Saturday, June 6, 2015

Lawnbot's off road test drive video!



 Here is an updated video of the actual lawnbot running on the lawn.  (water rationing in southern Calif.  So not much actual lawn!)
 Initially it kept stalling out a couple times a minute.  Pause for just a second.  I thought it was pulling to many amps and the code was shutting the circuit down for safety.  This was not the case.  It was actually reading the radio to soon and was dropping packets.  This result in a no connection result that put the code into a loop till it received valid data from the transmitter.
  I did find that the current readings were topping out at 10amps (each motor)  with a quick change of directions.  So I installed the - 30 to +30amp current sensors and have them safety checking at about 15-17amps. I'll check after the added weight of the propane motor and it's electric starter motor,  to adjust it any further.  Right now the pwm is limited to 90 out of a possible 255, so there is still a lot of performance possible in this build!
  In the video,  you will notice that it manages to drive up and over the border bricks.  This was a surprise since I didn't think it had that kind of torque/power.  Still quite a bit of fun left in all these recycled scraps!
  Motor mounts for the propane motor and it's custom starter motor are next on the list (after weight/amp tests).  So be sure to check back.  Any questions,  please let me know!

Wednesday, June 3, 2015

Lawnbot finally moves!


06/03/2015
  It's been a while since I have been able to work on this project,  but it is now sitting in the middle of my living room for some major upgrades and testing!  (and my wife is still talking to me?!?)
  I had a 30amp circuit breaker attached.  The first time I tested it, It blew the circuit breaker under load.  So I installed 2 ACS711EX Current Sensor Carrier -15.5A to +15.5A.  The PWM is set to max out at 60 (out of a possible 255) and with the current sensors shutting down the current draw at 13amps per motor,  it seems to run quite well.  Higher Current supply or different motors will be a future upgrade after I get everything running correctly.
  A couple other upgrades you may notice from the video: it has a separate wireless control that will be able to take control during autonomous activities or in an emergency.  The remote is made with an analog thumbstick/joystick,  an arduino nano clone,  4x16 lcd display and an nrf24l01 2.4ghz radio (2nd one on the lawnbot).
  I also added a mosfet and relay switch to allow the lawnbot's mega2560 to turn the high voltage/current to the motor controller.  The mega does a simple voltage check to ensure there is enough of a charge. If so will turn on the motor controller's,  motor current feed.  It does this with a simple voltage divide reading on analog pin 0.
  With the upgrades of the current sensors,  relay,  nrf24 radio,  new main power switch,  circuit breaker and various leds,  it has been rewired a couple dozen times and has some new 3d printed parts.  I think the controller will outgrow the enclosure soon,  so I may have to design another mount.  I'll try to get some newer pictures in the next few days.  (already much different than what is shown in the video above!)
  The next steps are to test the failsafe controller by driving it around the lawn for a bit. The living room is only big enough to do some parallel parking tests.
  If that works out well,  then I will start building a custom mount for the propane motor and it's starter motor as well.  These will always be manually controlled for safety reasons!

Wireless controller before 3d printed housing.  Uses external 5v USB batter

A view of the control assembly from a few weeks ago.  When the wires were still manageable! 

Monday, January 26, 2015

3D Printer Linear Motion without Linear Bearings!

 I currently use several modified Printrbot Simple Maker's Editions. They are very solid machines and make some impressive and fast prints. My largest print size is 200mm x 320mm. Roughly 8” x 14".
  The one thing that is a constant battle for higher speed prints is slop or flexing parts during the movements. I started experimenting with different materials to use in place of linear bearings for use in 3d printing. Although 8mm linear bearings and case hardened rods are the go-to solution, they still have play that is evident in the prints (especially when changing direction)..
  One quick, possible solution for my existing printers are the small round parts in the pictures. They are 8mm I.d. x 10mm x 10mm linear bushings. The are self lubricating bronze linear bushings with a hardened carbon steel outer casing and a PTFE (Teflon) inner liner. Extremely nice fit with no play, but they will require some new mounts to be fabricated to fit my existing bearing mounts.

The white part on rails is a test for my custom designed printer. The plastic is also self lubricating, but is HDPE. 
  I am very excited about this test as it is smoother than the bearings could ever hope to be and will easily be fabricated so there will be no play when mounted correctly.
  It's 5/8” thick HDPE and has had all 4 sides routed top and bottom, leaving a 3mm tab that fits snugly into the 3mm screw rail of  15x15 mm extruded aluminum (openbeam v2.1).
   This is an idea that I hope to develop in my custom printer and maybe even for the CNC machine I have been planning for the future.


  The last item is a roll of PTFE tape. I had considered using it to line the edges of an aluminum plate that could also run in the extrusions rail with HDPE wheels to provide side pressure, removing all play. That is something I will be trying and sharing the results here as well.
  It may be necessary to use aluminum/Teflon combo over the HDPE glide depending on how well my ideas work out to prevent heat creep on the heated bed and extruder mount. HDPE is my material choice for these projects. Wonderfully smoother linear movements without bearings.

Thursday, November 13, 2014

New hardware including 3d Printers are being added!

