Fallentines Tracked Rover... Almost complete. Pics+ Video!

Lynxmotion Rovers, Sumo, and Tracked
1

Here are some pictures of my rover, took them a couple weeks ago, i have a brand new 900mhz spy camera, but i haven’t decided where to put it. It’s a tiny camera, and it’s a small robot, but sadly the camera is still a little to big to place where i really want.

I’ll upload a video eventually, but i’ll do that after i fix my laptop and put the cam on my robot. Got a tv tuner for my laptop so i can control the bot wirelessly, and view the feed on the laptop.

It uses a bot board 2, and a Basic Atom pro 28m.
Arm is 5 degrees of freedom, base left/right, sholder, elbow, wrist up/down, wrist rotate.
Then the gripper which doesnt count as a DOF

Edit: video added aug 29th 2010


endedit

img294.imageshack.us/img294/1985/dscf0045g.th.jpgimg17.imageshack.us/img17/6245/dscf0039x.th.jpgimg805.imageshack.us/img805/9823/dscf0044.th.jpgimg411.imageshack.us/img411/3582/dscf0042o.th.jpg

The underside, the battery is held on by velcro, i haven’t found a better solution yet, and i’m a bit lazy.
img12.imageshack.us/img12/8033/dscf0041je.th.jpg

My 5v regulator and my pololu micro controller, will replace it someday… for now it works just fine.
img401.imageshack.us/img401/1962/dscf0037l.th.jpg

From left to right on the bread board.
Bluetooth uart adapter, 9600bps, i’ll be replacing it eventually with one that accepts AT commands, or buy an SPI programmer.
3.3v voltage regulator.
Red thing is a sparkfun 3.3v to 5v bidirectional TTL voltage converter, Bluetooth use 3.3v ttl, the botboard/basicatom uses 5v, weak signals caused problems before.
The white wire you see is my beefed up ground wire.
img121.imageshack.us/img121/1052/dscf0036wt.th.jpg

Video of it in action will come later.

Edit: Bonus images. My work area. It’s actually much messier now, i have a lot more projects going on.
img197.imageshack.us/img197/2359/dscf0048cm.th.jpgimg717.imageshack.us/img717/572/dscf0047u.th.jpgimg3.imageshack.us/img3/7740/dscf0049rj.th.jpgimg709.imageshack.us/img709/1751/dscf0046o.th.jpg

2

Heres some brand new pictures of my robot with the spy camera, it’s not attached, i just wanted to show you the size, so you’d understand why i’m having trouble picking a good spot.
My choices are, rewire the power wire so it uses a much thinner and lighter wire, then mount the camera on the side of the blue gripper servo.
Or use a second blue servo on top of the sholder servo then have the can tilt off to the side.
The robot will eventually have a nice array of 4-5 bright white LED’s and a laser module on the gripper to help aim.
Maybe even a high power dvd-burner laser module that can cut tape, light matches, pop balloons and what not. I’m just scared of equipping this thing with weapons… skynet and what not.

img3.imageshack.us/img3/4535/dscf0040r.th.jpgimg638.imageshack.us/img638/6236/dscf0041e.th.jpgimg210.imageshack.us/img210/15/dscf0042h.th.jpgimg838.imageshack.us/img838/7079/dscf0043.th.jpg

The second 2 photos, the camera is on a standard servo, to give you a perspective of size. It’s not a big camera, but it’s not a “micro” camera… Mini at most. Has fairly decent picture in medium to low lighting. Sadly the very low lighting might as well be a black image.

3

Update: While trying to find the best location on the bot for the camera, after spending about 30 minutes making a longer power cable for the camera, i came to a new problem. a very annoying problem…
The 900mhz camera interferes with my servos, causing them to vibrate violently.
This is very inconvenient. as i decided the best place is to put the camera directly on the arm.

The camera is made up of two boards internally, the camera module and a simple AV transmitter.

I could easily make an “extension”, separating the two, and putting the transmitter on some sort of spring supported stick.
But i’d probably still have interference problems with the servos.
I got 900mhz so it wouldn’t interfere with the bluetooth, i was sent a 2.4ghz cam by accident(identical boxes) and it disconnected my controlling laptops wifi, and caused problems with the bluetooth. So i sent it back and had a 900mhz camera exchanged. (i originally ordered 900mhz)

But i just can’t win, 2.4ghz glitched my laptops wifi and bluetooth, 900mhz glitches my servos, and 1.2ghz cameras are hard to find and are expensive, and still would probably glitch my servos.

Anyone have any ideas?

Edit: I’m definitely stuck. Tried making an extension, but the gauge of the wire is to heavy, the extension between the camera and the transmitter is stiff. And it barely solves the problem. Still get interference.
Doesn’t help the camera is so big. Also doesn’t help the view angle/field of view is bad. It’s a bit zoomed in, would need a wide angle lens, while still being very small.

