Cheap Bot, A Possible BSA Robotics Merit Badge Robot

2»

Comments

  • TtailspinTtailspin Posts: 1,326
    edited 2013-02-11 - 16:06:50
    I am a bit of a 'slowcoach', and still trying to figure out what the foam board is for??? :smile:
    InaPinch1.JPG


    -Tommy
    800 x 600 - 367K
  • Duane DegnDuane Degn Posts: 10,007
    edited 2013-02-11 - 20:22:47
    Ttailspin wrote: »
    still trying to figure out what the foam board is for???

    That is too funny!

    Very cool robot. Did I miss its figure 8 video?
  • TtailspinTtailspin Posts: 1,326
    edited 2013-02-11 - 21:59:28
    No, you didn't miss the video, but thats a great idea, Time to re-explore an old favorite. :)
  • Duane DegnDuane Degn Posts: 10,007
    edited 2013-02-18 - 13:27:50
    I was asked about how to get a robot to perform a figure 8.

    I thought this thread was as good as any to explain the process.

    I generally just try to make two side by side circles. To have a two motor robot drive in a circle generally you just need one wheel to turn faster than the other.

    It helps if your motors produce a consistent speed a particular setting. I initially had trouble with my cheap bot since the speed would vary depending on how freshly charged the batteries were. I have much better success when I used a higher voltage battery pack (Li-Ion) and had a regulator drop it down to a consistent 5.0V. I powered both the QuickStart and the servos with this 5.0V supply.

    I've found the CR servos only have useful speed control at relatively near the stopped pulse length (1500us). While a normal servo makes use of the full range of possible pulses (1000us to 2000us (or even more)), CR servos have a much smaller range of useful pulse lengths.

    For the cheap bot, I just guessed that a speed of 100us from stopped was within the useful range. I set these constants accordingly:
      DEFAULT_SPEED = 100
      DEFAULT_TURN_RATE = DEFAULT_SPEED / 4
    

    The full code for the cheap bot can be found in post #2 of this thread.

    So to turn, I set one wheel faster and one wheel slower by the turn rate (in this case the turn rate is 25).

    Here's the turn method with the ramping part removed:
    PRI SetSpeedTurn(localSpeed, turnAmount)
    '' A positive turn is counter clockwise.
      targetSpeed[PORT] := localSpeed - turnAmount
      targetSpeed[STARBOARD] := localSpeed + turnAmount
      pulseLength[PORT] := STOP + targetSpeed[PORT]  
      pulseLength[STARBOARD] := STOP - targetSpeed[STARBOARD]
      Refresh
      result++
    

    The constant "STOP" is:
    STOP = 1500
    

    The above code sets the left wheel to a pulse length of 1625us and the right wheel to a pulse length of 1575us.

    Once I have the speed and turn rate selected. I give the robot a try to see if I like how it's turning. The above speeds where just guesses on my part but I've had a lot of experience using servos so I was able to guess usable speeds on my first try.

    What I didn't guess correctly on my first try was how far around the circle the robot would get within a set amount of time.

    I could have just set the speeds and have the main loop wait a certain amount of time for the servos to do their stuff. The servo object automatically refreshes the servos so I don't need to keep sending pulses in the main loop. Instead of doing it this way, I instead "refresh" the servo by setting its speed 50 times a second.

    The reason I "refresh" has lots of reasons I wont go into right now (this technique is useful with complex movements likes hexapods). The main thing with a figure 8 is you need some way of timing how long the robot is performing each circle. I suppose it's possible to time how fast a robot travels at various speeds an compute the time to spend in each circle but IMO, it easier to figure out the amount of time required by good old trial and error.

    I left a bit of a record of my attempts in commented out parts of the CON section:
    REFRESH_RATE = 50
      STEPS_PER_FIGURE = 1150 '23 * REFRESH_RATE   ' 23 < ? < 24  with ramping
     
    

    I initially tried to have the "STEPS_PER_FIGURE" be some multiple of the refresh rate so I could just enter the number of seconds it should spend performing a loop. The required number ended up being a multiple of the refresh rate once I used a regulated 5V supply. Before using the regulated supply, I switched to using the larger number to make it easier to fine tune. I wasn't able to fine tune this number since it was a moving target prior to the regulated supply.

