Ok, time to come clean. The wind isn't the only reason I haven't made much quad progress this week. A new project has stolen my time for the week. It's not really multirotor related so excuse me for sharing it here...but I know a few people enjoy the DIY aspect of my thread and this fits in...and it is flight related so hold on for a few minutes. If you'd prefer skip to the end of the post and I'll share some more test video from the quad today
The new project is a large format plotter! I've been putting off building any more fixed wing craft for a few months because a friend offered me this plotter - but has been dragging his feet on actually making the transfer of ownership happen. Well, earlier this week he finally said "Come get it at lunch today" so of course I did:
It's an early 90's era HP DesignJet 600. B&W only but capable of 24"x33" prints since it uses 24" roll media. If I can get this bad boy going I'll be able to print out FT plans on one sheet! But there will be some challenges. The plotter is old, it only has serial and parallel - USB wasn't even a dream at the time this thing was built. It does have a network card though so that's promising. HP hasn't released drivers for it since Windows XP...but even Windows 7 has stock drivers for it built in. It's also in rough shape. It's been sitting in a friends bosses shop for who knows how long and was covered in dust and looked like it may have had something splashed on it at one point
I did some initial cleanup on it and tried to power it up...it made it most of the way through the initialization but then flashed an error about the window needing to be closed and gave a system error 010001. A bit of work on google turned up a service manual and I found that sure enough error 010001 means something is wrong with the window sensor. Tried to figure out how to get to the sensor...but no luck. Thankfully the service manual had good instructions on disassembly/assembly as well!
When I got the end cover off the problem was apparent. Something had been living in here. Or at least storing food in here. It was packed with old dog food! No bedding or feces which I normally see when a mouse has moved in...but one or two spots that looked like urine stains. Fun. Got all that cleaned up and noticed that the ribbon cable connecting the front panel to the main control board was damaged. 4 wires on it were broken. But with the work on the quad I've gotten quite good at soldering ribbon cable so I cut out the damaged sections, soldered in some new wire...and sure enough the window sensor worked again!
With that fixed I tried to load paper. After almost an hour of trying and failing repeatedly I finally read the manual. Yep...it should be sucking the paper in and isn't. Few more diagnostics and it's confirmed...the media sensor isn't responding. It's wires run in the same general area so I was pretty sure I knew what the problem would be. Took the covers back off and sure enough it looks like whatever was living in there went in and out through the same hole these wires ran though and eventually broke them against a metal edge. Took a bit longer to find the actual media sensor since it was about 8" away and the wires to it were totally missing. Eventually I found it and it looked ok...but only one leg still had any wire with insulation on it so I had no idea which wire went to which leg:
(Look close above the repaired wires and you can see the broken media sensor wires)
Did some more googling but couldn't find any photos of the sensor that would let me figure out which leg went to which wire. Did find that the sensor is just an omrom part available for <$8 new...but if you buy it from HP it goes for $50-$100! Then I found this site:
http://www.myolddesignjet.com/ and contacted the owner to see if she could help me figure out which wire went to which pin. She replied back very quickly and was extremely helpful. Sent me two sets of photos of a stock sensor showing the wires.
Wired up a new harness...and after a few tries I was able to get paper loaded!
Now I just need ink. It won't finish initializing until it can adjust the pens and it can't adjust the pens without ink. So I've got to wait a few days for some ink to show up (carts are still available but not locally.)
Fingers crossed that when I get the ink I'll be able to get it to plot...and crossed even tighter that I'll be able to get the computer to talk to it and actually send a PDF to it!
Now...back to the topic.
This morning I got to go up to the park for an hour or so to do some flying. Went fairly well. Did a test with the actioncam again and despite some horrible jello from my non-issolated mount it did pretty well:
After that things went downhill though. Hit a few trees, landed in a mud puddle, had some batteries that set of the beeper way sooner than I expected....but still got some nice flying in. Even started feeling more comfortable with acro mode (just not with the camera on!)
Really need to build a better mount.
Came home and got lunch with my wife and daughter...then while they went to the park (ironically the same on I flew at earlier) I got a few more minutes. All of my 2200 packs were in need of a charge at this point so I decided to try a 1500 for a change.
Worked really well. My 1500's are almost half the weight of my 2200's so it was a bit harder to get it to balance with the cam - but without it it felt great! Pretty sure I'll be using the 1500's when I finally start trying some real acro moves down the road
Even so I was able to do some more adventerous flying around the yard - I even made it through the carport which is something I'm pretty hit or miss with on the x1 still:
On the last 1500 I finally started adjusting my PID's. Was able to turn the P up almost as much as I had to turn it down before and still haven't gotten any wobbles yet. So that's promising.
I did try the position hold and RTH at the park and was fairly successful. The position hold doesn't hold as well as I'd like even with 11 sats locked on, wind still tended to push it fairly far away over time. But in general it held. And RTH definitely seemed to work...but I didn't like the quad flying itself since there were some other people at the park even if they were fairly far away.
Can't wait to get out to the desert again for some more intensive testing...still a few weeks before I can do that though