Comments on my Power and Chasis plan

[Justin]

This is my first robot, but I am an experianced programmer.  I'm going to keep it simple at first, but definately plan on buying some cool sensors and experinmenting with the mechanical side of things sooner rather than later.  It would be sweet if I could build something in the next year that could follow me on a dirt trail.  I'd like to get your comments on my design, before I take it much further.  I'm particularly interested in what you think of my battery choice, and what I might expect out of those servos performance wise on a bot this size.

Below is the link to my planned parts.  The only things I already ordered are the wheels, and a couple sheets of 3/16ths HDPE

https://spreadsheets.google.com/ccc?key=0AlpWF1Dft7l8dDlYeXZNTG9fRUUwdThLVXdfdkl3Y3c&hl=en&authkey=COPT2cQJ

Here is the link of my 1st attempt ever at a 3d design.

https://docs.google.com/leaf?id=0B1pWF1Dft7l8NDRiN2Y0MjAtNGY1Ny00MDNjLTg5MGMtNjU2NjhhNjg5OGIz&hl=en&authkey=CLrZmeQE

[rbtying]

I can't seem to download your 3D design, so I can't comment on it.  Your battery choice is fine; however, it would be to your advantage to get a SMPS to regulate it down to 6v for the servos - 7.4v LiPo batteries have fully-charged voltages of 8.2v or so, which will damage servos.  The commonly available RC BECs are generally suitable.  You will, if you use the $50 robot tutorial, need to modify it - a 7805 regulator will not be able to regulate once the voltage drops under around 7.5v, so you should use a low-dropout regulator instead. 

Regarding your goals: servo-based drive systems are fine for indoors, but generally aren't great for outdoor environments, since they're limited in their power (they're not that fast, either).  You may want to transition to more powerful gearmotors instead, since they will have higher torque and speed. 

[Justin]

I created a new link to the sketchup doc, hopefully it works for people now.  Thanks for the advice on the geared motors.  I do plan on following the $50 robot tutorial, but while waiting for the parts for that bot, I couldn't help but start to plan ahead and decide what upgrade parts I will want for version 2.0.  That way when I'm ready for them, they are ready for me.

As for modifying the design, I'm sure this belongs more in the electronics sub forum, but in the interest of reducing spam, I'll keep it in the same thread.

I really have no clue about electronics, I found a circuit simulation program so that I can practice, but I have no idea where to start.  I'll google for some tutorials, but if anyone has a link to one that they particularly like, or has a high relevance to the type of electronics design/troubleshooting most applicable to someone starting in robotics, I'd appreciate it.

[Soeren]

Hi,

[...] build something in the next year that could follow me on a dirt trail.

Quite ambitious.
Start with an indoor 'bot and when you cannot learn more from that, you may be able to take on the many extra challenges a dirt trail will be.

 I'd like to get your comments on my design, before I take it much further.  I'm particularly interested in what you think of my battery choice,

Good for the servo (and cheap), but...

and what I might expect out of those servos performance wise on a bot this size.

On a level plane, using two servos it should be able to go the stunning speed of 0.958km/h if it weighs a maximum of 1.75kg (with a very slow acceleration).
Even a senior citizen with a walker could outrun it.

A servo, however good it may be, aren't build to handle a 4" wheel in terrain and will break down in a short time.

Selecting a battery should be done after you have selected the motor(s).
The motor(s) should be selected for the weight of the 'bot, how fast you want its top speed and the wheel diameters.
The wheels should be selected after the total weight is known - too much weight on a wheel and it flattens and need more torque, too little weight and it will have less friction.
(OK, I'll stop now before you get discouraged  )

Here is the link of my 1st attempt ever at a 3d design.

https://docs.google.com/leaf?https://docs.google.com/leaf?id=0B1pWF1Dft7l8NDRiN2Y0MjAtNGY1Ny00MDNjLTg5MGMtNjU2NjhhNjg5OGIz&hl=en&authkey=CLrZmeQE

The link should be:
https://docs.google.com/leaf?id=0B1pWF1Dft7l8NDRiN2Y0MjAtNGY1Ny00MDNjLTg5MGMtNjU2NjhhNjg5OGIz&hl=en&authkey=CLrZmeQE

Gives you a page saying:
ROBOT CHASSIS.skp
No preview available
Download (43 KB)

But it wouldn't let me get the file.