I  got extremely sidetracked while waiting for a delivery of 10 ultrasonic sensors from China.  It basically took 3 months before the shipper would agree they were lost in transit.  In the mean time,  I received a Christmas gift that took over 100% of my hobby time.  A Printrbot Simple 2013 maker's edition 3D printer.  Since that time I have completely changed the design of that original printer and built 2 more.
  I started designing mod files and even recreating the original Laser-cut DFX files into Printable STL format. If you have a printer you can print your own 2014 Printrbot Simple 1405. If you don't you can commission an owner to print you one or even use a local 3d print service.  Those and all my files can be located at  http://www.thingiverse.com/metaled/designs
  I have a ton of subjects I would like to share hints and tips for.  Don't get me wrong,  I have ordered new Ultrasonic Sensors and do plan to continue sharing that robotics project.  But I  also plan to share Raspberry Pi,  PcDuino,  Arduino,  Intel Edison and lots of electronic circuits with the Lawn Mower Robot assembly!
  I'll try to find time to share as much as possible and if you have any questions,  please feel free to ask in the comment sections!




Sunday, October 20, 2013

Brains behind Arduino Lawn Mower!


Finally got a delivery from China.  It is the new Arduino Mega 2560 R3 compatible MCU along with a the v2 Sensor Shield.  (this combination had the same pin to usb short I discussed in an earlier post.  Be sure to check that post for the fix or it will fry your mega 2560! (Arduino Mega 2560 shorts against v2 Sensor Shield)
 It took a couple weeks to be delivered,  but the price is less than a quarter of what the cheapest price vendors are offering it here in the U. S..  Not only do a lot of local vendors order them from china,  but a lot of auction sellers will take your money and have it drop shipped from china.  Well worth it to do some research and cut out the middle-man! 


Now that I started to put the components into the battery box,  it is getting very tight with the new SLA 12Ah 12v battery.  The solution is to build a multi-level control board that will slip right in beside the battery. Seems like it works very well,  but it might get a little tight after I start adding sensors and more wiring. 

Thursday, October 3, 2013

Hobby RC remote for testing the Lawnbot!

I've been stuck indoors, unable to work on the mechanical parts of our robotic lawn mower.  I took advantage of this time and wired a Futaba FP-R127DF 7 channel Receiver to an Arduino Uno r3 and wrote a few lines of code to see if I could get it to read the signal from a Futaba T6XH-Super PCM1024 6 channel transmitter.  Was a lot easier to get the Arduino to display the signals than it is going to be for me to figure out what they all are.  This is a programmable aircraft/helicopter remote with settings for just about everything.  Ailerons, flap, Swash Plates and mixing of all the signals.  Wish it was a standard RC car remote!  It's going to take some research to program the transmitter.  With what I do know now, I believe I can get it driving the bot. 
 I started out the day knowing nothing about how to read or interpret rc hobby remotes.  I searched and found a real nice tutorial over at sparkfun.  Wasn't as complete as I thought,  but there is additional info in the comments that will help get it working with dual motor controllers.  The actual page is: https://www.sparkfun.com/tutorials/348
 BTW, not sure if the Raspberry Pi running Raspbian allows for enough time slicing in the multi-tasking to be able to poll the serial ports (usb)  for an rc transmitter combo.  I originally started using the Pi to program the Arduino as I always do.  When I opened the serial port monitor to view the signals, the cpu bogged down and brought the entire system to an absolute crawl at 9600bps.  Made it unusable.  Was forced to bring out another machine with an OS that I loathe! If it weren't my kids, I would install linux on it.  If you were to really delay the polling, you may be able to read it, but I would really check into other OSs or distributions if you want to use it with the Pi.  This is one of the reasons I  will be using the arduino with the Pi on this robot. 
 Tomorrow I hope to post some new updates on the chassis.  At the latest I hope to have something actually moving my the end of next week!  
 I also updated the current hardware page.  This is really starting to run together.  I think it's time to separate the parts by project.  That way if you are trying to duplicate something,  you will know exactly what I  did it with. 

Tuesday, October 1, 2013

Large mower-robot chassis progressing quickly !

Here's a quick update since I haven't posted in a bit.  Work is going on daily.  At least a couple hours a day is what I have been trying for.  Lately I  have been working on getting a physical chassis up and running.  What is pictured is a first mock-up of my 3 wheeled mower bot.  
The 2 rear wheels are salvaged, geared drive, dual 12v, 25 amp motors.  Taken from a kid's power wheels ride on jeep.  Along with the dual 45amp mosfets driven H-bridge should be more than enough to drive this bot through any urban environment . 
 The cutter motor is a 25cc, 4 stroke propane string trimmer (weed whacker) motor.  I had built the system around the string trimmer's 18" inch cutter head, but the body size of 18x24 inches is extremely large for my yard and would not only cut the lawn in a couple passes, but prevent me from maneuvering in the tight areas.  So I believe I  will not only shorten the cutter head, but also the overall size by about 2/3rds.
  The next step is to mount the battery box and some of the computer equipment so I  can start testing the base with a 6 channel remote control.  I will be sure to post some videos of this step as well. 
 The next hurdle is going to be cutting up the motor's clutch housing and fabricating a cutter head drive.  Obviously I won't be able to use the 8 foot drive shaft that came with the weed whacker. 
  So anytime I am not posting,  it's not that I have given up on any of this project!  It's more like I am so involved and barely have time to do anything but work on this bot.  I really want it to move forward quickly and have a working, remote controlled chassis in the next couple of days!  (did I mention a contractor has the front of my house torn off at the moment?  Ah,  Busy days!)