I’ve been wanting to do a foot sensor for the brat for a long time now. What I’m envisioning is a PC board with the exact same outline as the ARF-01. Either 3 or 4 micro switches like the ones on the reset and ABC inputs for the Bot Board. These switches would be the supporting points (one in each corner) for the foot. The board would have an LED for each switch also in each corner. I was thinking it would be cool to use a resistor ladder to make it possible to read the foot sensor data in from the SSC-32 or Atoms serial port. In fact it could be connected with a servo extender cable. Plug and play city… Please don’t hit me with a barrage of posts asking for I2C or other active solutions. I want this to be a simple and cheap board. The board will have the SES hole pattern right on it so it could be used in place of the ARF-01 for a stock BRAT, or added to the ARF-01 for a larger biped. This board would make the BRAT able to stay on a table top and know better than to walk off. I need to find out the force required to throw these micro switches to make sure the BRAT can work with them. Beyond that it sure looks like a no-brainer… comments? 8)
I’m on it… What’s your flavor for gerbers? 
You read my mind… lol
Let me find the force required for the 3 - 4 sensors first, then we can make boards. 
Ok, standing by sir… 8)
Maybe the forum noise fairy should clean this up. lol
Huh? I can’t hear you! speak up Jim.
All I know is every minute is dragging at work. I got parts waiting to be installed and I just can’t take it anymore! This is like Chinese water torture!
Aaaaahhhhhhhhh!
Is it that bad really? I mean I just mis-understood your post when you mis-understood Dale’s post who had mis-understood Jim’s post where he asked not to be bombarded by requests for active solutions like I2C… so it’s all Dale’s fault really. 
With only 4 switches you should get enough resolution from a R-2R ladder. I’m not sure I got the bit about replacing the ARF-01 with the PCB though, wouldn’t the switches be exposed to being sheared if they made direct contact with the tabletop? Either way it seems like you would be looking at a sub 100g actuator force on the switches, assuming flat footed you expect to have all 4 switches activated.
LOL, that’s right! Dale IS the root cause of all this madness. 
J/K Dale…
Back to the switches, each will have a certain amount of resistance, not talking about the electrical kind of resistance, but the physical kind. You have to press a certain amount force on them and if 4 are used it will be that amount x 4 which will increase the force required to press all four. My guess it with all for flush on the surface, they will all be in an open non-closed state cause you have the other leg as well so there are going to be eight little guys working together to keep the brat held up.
Huh, you still here… lol
Oh I found 100gf switches. So if we do 4 switches per foot, we would need 800gf to actuate all 8 of the switches. The bot weighs 1.95lbs (885g) with batteries etc. so it’s looking doable. The leg positioning can be altered to ensure adequate downward pressure when the foot is lowered to ensure the switches can enguage.
After thinking about it the board will have to have some active components to turn on the LED’s… I will whip up a schematic…
roflmao!
You got it exactly as I was planning it. 8)
The BRAT is relatively light and I had thought if the board were made with heavey duty weight PC board material it could be used inplace of the ARF-01. The switches would be exposed, but I don’t think there would be enough force to shear them off. Also I was thinking of this as a tabletop project and not intended for outdoor use. I also found this page on R-2R resistor ladders. Brings back memories.
en.wikipedia.org/wiki/Resistor_Ladder
Noooooooo, it wasn’t me! 
I truly did not remember that capacitive sensor being I2C. Besides, I specified something like that, not that exactly. I am just hoping for some ability to sense the actual force being applied by the foot at each sense point. If not this round, perhaps it will happen on the next round.
I am not sure replacing the ARF-01 with this board is a good idea. I think it would be to fragile by itsef. The board would be th same shape as the ARF-01, so perhaps it could be added to an ARF-01 instead of replacing it completely. The switches or whatever sensor ends up being used could protrude through the large holes on each end of the ARF-01, possibly supported on a short standoff. Otherwise you are not looking at a single board, but four smaller identical boards.