    The program uses half this value for each half of the loop.
    localIndex := STEPS_PER_FIGURE / 2
    

    The ramping used in the program adds a little bit of time at both the beginning and the end of the figure 8. Since this ramp time isn't significant, I didn't worry about it when trying to figure out how long the robot should continue to turn before switching direction.

    The code in post #2 should work with most robots using two CR servos like a BOE-Bot. You may need to adjust the DEFAULT_SPEED, DEFAULT_TURN_RATE or STEPS_PER_FIGURE constants to get your robot to perform a proper figure 8.
  • ajwardajward Posts: 1,101
    edited 2013-02-18 - 14:08:26
    PJ Allen wrote: »

    PJ... Thanks for the tip on this. Works great!

    Amanda
    Founder of the "Society for Aimless Tinkering and World Conquest"
  • GranzGranz Posts: 179
    edited 2013-03-25 - 19:39:19
    Could you possibly make this even cheaper by using the 1.99 micro servos, and cutting the body size a bit? Ive kind of wondered if they have enough power to pull around a really small robot, they are a little faster than the standard size servos.

    RWGast,

    Do you have a link for those micro servos? I am in need of a few for some animatronics projects.

    Thanks
    Art G. Granzeier III, President
    Granzeier Consulting
    www.granzeier.com

    Helping to Build a Better Engineer
  • Duane DegnDuane Degn Posts: 10,007
    edited 2013-03-25 - 22:00:33
    Art,

    HobbyKing sells some pretty good inexpensive micro servos. I've used the HobbyKing servos in both my hexapod attempts.

    I have links to other inexpensive servos in post #10 of my index.
  • GranzGranz Posts: 179
    edited 2013-03-25 - 22:48:36
    Duane Degn wrote: »
    Art,

    HobbyKing sells some pretty good inexpensive micro servos. I've used the HobbyKing servos in both my hexapod attempts.

    I have links to other inexpensive servos in post #10 of my index.

    Duane,

    Thanks, I have bought a few of those HXT-900 micro servos from Hobby King before. There's just something magical about that $1.99 price tag - it seems much more than just a dollar less than those '900s.
    Art G. Granzeier III, President
    Granzeier Consulting
    www.granzeier.com

    Helping to Build a Better Engineer
  • Duane DegnDuane Degn Posts: 10,007
    edited 2013-03-26 - 09:03:05
    Granz wrote: »
    Duane,

    Thanks, I have bought a few of those HXT-900 micro servos from Hobby King before. There's just something magical about that $1.99 price tag - it seems much more than just a dollar less than those '900s.

    I think there are occasional ebay deals with $1.99 servos. I could never find these deals myself. After seeing how well the HXT900 servos work, I stopped looking for cheaper servos.

    The HXT900 have a full 180 degrees of travel if you send pulse length outside the normal 1ms to 2ms range. They very strong for there size.

    I burned out two of my HXT900's when I attempted to power them from a two cell LiPo. I was attempting to boost their power in order to get better performance from my Popsicle stick robot.

    erco has managed to convert a HXT900 to continuous rotation.

    One other thing I've learned about these little servos is Polymorph can chemically bind to the plastic if it's hot enough. I used Polymorph to hold the servos to the Popsicle sticks. Most of these joints were easy to undo with the Polymorph pealing away from the blue servo plastic. On a couple of the servo, the Polymorph had combined with the blue plastic requiring the Polymorph to be cut away to remove it from the (now blemished) servo.

    I've been very positively impressed by these little servos.

    I've noticed on Let's Make Robots, the little robots with "heads" are more entertaining to watch. I'm wondering about increasing the performance of my cheap bot and give it some personality by adding an ultrasound sensor as a "head" with a HXT900 to provide it movement.
  • Duane DegnDuane Degn Posts: 10,007
    edited 2013-05-10 - 15:04:06
    Prof_Braino's LittleRobot(tm) thread, reminded me to update this one.