[Soeren]

Hi,

As for modifying the design, I'm sure this belongs more in the electronics sub forum, but in the interest of reducing spam, I'll keep it in the same thread.

Reducing "spam" would be to post electronics in "Electronics", mechanics in "Mechanics", software in "Software" etc. It would better your chances of getting usable answers too.

I really have no clue about electronics, I found a circuit simulation program so that I can practice, but I have no idea where to start.  I'll google for some tutorials, but if anyone has a link to one that they particularly like, or has a high relevance to the type of electronics design/troubleshooting most applicable to someone starting in robotics, I'd appreciate it.

Start with the microcontroller that you're gonna get anyway and start with a "Hello LED" program (the microcontroller version of a "Hello World" program).
Then each time you need to add components, read up on how they work and such. This will enable you to take it in a little at a time (and there's a lot that you don't really need to know to build 'bots).

As for the programming, you may have to adjust a bit. With PC programming these days, you may get away with not knowing the underlying hardware - on a controller you need to know every bit of it and you'll need to have the datasheet for that particular controller lying next to the PC.
Print an extra copy and hang it in your bathroom - you might as well learn how to make a memory dump while... 

[Justin]

[...] build something in the next year that could follow me on a dirt trail.

Quite ambitious.
Start with an indoor 'bot and when you cannot learn more from that, you may be able to take on the many extra challenges a dirt trail will be.

I tried to think of something ambitious enough to push me, and keep me motivated, but not so hard that it would take me 10 years and 50k.  I love to hike and bike the trails behind my house, and once I had that foundation, I could think of TONS of fun things to experiment with on the platform.  Would it be possible to ballpark total cost assuming a target climbing speed of 3mph, a 30 degree slope, and medium pack dirt?

and what I might expect out of those servos performance wise on a bot this size.

Selecting a battery should be done after you have selected the motor(s).
The motor(s) should be selected for the weight of the 'bot, how fast you want its top speed and the wheel diameters.
The wheels should be selected after the total weight is known - too much weight on a wheel and it flattens and need more torque, too little weight and it will have less friction.
(OK, I'll stop now before you get discouraged  )

Not discouraging at all!  It is so great to have this community to interact with.  What is discouraging is not knowing what the variables are, I can handle the fact that there are a bunch of them .  Following your advice on selecting a motor first, do you have any pointers?  I'm half tempted to just stick a microcontroller and some sensors on a RC truck chassis when I'm ready to start this project.

Here is the link of my 1st attempt ever at a 3d design.

https://docs.google.com/leaf?https://docs.google.com/leaf?id=0B1pWF1Dft7l8NDRiN2Y0MjAtNGY1Ny00MDNjLTg5MGMtNjU2NjhhNjg5OGIz&hl=en&authkey=CLrZmeQE

The link should be:
https://docs.google.com/leaf?id=0B1pWF1Dft7l8NDRiN2Y0MjAtNGY1Ny00MDNjLTg5MGMtNjU2NjhhNjg5OGIz&hl=en&authkey=CLrZmeQE

Gives you a page saying:
ROBOT CHASSIS.skp
No preview available
Download (43 KB)

But it wouldn't let me get the file.

Thanks, I'll see if i can host it somewhere else and fix the links.  Stupid google

[WaterPig Master]

Hi there,

You have chosen a very ambitious project, IMHO! Outdoor robots are trickier to say the least. Indoors you are working with a detailed and complex environment — still unpredictable but less so — and at least it is a little closer to the theorised 'perfect' environment. It'll certainly be a few years before I even contemplate an outdoor robot. But, if that's what you're interested in, go for it.

Following your advice on selecting a motor first, do you have any pointers?  I'm half tempted to just stick a microcontroller and some sensors on a RC truck chassis when I'm ready to start this project.

Sounds like you'll need reasonably sized motors, which mean big batteries, which mean more weight, which means more powerful motors, etc. How much money are you looking to spend? I don't think servos will get you far in this. You'll need some lightweight, efficient motors. Are you intending on gearing them down yourself?

You could well stick a MCU on a RC truck, but good luck getting sensors to register anything other than vague values when the robot's bumping around at quite high speed. You can get nice tracked chassis with motors built in that might be suitable for what you're trying to do? I've seen some good looking systems advertised here on the forum.

Thanks,
Barnaby