If anyone can find a 1.2ghz camera that uses NTSC, with at least a semi wide angle lens, smaller than the camera i have, all for under $50, please don’t hesitate to list it.

-sigh- this robot is just bad luck, i just can’t seem to win with it.

4

You could try shielding with aluminum foil, being careful to cover exposed electrical components with paper or similar to prevent a short.

5

The interference comes mostly from the tip of the transmitter antenna, closer it is to the tip the more interference.
It would be rather difficult to insulate the BB2 and the servo wires.
I never got around to finding the right size hobby enclosure. (4"L/3"W/2"H)

From searching, it seems many types of servos have problems with 900mhz transmitters.
I’ll probably have to get a 1.2ghz, or a 5.8ghz spy camera. The latter would need a 1W transmitter to get the penetration i’d need. (no giggles)

BTW, if you ever wondered why those cameras have such massive power plugs, theres a voltage regulator in it.
The cameras take 5v, but the transmitters which are a separate board use 8-10v.
The size can be reduced if you use your own regulator and mount it to the side, or even inside.

6

Ok interference is a two way street. I think zoomy meant for you to shield the servo, not the camera. If the ingress is affecting the servo directly this should help. You would have to surround the entire servo case in aluminum foil and ground it with the shortest wire length possible. Another path for ingress is the servo wires themselves. You can reduce this a lot by changing the wires from flat to twisted, or shield them. I do not know if it would have the same effect if you twisted the flat wire. I think this can be dealt with. :confused:

7

You know about bypass caps for motors, you can also put bypass caps on the power leads to the R/C servos, the camera, etc. And twisting the wires can help, as can lead dress (don’t bundle all the signal and power wires together). Keep the camera power separate, if possible.

Alan KM6VV

8

I’d experiment by placing an electrical insulating material over the body of the bot (like a dish towel to prevent an electrical short), followed by a sheet of aluminum foil. Then move the cam to various places around the bot to see if the interference continues. Connecting the aluminum foil to the bot ground might also help.

9

Off topic, but interesting, a popcicle stick biped. I wonder if there is some hot glue in there?

youtube.com/watch?v=DlmUeHkb … re=related

10

man that bot can barely wobble let alone walk.

My robots arm is cheap and fully functional.
Good range of motion, but poor control though, i can refine it, just don’t have the time.

The signal interference happens when the antenna gets near (within 8 inches) of the 2 standard servos, or their wires.
and within 2-3 inches on the smaller servos.

Oddly enough, if the antenna goes near the small servo’s wires, it’s the 2 larger servos that freak out. Not the smaller ones.
Covering my whole bot with tin foil is out of the question. As for the individual servos, i may be able to that, and i could probably find some insulated wiring.

If any of you have a 1.2ghz camera, would you mind testing to see if it interferes with any of your servos. I may just go with a different camera, i want a much smaller one. Like of of those ones thats as small as a penny or nickel.

11

Don’t mean to pick, but you already have insulated wire, you are looking for shielded wire.

12

Do you know what RG-8 or RG-58 coax is? (Radio Shack).

This coax has a metal braid, and the cable can be stripped apart to yield the braid. This braid can be slipped over one or more servo cables, and will provide shielding. ground at one end only, usually at the driver end.

There are other places to get this braid also. Buy by the roll (expensive). Look for surplus network cables. Those 15-wire XCVR cables are typically shielded, as are a lot of cables you can find surplus. Short bits of coax are also easy to find.

Alan KM6VV

13

Ya that’s what i meant, a bit light headed last few days, been sick. Was thinking shielded, said insulated.

I have some old USB shielded wires. It’s an all day project, so i’m not gonna try till i’m feeling at the top of my game.
Spacing is still a factor now.
My bot isn’t as roomy as it looks.
I’m going to remove the wires from the ends, split them, then braid them, i’ll lose a bit of length but it should help a little with interference.
If that doesn’t work, i’ll replace the wires altogether with shielded wire.

Yup, definitely an all day project, cant do it now, reflexes are shot, and i have butter fingers. But a couple days when i’m feeling better i should be able to tackle this no problem.

14

Well none of it worked, the braided wires looks really nice though.

shielding the servos and the wires didn’t help. Only thing that would help is to shield the antenna, which is obviously out of the question.
That or aluminum foil shielding the entire bot, wires, and arm included.
Anywhere the antenna gets close to any part of the bot that has any power, ground, voltage or signal, it causes some of the servos to freak, even if a little bit, and since the whole bots chassis is grounded i think, moving the antenna within 5 inches of the bot, caused servos to freak.

Which means my only choices are 1.2ghz, or 5.8ghz spy cameras. 2.4ghz wont work cause it glitches with the bluetooth and wifi.