    I'm been collecting peanut butter jar lids for a while, and probably have a half dozen or so. As PJ Allen and others have shown, jar lids make fine wheels. 'Though I've been surprised how fast my collection of jar lids has grown, I was hoping to find a more readily available alternative to the jar lid. I think I've decided to use Vex 2.75" wheels. The do add $5 to the cost of the robot but IMO, I think they will add a lot to a robot's "cool factor".

    WHEEL-LOWFRICTION.jpg

    I haven't yet figured out how to mount the Vex wheel to the servo, but I'm not very concerned about this. I've mounted a variety of Vex wheels to non-Vex motors on three different robots. I think these will be relatively easy to mount to the metal gear HobbyKing servos. One nice thing about these servos is the servo horn is held on with a M3 machine screw. These screws are available in a variety of lengths which makes attaching wheels to the servo much easier.

    I've got several other pressing projects right now, but once I have some extra time, I'll mount a set of these wheels to a cheap bot and test them out. I'll report back on my experience with them here.

    I just wanted to point out these wheels to others looking to make an inexpensive robots.

    I've also considered making wheels from Polymorph using the same mold technique are I used on my hexacopter landing gear. I was thinking I'd have curved spokes to act as shock absorbers like on the Mars rovers. I think Polymorph wheels would cost less than the Vex wheels and the servo spines could be molded into the wheel to make it easy to attach.

    The main downside I foresee with the Polymorph wheels would be the time it takes to make the initial mold. Making the wheels with a mold shouldn't take very long. I was surprised at how easy the landing gear were make.

    I don't think Polymorph wheels would look as good as the Vex wheels. I'm hoping the shock absorber aspect will look cool (and hopefully work), but the Vex wheels will be hard to top or match on appearances.

    Even though, I'll probably go with the Vex wheels, I don't think I can keep myself from trying to make a Polymorph version. Now that I've got the image of shock absorbing, servo spline hugging wheels in my head, I don't think I'll be content until I have a working version to try.

    I know OBC (Jeff) is using wheels cut from acrylic. They look okay, but I think it's worth the extra buck to purchase a set Vex wheels. Parallax's servo wheels should be easier to mount (if you use the correct brand of servo) but they cost more than the Vex wheels (and IMO, the Vex wheels look cooler).

    The Vex wheels cost $2.50 each but they only sell them in sets of four.

    I think robotics is a great way to get kids interested in STEM topics. I thought it was fun to control colors and shapes on my Apple ]['s monitor and this was a motivating factor to get me to learn some programming. I'd think controlling a robot rolling around in the real world should be much more interesting and hopefully provide kids with enough motivation to learn about electronics and programming.

    I'm very interested in what components others are finding in their attempt to build an inexpensive robot. I'm also interested to hear what your opinions are about the Vex wheels and other inexpensive wheel options.
  • Duane DegnDuane Degn Posts: 10,007
    edited 2017-04-06 - 21:26:03
    I tried mounting the Vex wheels to the CR servos today. I am very pleased with how they turned out.

    I initially thought I'd try attaching the wheel to a servo horn as I did with the Lego wheels but I didn't want any extra nuts or bolts interfering with the cosmetics of the wheel (I think the Vex wheels look cool).

    I decided to try drilling a hole in one side of the wheel and see if I could attach the wheel directly to the servo without using a servo horn.

    I used a 7/32" drill bit and drilled about 1/8" in to the wheel. I was then able to press the wheel onto the servo with a bit of effort.

    The wheel on the right is one that had been drill and pressed onto the servo.

    attachment.php?attachmentid=101676&d=1368933326


    I used a 20mm 3M machine screw to secure the wheel to the servo. Below you can see the length of the screw and the metal servo gear it mates with.

    attachment.php?attachmentid=101675&d=1368933324


    I the next photo you can see how far into the wheel the final gear was pressed. I looks like I could have drilled a little deeper into the wheel without causing a problem.

    attachment.php?attachmentid=101674&d=1368933322


    I was pleased with how straight the wheel was. You can see the results in the video below. (I'm still shopping for a better camera. My present camera seems to only want to record audio of camera body contact.)





