completely giving 100% reading on the Z axis and other weird values on y and x, roughly 2 min latter the values went back to normal but the board was still locked up and the sensor didn't respond to any movment.
Acc and gyro - After 2-3 min the whole thing went crazy again, same issue that was first observed the other day.
Lesson learnt, pay $27 for a board, get a cheap board. May just pop the 328 chip off and use it in another project.
Will stick to known brands of boards now/I'm looking into other solutions.
Naza is far to expensive, copter control is always out of stock, I don't enjoy KK that much any other options/what is a good multiwii board to look into? is 32bit that much more amazing?
-
NikTheGreek
- Posts: 348
- Joined: Thu Dec 08, 2011 4:17 pm
- Location: Greece
- Contact:
Re: Senors locking up and i2c errors
Hmmm.. i've read here that this kind of problems (weird readings) some times caused by BAD GROUND .
Re: Senors locking up and i2c errors
That would back up the conclusion of it being a cheaply made board if something isn't grounded properly.
Considering removing all the components from the board and just curdely wiring them up by hand, that would rule out a number of board related issues.
If I make any ground breaking conclusions I will post about it in case anyone else has similar issues.
Considering removing all the components from the board and just curdely wiring them up by hand, that would rule out a number of board related issues.
If I make any ground breaking conclusions I will post about it in case anyone else has similar issues.
Re: Senors locking up and i2c errors
@NikTheGreek thankyou for sparking an idea in my head.
The only way a bad ground would occur is a bad/non/dry joint on the board somewhere. Looking at the joints they all seemed a bit dull and crappy ($27 board...) and there was noway I could go through and redo ever one manually.
2KW of heatgun and some melted header pin plastic bits later and the board is all re-flowed. During the process I noticed some joints had a fair bit of soldier move around, most notably around the 3.3v regulater - at least I'm pretty sure that's the chip I mean.
Anyway after it cooled down I hooked her up and have had the board running for a solid hour now with absolutely no issues.
Not 100% sure what exact component was the cause but were ever it was it had a bad joint.
Thanx for the help guys.
The only way a bad ground would occur is a bad/non/dry joint on the board somewhere. Looking at the joints they all seemed a bit dull and crappy ($27 board...) and there was noway I could go through and redo ever one manually.
2KW of heatgun and some melted header pin plastic bits later and the board is all re-flowed. During the process I noticed some joints had a fair bit of soldier move around, most notably around the 3.3v regulater - at least I'm pretty sure that's the chip I mean.
Anyway after it cooled down I hooked her up and have had the board running for a solid hour now with absolutely no issues.
Not 100% sure what exact component was the cause but were ever it was it had a bad joint.
Thanx for the help guys.
-
NikTheGreek
- Posts: 348
- Joined: Thu Dec 08, 2011 4:17 pm
- Location: Greece
- Contact:
Re: Senors locking up and i2c errors
Great news.... 
-
mon_lolo_fr
- Posts: 40
- Joined: Tue Nov 15, 2011 9:50 am
Re: Senors locking up and i2c errors
So have nice flights now !
Re: Senors locking up and i2c errors
cardboard wrote:2KW of heatgun and some melted header pin plastic bits later and the board is all re-flowed. During the process I noticed some joints had a fair bit of soldier move around, most notably around the 3.3v regulater - at least I'm pretty sure that's the chip I mean.
you re-flowed the top or bottom side or both?
Did you just lay the board on the table or did you use a cushion to keep components fixated?
Re: Senors locking up and i2c errors
only components on the top side, only thing on the bottom is tracks and vias.
Sat the board on a wooden chopping board and went for it. I deducted the general area were the fault was most likely to be and just consentraed the heat there.
I could tell when I hit the magic temperature as the solider became shinier and in places were the joints were jagged it smoothed out.
I would have put it in the oven but with the header pins still attached I opted for the heat gun, at least I had some control over were the heat went that way and could reduce the amount of melted plastic.
Now some of the smds even look more aligned then when I received the board as well.
Sat the board on a wooden chopping board and went for it. I deducted the general area were the fault was most likely to be and just consentraed the heat there.
I could tell when I hit the magic temperature as the solider became shinier and in places were the joints were jagged it smoothed out.
I would have put it in the oven but with the header pins still attached I opted for the heat gun, at least I had some control over were the heat went that way and could reduce the amount of melted plastic.
Now some of the smds even look more aligned then when I received the board as well.
Re: Senors locking up and i2c errors
Hi,
That´s a perfect sign that there was a failure in the production process. They shoud always align to the pads due to surface tension of the molten solder, what you discribed showes that you got a board full of cold joints....
Greets for your successful repair.
cardboard wrote:Now some of the smds even look more aligned then when I received the board as well.
That´s a perfect sign that there was a failure in the production process. They shoud always align to the pads due to surface tension of the molten solder, what you discribed showes that you got a board full of cold joints....
Greets for your successful repair.