Shop OBEX P1 Docs P2 Docs Learn Events
Which parallax bot should I purchase? — Parallax Forums

Which parallax bot should I purchase?

rwgast_logicdesignrwgast_logicdesign Posts: 1,464
edited 2013-03-20 07:01 in Robotics
I really need some advice on purchasing a robot. My birthday is coming up in a few weeks. Im torn, between a scribbler2, BOE BOT hardware/electronics kit, and a stingray.

My first choice was the stingray, its a nice robot chassis and is very expandable with good speed, ive also got most everything needed to turn this in to a really nice finished bot.

Im also really thinking aboit getting to boe bot kit minus the bs2 board. I have a bs2 HW board and prop boe. I like this idea becuase it will allow me to work through all the boe bot materials and learn more about the software side of robotics, which is where im lacking right now! Im also rebuilding a roomba so i already have a stingray size bot, and a boe bot sized bot is a little easier for experementing. My biggest concern about the boe is its size it doesnt look very expandable to play with other stuff that needs pan and tilt.

Then theres the scribbler witch i cant really see buying as its not very expandable at all. I really need room to grow as this may be my last big purchase for a long while. I think a stingray with hb25s would be compat with all the boe bot material but hb25s are pretty spendy, compared to a solution that doesnt use rc style pwm.

what do you guys think?
«1

Comments

  • Cats92Cats92 Posts: 149
    edited 2013-01-25 01:18
    Hello,

    Chose something very open for progress.

    I dream of a bot with a Propeller and a card for image analysis.

    Perhaps this one is too big : http://veterobot.com/index.html ?

    But the general idea seems good : with a robust chassis ( Dagu Rover 5 Tracked Chassis )

    And a lot of work for next months !
  • NWCCTVNWCCTV Posts: 3,629
    edited 2013-01-25 06:44
    http://veterobot.com/index.html Not available yet according to their site.

    To me the Stingray would be a good choice with the heardware you already have. Plus, it's on sale! I have a Boe Bot and although it is great to work with, it is quite small for expanability. A Scribbler however might be a good base to hack!!!

    Edit: I have some HB-25's.
  • Martin_HMartin_H Posts: 4,051
    edited 2013-01-25 09:25
    I don't own a String ray, but for a hundred dollars it looks like a decent platform. The HB-25's are nice, but are probably overkill for motors that size as they can handle 25 amps. There are a number of serial motor controllers that would work with a BS2 and those motors. If you switch to a Propeller chip you can use a cheap H bridge chip like the sn75441one.
  • Matt GillilandMatt Gilliland Posts: 1,406
    edited 2013-01-25 09:53
    Hey RW - yes you could use an HB25, but that's significant overkill. Two other options would be the L293 or the Polulu device. The L293 is a real inexpensive solution, but perhaps slightly under-powered. We've done it, and it works. The Polulu controller should work really well, although ironically, I don't think we've ever taken the time to try the circuit ourselves - at least I know I haven't yet.

    -MattG

    PS Hey everybody - RW is about to get his 4th star:)
  • rwgast_logicdesignrwgast_logicdesign Posts: 1,464
    edited 2013-01-25 11:15
    I understand the HB25s Are wayyy to much, I acually already have some 2.8/5.5 amp hbridge boards around that ive made, these l6205s work well on a roomba and Im sure are perfect for a sting ray. So I essentially have all the hardware I need to get a stingray running out of the box!

    The only reason I brought up the hb25s, is so I could use the BOE BOT book to learn from, since the BOE BOT uses CR servos, I figured using something like an HB25 would let me use a BS2 would allow me to follow the BOE BOT Robotics book line for line. I understand PWM based speed control and vasic open loop stuff to get my little guys rolling around. But when it comes to running a figure 8 or driving in a circle based on a pings readings im stumped! This is kind of why I wanted to go through the BOE BOT book. Can anyone else recomend a less inexspensive solution for CR type motor control? I could just use a micro to recive RC pulses then scale them to a regular old PWM signal, but thats a project in it's self.

    NwCC, im sorry been meaning to mail you back, just been extremely busy. Maybe we can work out something on an HB, when I get time to write a proper mail.
  • NWCCTVNWCCTV Posts: 3,629
    edited 2013-01-25 11:29
    rwgast_logicdesign wrote: »
    NwCC, im sorry been meaning to mail you back, just been extremely busy. Maybe we can work out something on an HB, when I get time to write a proper mail.

