The responses came out very well for the first two minutes or so. After a while with the GPS connected however, I noticed that the servo's were starting to jitter. There is no manual override in that case. Taking out the electronics carefully, I noticed that the GPS lights were continuously flashing (A, B + C), much like in the startup sequence.
With the GPS connected, the board is continuously resetting itself. I know that, because after taking off the GPS plug directly, the flights eventually dimmed, then blinked for a while, a servo moved and then the lights went back into flashing mode (no fix). After that, the servo's could move again.
It is important to mention here also that also in manual mode, nothing worked consistently well. The signal would arrive at the receiver eventually, but there were freakish servo outputs of full rudder or elevator or just jitter. Basically, should this happen in the air it will probably crash.
The power feed is ok. I removed the solder jumper, because the board gets hooked up to the battery directly and the servo's through the receiver. In case servo's draw huge amounts of current (when locking up?), the board won't get affected by power dips or surges. So, I'm not powering the entire plane through the USB port or anything.
I don't have much more time today to look into this, but I'd like to ask anyone else if they experienced the same issue. My GPS was not shipped with the ardupilot settings, but I got it from coolcomponents and had to upload the configuration myself. The firmware was already v6.02. It's a ublox. The concern of a colleague of mine was regarding the battery on the adapter board, which seems to sometimes touch some components on the other end.
Please let me know if someone has seen this before and if this was resolved somehow.
You need to be a member of diydrones to add comments!
Replies
http://www.electronics-lab.com/articles/LM317/
You could take voltage from 11V to 9V and then let the on-board regulator do the rest if you need to in order to distribute heat between more than 1 regulator. That's what I did on mine.
I also haven't seen any formal replies regarding the voltage of 7 or 6 V. The manual states this:
http://code.google.com/p/ardupilot-mega/wiki/Hardware
My board has 6+ V printed on that location.
Also please notice that the manual doesn't specifically exclude RC usage, but mentions the use of separate battery's as if it were a personal preference. If the intention is different, some clarification would be required there.
If the shifting voltage still doesn't get rid of the excessive heat issue, then I'll look into a little heatsink or perhaps consider options for replacement of that particular component.
Can you explain more why you chose not to use the recommended power source?
With only the receiver powered: 100 mA.
With the board powered: 250 mA.
With the gps powered as well: 350 mA.
So the board seems to require 250 mA overall and in the future a little bit more when I get my airspeed sensor attached to it as well and the telemetry xbee.
As soon as the GPS is attached and a little while later, the board starts acting funny. No one is moving the sticks and the tx is all the time in MANUAL mode. The LM317 is getting hotter and seems to start to put out very ugly voltages, causing strange behavior. At the bottom side of the print near the regulator, it's getting so hot that it wouild cause burns if touched for a short time.
I noticed that at cold startup, the ppm led blinks happily and very periodical. But during the strange behavior, the ppm encoder led starts to blink irregularly, the mux led starts blinking, remains on, then goes off and the board seems to restart (the gps leds start flashing again, etc.). It's probably caused by the microprocessors going haywire.
The specification of the LM317 seems to imply that this regulator should be able to handle 1.5A rather easily (although the apm spec says it's 1A peak). Would you say that my regulator is definitely broken? We calculated that with 250mA and a bit more going through the board and regulating the voltage down from 12V to 5V, the regulator has to consume 7 * 250mA of power. 1.75W seems a bit steep for such a tiny thing.
I'm now considering pre-regulating the voltage down with an adjustable switching voltage regulator from "Dimension Engineering". These will leave a bit of a ripple on the voltage after stepping down, but I expect the LM317 to shave this off afterwards. That means the LM317 will then only have to reduce 1-2 V only @ 250/300mA, which means consume between 0.25-0.75W only.
One ambiguity for the board developers here that needs to be figured out. The technical details state that it's a regulator with 7-14V range. On the board itself it says "6+ V" near the power supply pins. Is it 6 or 7?
Let me know if the voltage regulator is simply definitely broken.
http://code.google.com/p/ardupilot-mega/wiki/Encoder
Even after that I had still some problems with power coming from ESC or with jumper unsoldered and direct powering the board with a second lipo
http://code.google.com/p/ardupilot-mega/issues/detail?id=29
I am now waiting for a non switched ESC to see if the jitter is coming from that
"Normally" with usb power , I do not have any issues once the 328 reprogrammed.