    (Skip ahead to 1:40 to see the wheel spinning.)

    I hope to get these servos and wheels mounted to a robot soon and take them for a test drive.


    So far I'm very pleased with these $2.50 (each) wheels.
  • ercoerco Posts: 19,560
    edited 2013-05-18 - 21:09:28
    Sweet results, DD! Hand drilled, or more likely drill press?

    I had a busy, challenging day, too. Nothing robotic or electronic for a change. Took the girls to Disneyland solo. Fear Factor goal #1 accomplished: taking twin girls into a public MEN's restroom. :)

    mickey.jpg
    768 x 576 - 188K
    "When you make a thing, a thing that is new, it is so complicated making it that it is bound to be ugly. But those that make it after you, they don’t have to worry about making it. And they can make it pretty, and so everybody can like it when others make it after you."

    - Pablo Picasso
  • Duane DegnDuane Degn Posts: 10,007
    edited 2013-05-18 - 21:29:01
    Wouldn't a nice quiet game of rugby be a better convalescing activity than taking twins to Disneyland?

    That's a fantastic picture! Thanks for posting it.

    I did use a drill press. Even though the original holes were square, they still helped guide the bit so drilling a reasonably centered hole wasn't too hard.

    The size of these wheels are different than the Lego wheels. I'll need to tweak the numbers to get a figure 8 out this robot again (or use the compass).

    It's been a while since myself or anyone else has added to the figure 8 thread so testing the new wheels on this robot will be good to have an excuse to post another video to the collection.
  • Cluso99Cluso99 Posts: 15,540
    edited 2013-08-03 - 15:56:16
    Just reading this thread. I purchased a Tamyra dual motor setup with interchangeable gears. IIRC cost was ~$16 and I bought 2 cheap wheels. This would seem comparable with the cost above using servos.
    I made a tiny pcb with a motor controller IC that mounts to this gear chasis and can be driven by any micro. Unfortunately I have never had the time to build the pcb and test it out. I thought this setup would make for a really cheap robot.

    The battery box with switch, and the breadboard can be purchased cheaply on eBay. They may not be the best of quality, but considering their low cost they certainly make a low level cost entry possible.
    My Prop boards: P8XBlade2 , RamBlade , CpuBlade , TriBlade
    P1 Prop OS (also see Sphinx, PropDos, PropCmd, Spinix)
    Website: www.clusos.com
    P1: Tools (Index) , Emulators (Index) , ZiCog (Z80)
    P2: Tools & Code , Tricks & Traps
  • Duane DegnDuane Degn Posts: 10,007
    edited 2013-08-04 - 12:14:35
    Cluso99 wrote: »
    Just reading this thread. I purchased a Tamyra dual motor setup with interchangeable gears. IIRC cost was ~$16 and I bought 2 cheap wheels. This would seem comparable with the cost above using servos.
    I made a tiny pcb with a motor controller IC that mounts to this gear chasis and can be driven by any micro. Unfortunately I have never had the time to build the pcb and test it out. I thought this setup would make for a really cheap robot.

    The battery box with switch, and the breadboard can be purchased cheaply on eBay. They may not be the best of quality, but considering their low cost they certainly make a low level cost entry possible.

    I like the motor, gearbox and h-bridge idea because it seems more "honest" as in that's how "real" robots work. But there's also a lot to be said for keeping it simple by using a servo.

    The metal geared servos from HobbyKing are pretty darn powerful. I'd be surprised if the Tamyra dual motor setup would be more powerful. I have one of those setups so I should compare it with the HobbyKing servos.

    Another reason I like the servos is I don't mind stocking up on them. I already have at least 20 of them. I intend to resurrect my popsicle stick robot with these servos at some point. I think I'd have a hard time coming up with uses for surplus Tamyra gearboxes.

    I think I'm no longer going after the very least expensive bot possible. I'm going to try to put a kit together that has the most bang for the buck.