    Not a problem. Just let me know if you decide to go that route. I still have everything that I bought when I purchased the HB-25's. As long as my business stays busy I am not too concerned about recouping the money I had spent. However, if things get slow again I will be putting some of the items on EBay and or the classifieds here.
  • ercoerco Posts: 20,255
    edited 2013-01-25 11:39
    A few Scribbler 1's on Ebay now in the $40 range. that's a ready-to go mobile robot with motors and many sensors. At that price you can afford to hack them and learn a lot in the process. Old Blue is tried & true!
  • NWCCTVNWCCTV Posts: 3,629
    edited 2013-01-25 11:52
    @erco is right, they are quite cheap right now. I am not sure what major difference there is between an S1 and an S2 other than the Prop, but you could probably hack it and make it an S2 if the components are the same or close to it!

    Edit: I just found the difference. It looks like the S2 has a few more components that could be easily incorporated in to the S1 along with the Prop. Although not listed, the S1 also has the Pen Port and Hacker Port.

    S1 Specs:
    3 photoresistor light sensors
    2 Infrared object sensors
    2 Infrared line detection sensors
    2 independent DC motors
    Stall sensor
    Speaker with full range of notes
    3 LED indicator lights

    S2 Specs:
    • 3 light sensors
    • 2 obstacle avoidance sensors
    • 2 line following sensors
    • 2 independent DC wheel motors
    • Wheel encoders for precise maneuvers
    • Stall sensor for wheel motors
    • Pen port for drawing on paper
    • Speaker to make a full range of notes
    • Programmable indicator lights
    • Microphone for detecting tones from other S2 robots
    • Bi-color LEDs for visual feedback
    • Hacker port for connection to external sensors, RF devices, and servos
  • RobotWorkshopRobotWorkshop Posts: 2,307
    edited 2013-01-25 12:59
    Well, since it is for your birthday you could aim high and ask for an Eddie platform.....

    If you want a mid size robot then the Stingray chassis kit is nice. With it on sale now it is a good solid platform to build on. Or if you want one that is ready to go I would probably go with the S2 robot. There were some pretty cool examples using the S2 where it could draw logos or word from data in an Excel spreadsheet.

    You may want to ask for the Stingray or Scribbler S2 as a birthday gift. If you want a Boe-Bot I've sure you could probably find a pretty good deal on a used one.
  • ercoerco Posts: 20,255
    edited 2013-01-25 14:09
    For small bots, the math is pretty simple:

    Two HB-25 modules for 2 motors=$100 (16V,25A)

    One L298N for 2 motors=$4.18 (35V, 1A/motor)

    http://www.ebay.com/itm/L298N-Dual-Stepper-Motor-Driver-Controller-Board-Module-5V-For-Robot-Smart-Car-/180793346663?pt=LH_DefaultDomain_0&hash=item2a181f8a67
  • rwgast_logicdesignrwgast_logicdesign Posts: 1,464
    edited 2013-01-25 14:41
    @Ereco, like I said ive got plenty of h-bridges around. Doesnt an l298 only push 3A? Matt also said the l293 will work. The included motors stall @ 4.5a, wouldnt this smoke either of these chips?

    The only reason for the hb25 is to make the thing BOE compatible. I didnt realize they only drove a singal motor though!
  • rogersydrogersyd Posts: 222
    edited 2013-01-25 15:34
    I have owned them all at one time or another. Today I keep both a boe chasis and a sting ray available for tinkering. I owned a S2 at one point, but blew it up. I would suggest that the S2 is great if want a stock bot, ready to roll, without purchasing additional hardware. If versatility and horsepower (sort of speak) turn you on, then the Stingray is the winner. The BoeBot platform fits very well between the two. Tough choice my friend! If you are serious about the pan and tilt route, the Stingray is probably your most practical option simply due to its deck space.
  • Martin_HMartin_H Posts: 4,051
    edited 2013-01-25 15:37
    rwgast_logicdesign wrote: »
    @Ereco, like I said ive got plenty of h-bridges around. Doesnt an l298 only push 3A? Matt also said the l293 will work. The included motors stall @ 4.5a, wouldnt this smoke either of these chips?

    The only reason for the hb25 is to make the thing BOE compatible. I didnt realize they only drove a singal motor though!