15

Have you grounded one end of the shield braids? Have you added small bypass caps to the power leads?

You can also generally add quite small bypass caps to the data lines.

These have all been mentioned and discussed before, and are known methods of keeping RF out of signal and power lines.

Alan KM6VV

16

So a cap between the positive and ground voltage, and a second cap between ground and signal?
I assume since all servos share the same voltage lines i only need 1 cap for the whole array.
Then each servo gets an individual cap for the signal lines.

I found separating the entire bot from the antenna with a simple grounded foil sheet, didn’t even work, as the RF interferes with the ground. It caused all servos to freak.

Would a 0.1uf cap work? i have plenty of ceramic capacitors i can salvage, just let me know what uf to use, or what number i should look for on the capacitor. ie 104’s for 0.1uf.

My budget wont allow for a new camera for a few months, so if i can get this 900mhz camera working, awesome.
Having a bit of trouble getting the shielded wires.
Been using wires i salvaged from other stuff, but getting enough, and the right size is becoming a pain in the butt, but that’s my problem.

I’ll probably be redesigning most of the bot in a little bit, after i get more of those MG90 servos, small and super strong. MG90S’s might work, the S’s are $1.50 cheaper, but they have sleeve bearings vs the MG90’s ball bearings. I don’t know if it makes that much of a difference 2kg/cm sleeve versus 2.2 to 2.4kg/cm for ball bearing.

17

.1 caps near the power pin, if you have a few servos, you can even put them right across the bottom of the connector. Put .001 caps on the servo data lines, (test), shouldn’t affect the data. Put caps on the camera power lines. Are the camera and servo power supplies isolated? Have you grouped the servo lines together as much as possible, and kept them near the metal chassis? The camera lines could also be routed close to the chassis, (put away from the servos). Don’t forget to ground one end of the braids.

Alan KM6VV

18

Thanks.
I’ll try a 0.1uf(104) cap on the power, and .001uf caps (102’s) on the primarily effected servos and test that.
As far as the braid, i’m still having trouble finding the right kind, it’s not as easy as i originally thought, but no problem, i’m sure i can find some.

The camera is on completely separate power at the moment. They were originally supposed to share power, but i think i need to get it working as is, before i re-modify it again.
I returned the camera back into it’s original unmodified state.

I’m still a bit sick so i’m taking my time working on it. Robots don’t like to be sneezed/coughed on.

Edit:
Well… it didn’t work.
The 104 capacitor over the power helped a tiny bit. but the 102’s between the grounds and the signal wires did nothing, it may have even made it worse.

The shielded wire (i tested it on 1 servo and only had 1 servo connected) cut the interference down to about 25-35% of what it was.
Which is pretty good… but not enough. For it to work i’d need to get rid of the interference completely.
Since the camera gets attached to the arm, and this is a tiny robot so everything is within just a couple inches of eachother. It just didn’t work.
The interference of having the camera within 3 inches of the servo caused it to glitch. I don’t have an oscilloscope to do other testing.
If anyone has a 1.2ghz camera, mind testing it near a few active servos, see if it glitches at all?

Edit again:
What about ferrite chokes? I see them on just about every device, wouldn’t those help a little?

Edit some more:
Well ferrite chokes helped a bit. 1 servo seems pretty stable now. The wires are MUCH less sensitive, but i don’t have enough shielded wire to work in addiction to chokes, but the chokes seem to reduce the wire sensitivity by like 80%.
Does the color of the ferrite choke matter? I’m salvaging them from old fried motherboards, so i wanted to know if the color matters.
The servo with the black painted choke seems to be doing a little better than the servo with the green painted choke.
They are really hard to get off without breaking them… Like 5+ capacitors in the way.

I’m only able to get 3 wraps with ferrite chokes and braided wires. Any more and i lack enough length, and it simply doesn’t fit.
Capacitors had no effect.

Right now i have chokes and braided wires… Still getting to much interference to make it safe to use at a range.
My goal is to use my bot in a few air ducts. So i absolutely can’t have it freaking out.

19

It would be up to the choke mfg what the colors meant. I got some from SuperDroids for DC motors, should be useful. Even then, they didn’t have specs.

The robot chassis is aluminum, right? use it as a ground at one point for the (-) power to the motors. Tie the braid to the chassis, one end only.

Any more then that, and I’d need some really close-up, well illuminated pictures to get a better idea of your setup, and what to suggest.

Sometimes you can make up a “probe” of turns connected to a shielded cable into an audio amp.

Yeah, eliminating noise in mixed systems is more of an art then a science!

Glad you mentioned the ferrite chokes, I’d forgotten about them.

Alan KM6VV

20

It might be worth reading up on the operation of ferrite chokes, and see if something similar can be made using heavy iron based wire from the local hardware store.