    I'm also debating the use of the foam core board. I recently purchased a CNC router and I think I'll use it to cut expanded PVC to make a simple robot chassis. I learned about expanded PVC from Gordon McComb, and I find it really easy to work with. I'm hoping a chassis made from this plastic will be more satisfying to the young roboticist.

    Presently I'm holding off on having a robot class until I get my CNC machine up and running.

    Thanks for adding your thoughts and suggestions.
  • Cluso99Cluso99 Posts: 15,540
    edited 2013-08-04 - 19:54:36
    Thanks Duane. Wishing you success with your project.
    My Prop boards: P8XBlade2 , RamBlade , CpuBlade , TriBlade
    P1 Prop OS (also see Sphinx, PropDos, PropCmd, Spinix)
    Website: www.clusos.com
    P1: Tools (Index) , Emulators (Index) , ZiCog (Z80)
    P2: Tools & Code , Tricks & Traps
  • mindrobotsmindrobots Posts: 6,506
    edited 2013-08-04 - 20:58:48
    Re: foam core boards. We built some of the cheap robots using the foam core boards as a chassis plate. The double sided tape worked well for attaching things...too well, in fact. A couple of the robots had the foam core boards delaminating from the weight of the motors. They were stored without weight on the wheels and the paper pulled away from the foam core. Easy to fix but still a frustrating failure for kids and adults alike!
    MOV OUTA, PEACE <div>Rick </div><div>"I've stopped using programming languages with Garbage Collection, they keep deleting my source code!!"</div>
  • ercoerco Posts: 19,560
    edited 2013-08-04 - 21:14:01
    Foam core has its uses. It's lightweight and cuts easily with an exacto knife, but that's the best I can say about it. I'd stay clear for using as a robot chassis for a few reasons. Besides delaminating, it creases easily and that's the end of it. It's not all that inexpensive. I bet Home Depot hardboard (with or without pegboard holes) is about as cheap, and much stronger & durable. $5 for a 2'x4'x1/8" piece.
    "When you make a thing, a thing that is new, it is so complicated making it that it is bound to be ugly. But those that make it after you, they don’t have to worry about making it. And they can make it pretty, and so everybody can like it when others make it after you."

    - Pablo Picasso
  • mindrobotsmindrobots Posts: 6,506
    edited 2013-08-05 - 04:15:40
    I'm going back to my Baltic Birch plywood. It's easy to work and takes a Danish oils finish very nicely plus it can be stained to match any decor! :lol:
    MOV OUTA, PEACE <div>Rick </div><div>"I've stopped using programming languages with Garbage Collection, they keep deleting my source code!!"</div>
  • whiteoxewhiteoxe Posts: 794
    edited 2014-04-29 - 20:43:02
    Duane Degn wrote: »
    The chassis and tail dragger were made from foam board. Here's a picture of the two pieces of foam board used to make the robot.


    attachment.php?attachmentid=99076&d=1360118436

    Neither of the pieces where measured. I just help up the pieces I wanted to fit on the robot and cut the rectangle to general size I wanted. The tail dragger was made long enough to keep the robot relatively level with wheels I planned to use.

    I used Gorilla Tape to attach the tail dragger piece to the rest of the chassis.

    Attachment not found.Attachment not found.

    I used some hook and loop material that's different than the usual "Velcro". This particular product sticks to itself so it doesn't require two different types of material.

    Attachment not found.

    I used this material to secure the battery pack to the chassis and the voltage regulator to the battery pack.

    I used "servo tape" to attach the continuous roation servos to the chassis.

    Attachment not found.Attachment not found.

    The breadboard was secured to the battery pack with double sided foam tape (included with the breadboard).

    attachment.php?attachmentid=99080&d=1360118436

    I inially used NiMH cells in the battery pack since four NiMH cells are close to 5V and can be safely used to power the servos directly. I found the change in voltage over time from the draining batteries made the speed produced by the continuous rotation servos inconsistent between runs of the robot.

    My solution to produce a more consistent speed was to use four Li-Ion cells and then regulate the voltage down to 5.0V. This power supply system worked much better than the original NiMH system.