    The proper engineering approach would be to include some 3A fuses in the event of a prolonged stall. Even without that many H bridge chips include a thermal shut down that will get activated when the chip overheats during an overcurrent condition. Worse case scenario is that you are out $4 which is cheap compared to over engineering.
  • ercoerco Posts: 20,255
    edited 2013-01-25 17:36
    For my 24V motors, the 298 is perfect. Their stall current is 1A@24V, average running current is far less. Per M_H, a little planning/engineering up front leads to a better and more cost effective solution.
  • ercoerco Posts: 20,255
    edited 2013-01-28 17:34
    An S1 for $27 (incl ship) would be sweet... http://www.ebay.com/itm/281056416584

    Somebody get it before I do!
  • Duane DegnDuane Degn Posts: 10,588
    edited 2013-02-07 21:56
    rwgast asked some question in a PM. He said it would be okay for me to quote him in this thread in order to make easier to answer his questions.
    rwgast_logicdesign wrote:
    . . . ive been looking at every possible robot I can purchase. . ..
    Since I believe you own both of these bots (you do own a sting ray correct). . ..

    I don't own a Stingray, but I do have a Rover 5 and a bunch of other robots.
    rwgast_logicdesign wrote:
    The first would be a stingray, with homeade encoders. Speaking of homeade encoders do you see any advantage to using one of those digital TSL light sensors parallax sells over a standard photo transistor/schmitt trigger?

    I think the TSL sensors would be too slow to use with encoders. You just need to know if the sensor is sees light, not how much. The TSL sensor is very useful for other light sensing applications though.
    rwgast_logicdesign wrote:
    The second is a rover 5. . . . I would be adding these wheels https://www.sparkfun.com/products/10555 using two ebay l298 driver boards and . . . some micro servos for pan/tilts and a . . . IMU.

    As I've mentioned in the Rover 5 thread, I'm still not convinced either way about the Rover 5. It is a relatively inexpensive way to purchase four geared motors with encoders, but it does have some drawbacks.
    rwgast_logicdesign wrote:
    Im really up in the air because they both have there ups and downs. The rover will be able to travel outside, and eventually to GPS way point navigation, which is a big deal to me. Ive seen a video of this thing with a nice expansion platform carrying about everything I would ever need to straped to it, and it managed to get up a pretty steep incline.

    Now I get the feeling a stingray will be allot easier to learn with, although my roomba chassis is basically the same set up. The stingray is also allot more rugged and in general a higher quality product, but that omni wheel makes it indoors/sidewalk friendly only, just like the roomba correct?

    I agree, this is a tough call. I'm not so sure I'd concider the Rover 5 an outdoor robot. I also don't think the omni wheel caster on the Stingray would keep it from working outside. I probably wouldn't use it to go off roading but I'm pretty sure I've see video of people using the Stingray outside in the street.
    rwgast_logicdesign 2ndPM wrote:
    This is a very tough choice for me, as about 80% of the things I want to do with a robot are pretty much indoors and dont translate well outside, this would include fixed and moving beacons, spacial mapping using wii cam's and lasers, etc. But the last 20% of outdoor use may be where the bot ends up getting the most use, I.E, waypoint navigation using multiple systems, animal monitoring using beacons.
    . . .
    Hey I also forgot to ask, you having taken apart a rover, does it look feasible to control the leg incline using servos? So the clearance can be st on the fly.

    Short answer no. You can kind of see how the motor/gearboxes are held in place in this post. You'd need some sort of custom pivoting system to be able to move the gearboxes on the fly (I'm actually working on something like this with the bot I'm building now (not a Rover 5)).
    rwgast_logicdesign wrote:
    BTW Id really like to use your PWM/Encoder code on my roomba, especially if you can set it up to work like a CR servo. The whole point of the roomba is to have a very easy to play with platform to learn on, that im not shy about damaging. I could run through all the BOE bot material using it.

    I would be great if my code made it easier for people to use BOE-Bot code in robots with h-bridge motor controllers instead of CR servos.

    (Answers to a few other questions.)
    I haven't looked for a new source of Nordic nRF24L01+ modules yet. I think Gareth just posted about a vender he is using. I'll need to edit my index to fix the dead link.

    I have one of those LinkSprite camera from SparkFun, but I don't know how to use it with a Propeller. I think trying to capture JPG files would be more work than it would be worth and I don't know how to capture raw images (or if it's possible to do so).

    Besides the NTSC video capture options, the CMUcam4 and the Parallax Range Finder are the two most promising image capture options I know of for the Propeller. As has been described many times, a big problem with trying to capture an image with the Prop is where to store it.

    I found the following in the text editor when I first began this message. I think it from a post I never submitted. I think it dates back to when you first started this thread. I decided some of the information may be useful.

    BTW, The bot I mentioned earlier (with rotating motors and gearboxes) will be using the Pololu motors mentioned below (they seem to work well now that I've had a chance to test them a bit more). The robot will also likely be using two sets of Vex treads (I hope the real version is as cool as the vapor version in my head).