I just happen to be looking at the BRAT assembly guide, and it looks like there is plenty of clearance on the end holes to put a small sensor board and have the sensors look down over the edge of the ARF-01. It will probably be easier for me to CAD it up rather than try to explain it further. A smaller single sensor board would likely be much easier to sell and more general, which would allow it to be used in other applications also.
8-Dale
It doesn’t HAVE to replace the foot part, but it could. Fiberglass is very tough stuff. The way I am envisioning it, the board can be used by itself, or added to the bottom of the ARF-01. You know I will make it work in as many configurations as possible. We will use through-hole assembly for the switches to enable replacement if they become damaged. The switches don’t have to withstand an F5 tornado, just an innocent little robot trying to walk on a table without falling off.
You 0.125" FR4 would be more than solid enough to replace the ARF-01. I think 0.093" might be a little prone to flexing without some support along the edges but then again most people don’t have the BRAT tap dancing.
It would be nice if the foot was the size of an ARF-02 with a 50% v-score to allow it to optionally be sheared to the ARF-01 size.
A manufacturer with both reflow and selective wave capabilities could reflow the LEDs and any active parts on the top side, then wave solder the through hole switches on the bottom pretty cost effectively.
I understand what you have in mind. 
I was just thinking that a single smaller board with one sensor on it would be usable in even more ways and places - limit switches, for instance. You would need additional mounting holes in the ARF-01 (and ARF-02) to make this setup workable.
I know you’ll make it work in as many ways as you can. That’s the true beauty of the SES. I’m just trying to push the design just a bit further.
Innocent little robot? Hmmmmm, isn’t that an oxymoron?
8-Dale
Well, maybe not yet…
This would work if the board is the same size as the ARF-02 to start with. With my way of doing it, you wouldn’t need to do that though - just position each sensor where you want it - not necessarily on a foot.
Consider the new board to be a jig to mount four small sensor boards with the switches or other sensor on them. This could be added to the bottom of the current ARF-01/ARF-02 or replace them completely. This would also allow for creating new sensors to add to feet and elsewhere on a robot. All you would have to do is make a new single sensor board to mount on the new “jig” board or anywhere else. You only need a “jig” board for each foot style.
It would work.
8-Dale
Well I have the boards outline done but I am having problems placing the relief cut holes in the right spot. What I need are dimensions using the board origin, the " + " symbol at bottom left to get centroid data for each hole based off this origin. When I try to use the measurements in the drawing, the holes do not seem to end up in the right spots. the only hole that worked out right is the bottom left hole. all the other holes were eyeballed.
There might be some problems using Eagle Cad lite, the free version only allows 4 x 3.2 inches board sizes and it won’t let me place the switches past the point of where the top two switches are currently at. 
Here’s a screenshot of what I have so far.
I just finished the new 3D model of the ARF-01 foot, so can place holes where ever you need them to be. You could use my model as the basis for the new board outline.
8-Dale
Instead of direct contact switches, is the thought of using an IR pair of detectors/transmitters mounted front and rear and pointing at an angle down usefull.
I have a small toy robot made by Tiger Toys with a downward pointing sensor that works quite well at staying on the table.
And one other thought is to use conductive foam that returns a variable resistance when squeezed. You could wafer it between two ARF-01’s
Have any of you checked out this bugger?
Capacitive Sensor Board - AD7746 Breakout
sparkfun.com/commerce/produc … ts_id=7918
It seems to come with some sort of external sensor which can be mounted on the underside of the foot.
Another “simple” option would be to tape a few
FlexiForce
tekscan.com/flexiforce.html
Sensors to the bottom of the foot. Possible four, one in each corner. Or six like Mike posted.
Or like Putt Putt said, use IR… Lynxmotion does sell a IR sensor which can be used for this purpose.
I have and already suggested something along this line. Jim does not want anything I2C though.
8-Dale