Yeah, start with a BRAT. 
Then add more DOF’s as you work. Think of all the other BRAT builders who will be praising your name Nick. hehehe
The switch / foot PC board is a good idea. I want to have a PC board made with the switches and even LEDs, that can either replace the aluminum foot for lightweight bots, or be added to the aluminum foot for heavier bots. It’s on the drawing board anyway. 
whats wrong with micro servos?
Better add a place in the bottom of each foot to hold a couple of 2 cell LiPOs that can be joined in parallel and regulated for servo power. 
That’s my story and I’m sticking to it, Harry
I’ve been thinking about this quite a bit and I think it’s the way to go for me to start looking at balance in humanoids. The configuration that seems to make the most sense to me is to add another servo just above the foot with the same SES components that are on the similarly placed servo on the BRAT. This would require an ABS-04, ABS-06 and ABS-09. Jim, what do you think? I remember you suggesting that adding another DOF at the hip may make more sense.
Any thoughts from other BRAT owners or SES experts?
This looks really interesting too. Any chance that it’s moved off the drawing board yet? 
Thanks, Harry
No sorry, I have my PCB engineer working on our new Dual H-Bridge, and he will not be able to work on other things untill it’s completed.
Pleeeeeeeeeeeeeeeeease tell me that it will be at least 5A!!!
It must be! (and cheap, please!)
theChipmunk
Heh heh, I just got my first issues of Robot, SERVO, and Nuts and Volts this week. I am in information heaven right now!
Wow, I know just how you feel. This should keep you bussy in the bathroom for hours.
It’s 10 amp, two channels. It will be like the SSC-32… Feature packed but inexpensive.
Oooh.
That sounds very useable, Jim.
How inexpensive!? If it was a good price, I would hold fire on buying the Polulu.
BTW, do you ship to N Ireland for free…?
theChipmunk
It’s way early to quote a price. I probably should have kept my mouth shut. 
Unless UPS starts delivering my packages for free, I’m gonna have to keep charging for the service. Sorry…
Nope. We appreciate the fact that you can talk to the Head Honcho…
theChipmunk
cough Thread hijackers. cough
I finished the schematic for this bugger, today:
img1.putfile.com/thumb/9/24517202050.jpg
Hopefully, my idea of using LEDs to visually confirm that the buttons are depressed will work.
I tried it on a breadboard, and it seems fine.
Hopefully, it won’t affect the 20ms button debouncing in the software.
Each LED should be getting around 5mA (5V / 1000 Ohms), so I’ll be wasting 40mA on those LEDs alone, but I feel it’ll be well worth it.
Especially since I’ll have much more than an amp of logic power to burn.
Does it look like I’ve got the seperate batery-source thing correct?
Should I use a common ground as I did above, or should I keep them seperate?
I was going to put a schottky diode to protect for reverse connection in the servo supply as well, but I opted against it.
At 8.4V max, 7.2V nominal, I’m cutting things closer than I’d like.
I’ll be regulating VS to 6V, so if I assume a .5V drop, there’s little leeway for another small drop with a schottky.
EDIT: Sorry about giving three capacitors the same name (“C2”).
Thus are the limits of copy/paste truly realised.
I don’t think that will work as you intend. It would require that the PIC pins are being inputs and outputs at the same time. Unless you plan to switch from I to O at a high rate in the SW?..
But remember you have to take into account the power dissipation of the regulator. With 7.2V input, and 1A output, your regulator must dissipate 2.2W, which is a fair bit of heat. You would need some sort of heat sink at a minimum.
Yes, I think it’s OK as drawn. However, you should watch out for “ground loop” problems since you have the servo circuits that draw relatively higher currents. What you should NOT do is “daisy chain” the ground circuit that includes U1 or the servo grounds. That is, do NOT run a wire from the power-input, to one device, then to another device, etc.
Instead, from the power-input, run a ground wire from there to U1 and to the servo connectors. Then run a separate ground wire that starts at the power-input connector, and goes to all the low-current devices and connectors (it’s OK for the low-current stuff to be daisy-chained).