    *** restored text ***

    The main disadvantage I see with the Stingray is the lack of encoders on the motors. I'm pretty sure these encoders from Lynxmotion will fit on the Stringray's motors so you could add encoders later.

    I think I've mentioned before Pololu's motors with encoders. I picked up four during the Black Friday sale. I've just barely started testing them out so I can't really comments on how well they work.

    The Scribbler has a hacker port but it uses a level shifter that interferes with some devices. The Scribbler also needs a very flat floor to work properly. I've had mine get stuck on carpet to vinyl transistions.
  • rwgast_logicdesignrwgast_logicdesign Posts: 1,464
    edited 2013-02-07 23:04
    I decided to go with the stingray in the end, for quite a few reasons. I read more from the guy with the tricked out rover 5 using Dagu off road wheels and found out even though there advertised to work, you have to cut the rovers axels, this means you cant go back to treads. The axels are also long and soft that he bent one shortly after shooting his video, and is working on a way to install better axles and and wheel hubs (fussing with this kind of stuff is what im trying to avoid, ive got project chassis, I want a reliable one)

    Here is the bot i was inspired by, vids and all, its really nice just needs to much modification to make it rugged enough for real use with off road wheels
    http://arduino.cc/forum/index.php?topic=89196.0

    The sting ray is just a really nice looking bot, has no problems, and encoders are easy enough to make for it, already have all the stuff on hand. The Stingrays motors are also a lot nicer/rugged than the rover's. I acually thought about it for a while and the stingray could be a decent off road chassis, all one needs is to swap to off road wheels, buy another chassis later and mount the two togather back to back to make a 4wd rover. Basically the stingray will be nice to learn with and can be upgraded to a rover with very little work.
  • Duane DegnDuane Degn Posts: 10,588
    edited 2013-02-08 07:43
    rwgast_logicdesign wrote: »
    I decided to go with the stingray in the end, for quite a few reasons. I read more from the guy with the tricked out rover 5 using Dagu off road wheels and found out even though there advertised to work, you have to cut the rovers axels, this means you cant go back to treads. The axels are also long and soft that he bent one shortly after shooting his video, and is working on a way to install better axles and and wheel hubs (fussing with this kind of stuff is what im trying to avoid, ive got project chassis, I want a reliable one)

    IMO, you made a smart choice. The axles on the Rover 5 are soft. It's something I forgot to mention in the Rover 5 thread.

    The Rover 5 in the video is pretty cool. It looks likes he made good use of the Arduino's shield concept. I'd be surprised if he's reading the encoders with an Arduino (I think four quadrature encoders is too much for an Arduino to handle).
  • Martin_HMartin_H Posts: 4,051
    edited 2013-02-08 08:08
    The modified rover 5 bot is pretty neat, but given the required amount of work he might have been better off scratch building. From the video it sounded like he's using an Arduino mega which has a little more oomph than an Arduino
  • Duane DegnDuane Degn Posts: 10,588
    edited 2013-02-08 10:12
    Martin_H wrote: »
    The modified rover 5 bot is pretty neat, but given the required amount of work he might have been better off scratch building. From the video it sounded like he's using an Arduino mega which has a little more oomph than an Arduino

    Even though I have two Rover 5 bots, I'm still undecided about recommending them or not. It is a quick way of getting four motors with four encodders but the drawbacks mentioned can put a damper on the fun.

    I don't know much about the various Arduinos. One reason I suspect an Arduino (even a Mega version) would have trouble with the encoders is based on Cliff's experience.

    I do think driving all four motors on the Rover 5 while reading all four quadrature encoders is a good example of why the Propeller is such a great controller for robots.
  • rwgast_logicdesignrwgast_logicdesign Posts: 1,464
    edited 2013-02-08 13:12
    If you read about that rover, it on a few other sites, diydrones, lmr, etc, He had not implemented the encoders yet and it had problems getting through depressions, this is from a lack of suspension, and traction control. Basically one tire will lose grip with the ground and spin full speed hogging all the current, encoders could fix this fairly easily. Leaving tracks on this thing and running the 4wd version is what seems to cause alot of the problems with the tracks slipping. I was thinking a two wheel drive version would probably be the way to go if your leaving the tracks on it. From what I had read throwing wheels on the 4wd chassis solved alot of issues, but then it turns out the axels are really long and soft, which means they need to be replaced if one wants to use wheels :/. I was under the impression I would remove the tracks and rollers and then screw the new wheels on, this is not the case. Not only that but switching the wheels also doesnt totally fix the problem with these chassis coming boed from the tension of the treads.
  • ercoerco Posts: 20,255
    edited 2013-02-08 14:48
    Actually encoders would not fix that all-terrain problem at all. The untold truth about mecanum wheels is that they only function properly on perfectly flat terrain, and they still need to have a chassis pivot or suspension to distribute the weight equally and make sure all wheels are touching the ground. As soon as one wheel lifts or loses traction, random motion occurs. At best, an encoder might indicate a wheel lift condition, but could not provide any useful motion data.
  • Duane DegnDuane Degn Posts: 10,588
    edited 2013-02-08 16:50
    erco wrote: »
    Actually encoders would not fix that all-terrain problem at all. The untold truth about mecanum wheels is that they only function properly on perfectly flat terrain, and they still need to have a chassis pivot or suspension to distribute the weight equally and make sure all wheels are touching the ground. As soon as one wheel lifts or loses traction, random motion occurs. At best, an encoder might indicate a wheel lift condition, but could not provide any useful motion data.