    The switching regulators from ebay had solder pads for additional capacitors which I used to solder headers to the PCB to make it easier to connect the regulator to the power supply and the breadboard power rails.

    attachment.php?attachmentid=99077&d=1360118436

    The regulator on the right (in the above photo) is how the regulator originally came from the seller. The regulator on the left is the modified version with headers soldered to the board.

    Edit: See post #5 for more assembly photographs. I was going to add them here but the forum software limits attachments to ten (I have eight so far and I wanted to add three more).

    See post #4 for a bill of materials.
    can i cut that foam board with a sharp blade or can it also be cut with a hot[wire foam cuter ? and they are really Leggo wheels ? i want a couple :)
    there is nothing either good or
    bad, but thinking makes it so.
    Favourite quote of all. Seems to damn the ideas of morally that sam harris champions. He makes claim that science can teach morality better than religion. That might very well be true but who can say or define what is bad and speak for all cultures and times? Its all a bit of an illusion.
  • Duane DegnDuane Degn Posts: 10,007
    edited 2014-04-30 - 08:55:42
    I was asked about the regulators shown in post #3 and some other places in this thread. These are the inexpensive regulators available on ebay. There's a link to an ebay seller in the BOM on post #4.

    Post #35 mentions why using a regulator is important for consistent speed.

    There's some discussion about these regulators in this thread.
  • Duane DegnDuane Degn Posts: 10,007
    edited 2014-04-30 - 09:07:22
    whiteoxe wrote: »
    can i cut that foam board with a sharp blade or can it also be cut with a hot[wire foam cuter ? and they are really Leggo wheels ? i want a couple :)

    I don't know how I missed this post. I didn't see post #51 until after making post #52. I think it's because I had this page already open in a tab and I hadn't refreshed it prior to making a post.

    I used an Xacto knife to cut the foam board. I don't think a hot wire would work well cutting through the paper.

    Yes, those are Lego wheels from a Technic motorcycle set. I agree, they are nice.

    As I mention somewhere else in this thread, I'm planning to use Vex wheels. At $2.50 each, they're a pretty good deal IMO.

    Concerning using a QuickStart board or a Gadget Gangster board, which you asked about in the other thread, either board should work fine. One thing I like about the QuickStart is I can use the pins connected to the LEDs for signal pins and I can verify the signals are being sent by watching the LEDs flash. I really like using the LEDs with h-bridge robots.

    Never underestimate the value of flashing LEDs. Unfortunately a few flashing LEDs can be more impressive than GPS navigation or other cleaver software/hardware additions.
  • whiteoxewhiteoxe Posts: 794
    edited 2014-04-30 - 19:26:32
    Duane, this morning I went to the biggest electronic store in brisbane 'Jaycar' usually they have been great. Today, hopeless . I have the foam board because i was in town and near the shop that sells it, so i am going to try out a design similar to yours. I am about to order a couple of continuous servos. I went to the electronics store today to ask them what do H-Bridge motors do and how do i connect them to a motor controller. Well the chaps there this morning, there were half as many as usual, didnt have a clue even what a H-brodge motor controller was much less what an encoder was! Often there are people there that play with microcontrollers qnd all sorts but today they must have had the morning off ? Can you link me to a H-briridge motor controller for the continuous servos, i think ive got the encoders from spark fun. Putting these together will be just another challenge !

    Ps. I have ordered a bunch of different wheels from toy and remote control cars cheap from ebay.

    I looked on ebay for H-Bridge MOTOR CONTROLLERS, There were heaps but 99% stated they were for stepper motors/servos. The only one not for a stepper motor was this , do you think two of these would work with the continuos type and an encoder ?

    http://www.ebay.com.au/itm/1-Pcs-Motor-Drive-Controller-Board-L298N-Dual-H-Bridge-Module-For-Arduino-New-/141145789355?pt=AU_B_I_Electrical_Test_Equipment&hash=item20dcf1d3ab#ht_4940wt_983
    there is nothing either good or
    bad, but thinking makes it so.
    Favourite quote of all. Seems to damn the ideas of morally that sam harris champions. He makes claim that science can teach morality better than religion. That might very well be true but who can say or define what is bad and speak for all cultures and times? Its all a bit of an illusion.
  • Duane DegnDuane Degn Posts: 10,007
    edited 2014-04-30 - 20:04:12
    whiteoxe wrote: »
    Can you link me to a H-briridge motor controller for the continuous servos,

    You don't need an h-bridge to control a CR servo. The h-bridge is included in the servo electronics. This is why they are so popular with hobby robotics.
    Duane Degn wrote: »
    CR servos are popular with hobby robotics because they're so easy to control. Just the single control wire. A h-bridge circuit usually requires at least two control lines per motor.