Good call. What you should do instead is put a diode (regular type is OK) across the power-input connections, rather than in series. If you apply power backwards, the diode will take all the abuse. Since you don’t want the diode to explode, also add a fuse in series (before the diode). Then if you apply backwards juice, the fuse blows and you’re safe.
In your schematic, on pin 1 of the PIC, R3 should go to Vdd rather than Vss, and you should put it in series with a small diode. As is, the /MCLR line will be low all the time except when your PIC programmer is pulling up the VPP line. See the schematics I’ve posted, for example my sensor board with the 16F88 on it. The diode can be almost anything, such as 1N914, 1N4148, or 1N4001 (that last one is physically larger, but it doesn’t matter otherwise).
The diode needs to be there because the PIC programmer will apply about 13V to that pin during programming, and you don’t want the 13V to force itself back into the other circuitry. The way you drew it is “safe” because R3 goes to ground, but it will keep your PIC in reset…
Pete
BTW Nick,
Perhaps what you want to do for the button/light thing is like this:
+5
|
SW
|
+--- to PIC pin
|
R
|
LED
|
GndPete
Holy schematics Nick!
I saw all the finely detailed schematics in your putfile collection! Very well done and very professional looking! They look like data sheets!
Doh!
I entirely forgot about the pin direction!
Well, I don’t actually intend to rip an amp through either of the regulators, but it’s nice to know that I have sufficient headroom with all that overkill.
Since one of the feet will be entirely dedicated to the regulators, I’ll have ample room for a heatsink for both of them.
I gotcha.
In other words, make direct connections to the input grounds for all the biggies.
I’ll take that a step further and run direct connections to the regulator outputs for the positive-side of the biggies as well.
Woah…
I almost did something right?!
Hmm…
Doesn’t the fuse sort of defeat the purpose of reverse protection, though?
If the fuse blows, then the path of least resistance will be through the circuit, backwards.
So, this is sort of just a WAKE UP, STUPID!!! kind of warning?
Certaintly better than nothing…
Oh dear…
That would have been really bad, sitting around with a micro held in reset and wondering what I did wrong.
The sad part is, on all my other schematics I’ve pulled it up.
If only I could get some time to work on these things while I’m still awake…
I checked out your sensor board schematic and I see what you mean.
Does that itty bitty cap decouple the arguement between the two different voltage levels at that junction?
AHHHH!
So, now I’d just set that pin for input and poll it from time to time.
And, the LED would be on whilst the buttons are depressed.
Sweeeeeeeeeet.
Thanks so much, Pete!
You wouldn’t happen to have a nice big logo that I can slap on this bot when it’s done, eh?
Perhaps I should just make one that says:
[size=150]Pete > You Humans[/size]
I’ll post back later tonight.
I should have my shrine for Pete errected by then and the schematic revised.
Oh, and Mike, those pretty schematics in my file won’t actually work.
I tend to make things pretty to cover up the huge lack of knowledge between my ears.
I’ve found that thinking outloud helps when I have no clue what I’m doing.
That way, I can look back and see where I went wrong, rather than having to redo everything.
Usually, though, I end up messing up something crucial in the beginning, and it’s all a waste of time, anyhow.
I finally got a chance to get back on and post updates to all my ongoing projects.
Here’s a pic of what I’ve got done with the mini biped.
img293.imageshack.us/img293/5969/p1010104av7.th.jpg
Not much.
The limit switch below the bot is one of the eight that I’ll be putting on the bottom of the feet to get it to balance.
I spent a lot of time in RadioShack looking for ordinary switches, but even the PCB ones take too much force to fully depress 8 of them.
So, I finally decided to give in and go with my backup plan of limit switches.
These ones are small, at least as limit switches go, so they should work well.
With most of the mechanics now solved, and a pretty simple electronics schematic to follow, I should be all set.
Well, I should be, except the damn battery charger IC is holding me back.
I can get it to the brink of charging the batteries, where it’s passed all it’s necessary internal tests (error LED off, status LED off), but I can’t get it to start charging (status LED on).
Of course, the data sheet says that if the error LED is off, then the status LED has to go on, and the charger has to start charging.
All my tests say otherwise.
It also doesn’t help that the Microchip people don’t like to answer my questions.
Until I get the battery issue fixed, moving on to the rest of the project is rather pointless.