    Erco, I'm inclined to disagree about much of the above.

    I think in the encoders would have helped with the trouble the "all-terrain" Rover 5 had. If the robot could sense the wheel spinning it could give reduced power to that motor leaving enough current for the other wheels to move the robot free from the obstacle.

    Rich previously corrected my mistaken assumption about Mecanum wheels having trouble in an off-road environment. I do agree a lifted wheel could cause course disturbances, but I don't think Mecanums are necessarily problematic in rough terrains. (Remember Rich's robot driving on top of the Roomba? And my Mecanum driving over dinner plates without much trouble.)
  • rwgast_logicdesignrwgast_logicdesign Posts: 1,464
    edited 2013-02-08 17:26
    Ya that was my thought on the all terrain wheels, detect if one motor is at max speed while the rest arent, then apply appropriate traction control. I dont understand why the higher end Lynxmotions and 3-500 dollar rovers dont include suspensions. Im almost tempted to say a decent RC chassis is the way to go for all terrain. While the rack and pinion make things a bit harder to deal with the suspension is worth it.

    As far as mecanums in any all terrain situation im not sure I would trust the rollers to not get dirt in them and get stuck.. but since I dont own a pair I this is just an assumption.
  • rwgast_logicdesignrwgast_logicdesign Posts: 1,464
    edited 2013-02-08 17:31
    As an update to this, I just purchased a stingray with some ebay servos for a pan/tilt and nordic wireless modules. Weird thing is the ebay Nordics dont have any pins labeled, nor does the buyer say what the pins do, but there are a lot of sellers with the same Nordic modules.

    http://www.ebay.com/itm/1X-NRF24L01-2-4GHz-Antenna-Wireless-Transceiver-Module-For-Microcontroller-EA-/350703013778?pt=LH_DefaultDomain_0&hash=item51a7874f92

    anyways I think I made a good choice, already have a Nimh pack/charger, h-bridge, power supply, etc...
  • ercoerco Posts: 20,255
    edited 2013-02-08 17:49
    @Duane: Fact: with one mecanum wheel lifted, the bot will steer completely differently. Sure it can stumble along randomly in any terrain, but the mecanums lose every advantage. Not so bad if you're driving it like an RC car, but for an autonomous robot, that's real trouble.

    I'll stand by assertion that encoders bring nothing to the offroad party. Your move, my learned opponent! :)
  • TtailspinTtailspin Posts: 1,326
    edited 2013-02-08 18:21
    You have made a great choice!

    Please, Please be sure to film the first few seconds of powering up your new Stingray.
    And do Remember to always check your backdrop when you take aim... :)
  • ercoerco Posts: 20,255
    edited 2013-02-08 18:32
    Ttailspin wrote: »
    You have made a great choice!

    Please, Please be sure to film the first few seconds of powering up your new Stingray.
    And do Remember to always check your backdrop when you take aim... :)

    Range, elevation... FIRE! :)
  • suspect_devicesuspect_device Posts: 4
    edited 2013-03-19 22:39
    I just bought a Stingray and am using my BOE Shield with an Arduino MotorShield stacked on it to control the motors. It's cheaper than the HB 25s.
    IMG_20130320_013424_zpscb3e1937.jpg
    IMG_20130319_150107_zpse3aa7ffe.jpg
  • rwgast_logicdesignrwgast_logicdesign Posts: 1,464
    edited 2013-03-20 00:12
    Your stingray looks nice! What kind of sonar sensors are you using? Im just curious if they are the e-bay kind I like the way you mounted them.

    You may want to look in to a different motor driver though, that shield uses the l28 if im not mistaken, the sting ray motors stall at 4.5 amps which is over an l298's rating if im not mistaken.
Sign In or Register to comment.