    If you want a h-bridge to control small motors similar in size to those found in servos, you could use one of these h-bridge chips.
    Duane Degn wrote: »
    Small motors can be controlled with a chip like the L293 or SN754410.
    whiteoxe wrote: »
    i think ive got the encoders from spark fun.

    Which encoders? There are lots of different kinds.
    whiteoxe wrote: »
    Putting these together will be just another challenge !

    I concur.

    My four encoder plus four PWM driver provides feed back from up to four quadrature encoders and will provide PWM signals to up to four motors. If you set the direction pins to -1 the driver can be used with CR servos.
    whiteoxe wrote: »
    The only one not for a stepper motor was this , do you think two of these would work with the continuos type and an encoder ?

    http://www.ebay.com.au/itm/1-Pcs-Motor-Drive-Controller-Board-L298N-Dual-H-Bridge-Module-For-Arduino-New-/141145789355?pt=AU_B_I_Electrical_Test_Equipment&hash=item20dcf1d3ab#ht_4940wt_983

    A lot of people do not like L298N h-bridges. Apparently they drop the voltage to the motors by about 2V. I found the inexpensive price too hard to resist and I use two L298N boards, very similar to the one you linked to, in each of my two Rover 5 projects.

    In my more recent projects, I'm using these h-bridge chips. Apparently they are supposed to work much better than the L298N chips. I haven't used the MC33926 drivers enough to have a strong opinion about them.
  • Duane DegnDuane Degn Posts: 10,007
    edited 2014-12-12 - 20:39:59
    As I've mentioned previously, I'm attempting to find an inexpensive board to use as a robot controller. My initial choice of the QuickStart board was revised when the price of the board was increased to $35. While I think the Propeller Project Board is a viable option, they weren't available to purchase for several months. I've decided to try my hand at making my own Propeller boards to see if I can bring the price per board below the price of the Propeller Project Board.

    One possibility I'm exploring it to use a small Propeller board as the brains and a larger board to supply the various connections for sensors, motors, servos, etc.

    As I mentioned my my NonaProp thread, I can fit nine Propellers on a 10cm x 10cm PCB (a common size at many discount PCB fab houses).

    attachment.php?attachmentid=112127&d=1417501373

    As mentioned in the NonaProp thread, I included a capacitor inline with the reset line so I can program the Propeller with these inexpensive USB to serial adapters.

    attachment.php?attachmentid=112128&d=1417501375

    I'm still not sure what connectors I'll include on a companion board but I'm hoping I can make a board to allow easy use of some of my favorite inexpensive sensors and components.

    I'll list some of my favorite inexpensive components in a separate post.
  • Duane DegnDuane Degn Posts: 10,007
    edited 2014-12-12 - 20:41:28
    I recently had someone ask me about my experience with inexpensive sensors and other components one finds on ebay and if these inexpensive sensors will work with the Propeller.

    While these inexpensive sensors have been discussed many places on the forum, I can see how someone new to the Propeller would have a hard time trying to find all this information (though much of it is listed in my index). I thought my "Cheap Bot" thread would be a good place to list some of these components with links to drivers to use them with the Propeller. For now, I'm only listing drivers written in Spin/PASM. It should be possible to convert many of the Arduino drivers to Propeller C but I haven't done this myself.

    I haven't decided which sensors I'll want to include in any sort of "robot school" I may hold, here are some of my favorite inexpensive sensors and components.

    The HC-SR04 ultrasound sensors aren't as nice as Parallax's Ping sensor, but most of them work reasonably well. Post #32 of erco's "One Buck Ultrasonic Sensor" thread includes code to use these sensors with the Propeller.

    The ultrasonic sensor code is just an adapted version of the PING))) code in the Propeller library. I recently wrote a modified version of the PING))) code which allows one to set the value used for the speed of sound in the distance calculations. There's also a method which calculated the speed of sound based on the temperature. I doubt the temperature compensation method is very useful in the real world but if there's interest, I'll adapt my new Ping object to work with these inexpensive sensors.

    BTW, check your local Radio Shacks for Parallax's PING))) sensors. I found several locally for $7 each and I was able to purchase six genuine PING))) sensors on ebay for about $5 each.

    I know Phil has had some bad experiences with the cheap HMC5883L sensors but the few I've tried have worked well. Here's a link to the thread about the object I wrote to interface with these sensors. I'm pretty sure the thread includes a link to the seller from whom I purchased my sensor.

    While not specifically a robot part, I like these little OLED displays.

    attachment.php?attachmentid=108626&d=1399775726

    I've only been able to get them to work with the display configured as a SPI device. While the object in the OBEX works great, the SPI section of the code was pretty inefficient. I attached my version of the driver to post #14 of this thread discussing small displays.

    Another non-robot component I like are these MRFC522 RFID readers.

    attachment.php?attachmentid=112264&d=1418441768

    I haven't figured out all the features of these readers and cards but the object attached to post #40 includes methods to read data from the cards as well as methods to write data to the cards. These 13.56MHz readers and cards are not compatible with Parallax's 125kHz readers and cards.

    It looks like the Sharp distance sensors erco told us about are still selling for about $6. You'll need an ADC chip to read the analog output from these sensors. This thread and this thread include recommendations on ADC chips.

    Erco mentions several inexpensive options for real time clocks in this thread. As mentioned in the thread, there are objects in the OBEX with work with these RTCs.

    One of my favorite sensors doesn't have a full Spin/PASM driver (that I'm aware of). It's the MPU-6050 gyro/accelerometer combo. I've been very impressed with the Arduino code used with this sensor. While I'd like to have a Spin/PASM version of the code, I've decided for now to use a small Arduino clone as a slave controller to interface with this sensor. Since the sensor is a 3.3V device, I'm presently using a 3.3V Arduino clone. My plan is to communicate with the slave board over the slave's UART.

    I've mentioned many times I've a fan of the nRF24L01+ transceivers. I have information about these transceivers in post #1 of my index.

    While not a cheap component, I have recently become a fan of these wireless Nunchucks. I've found many clones of game controllers are very poorly made but I really like these "Kama" controllers. The joysticks have 8-bit resolution like the original. I've found some clones use a scaled a 5-bit value for their joystick output but these Kama controllers don't have reduced resolution. There are objects in the OBEX which makes using a Nunchuck with the Propeller really easy. SparkFun sells a little PCB to allow one to connect to a Nunchuck (or Nunchuck receiver unit).

    The receiver of my black wireless Nunchuck can be seen on the top of my Halloween Hex.

    attachment.php?attachmentid=111915&d=1415586382

    I'm using a small PCB I designed (which is pretty much a copy of the WiiChuck) to interface with the receiver. The receiver has a LED on it. It's the single LED behind the LED arrays.

    There are many more inexpensive sensors and components available. The ones I listed are some of my favorites. I'll likely add to this list as other sensors and components come to mind.

    A lot of the links in this post are to ElectroDragon. I'm a fan of ElectroDragon since their prices are usually pretty low and even though they charge for shipping it isn't very expensive and it's usually faster than the shipping ebay sellers use.
  • Duane DegnDuane Degn Posts: 10,007
    edited 2015-06-02 - 10:36:48
    Here's another item I want to add to my list of cheap robot parts.

    5043.JPG

    It's a microSD card reader.

    I'm using this reader my CNC controller project and it appears to work fine.

    I also like Parallax's Propeller Memory Card.

    40004.png?itok=wsRw6Tgy

    The PMC has additional memory chips to experiment with.

    I'm using the PMC in my robot remote project.
Sign In or Register to comment.