Micro Hex Walker input requested

Yeah, I am considering kits. Depends on how well it all goes. I mean, who wouldn’t want a walker that fits in the palm of your hand?

I want one ;D

Just got my V1 boards… ;D

Edit: I was able to fit RTC and accelerometer in for V2. But I must first validate the PSU on V1 before I move to V2…

And top side.

Now the fun starts… ;D

That board looks awesome. But there are many via’s that seem to overlap, did you do that on purpose?

@ Wouter Huysentruit: Yes, all the “big” overlapping vias are connecting the top and bottom ground plains together. Had to overlap them to fit the max number of vias. There will be a fair amount of current in the ground plain due to the servos and I’m trying to spread the current…

I ran Dorkbot’s Design Rule Check to make sure that nothing overlaps that shouldn’t. It didn’t tell me that one of the servo connectors aren’t connected though… :-[

Sweet! Can’t wait to try it someday :wink: I have some mini hexapod guts laying around for some time now - never had time to finish that project. But I have high hopes for your board. ;D

[title]Everything is Go[/title]

[ulist]PSU works!
Fuel Gauge works!
Soft power switch works!
(Fuel Gauge is 1Wire. If you pull the 1Wire pin down for 1 second then the fuel gauge disconnects the battery from the rest of the circuit. If you let it go high again then it rearms. Pull it low again and it turns the power back on)[/ulist]

(Yeah, I melted the I2C connector a bit. Will replace that later…)


Congrats! :dance:

My task for tonight is to bring up the USBizi. But I don’t expect any issues there.

I don’t have the exact PNP in my stock that is on the USB pull up, but i’m sure any low current PNP will do.

Also, I’m still waiting for my reel of 33 ohm 0603 resistors(should be here on Wednesday), so I might just slap two 68 ohm 0402s in parallel, which will give me 34 ohm.

Let this weekend be fun… :slight_smile:

So far, so good.

USBizi up and running. Loaded newest firmware using USB. Everything OK.

Lipo charger is happy to supply the board with power without a lipo connected, which is nice for development.

Need to test that bluetooth is connected correctly to the USBizi.


How are you running 18 servos from the 6 PWMs offered by the USBizi?

See OutputCompare in the documentation.

I’m aware of OutputCompare, but that would limit the device to moving a single leg at a time, and eliminate the servo’s ability to actively “hold position” … is that desirable?

Errol, how was the (I assume) hand-soldering of the USBizi?

intresting project. Humm. Not only will you need to control 18 servos but you have to do the controls at a certain speed. Fez can do the calculations but you will need some smarts to get the timings right.

I used hot air and solder paste to solder the chip which was a breeze… :). For this board hot air is a must. None of the other chips have legs, all their legs are under the chip, which makes hand soldering REALLY hard.

For servo control I plan to use hardware assisted OutputCompare. The USBizi has 5 timers, one used by the OS. All the timers have 4 “interrupt on match” functions which I plan to use to control the servos using RLP. This should use minimal CPU while being able to run 16 servos in a 2.5ms window, 32 servos in a 5ms window. That is 200Hz updates for 32 servos, which my servos can apparently do. Then the managed code can just give the RLP code servo timings 50 times per second, or what ever is enough for fluid motion, and the RLP can do the rest.

At this point I’m focused on 12 servos for the legs, possibly two extra for the head. Depending on how things go I might play with 4 legs with 3DOF(still 12 servos) then 6 legs with 3DOF(18 servos).

Edit: CPU has 4 timers, of which 1 is used. Leaves 3 timers with 4 Match registers each, for 12 servos per 2.5ms frame, 24 servos per 5ms frame…

Some more feedback.
Bluetooth works as expected. Using it to relay fuel gauge data to the PC while plugging in and out the usb charging cable.

This is some of the data that I’m getting from the fuel gauge:
Temp:38.375C, Charger:True, StateOfCharge:0%, Current: 399.6875mA, Volt: 3.37134004V,
Charge Remaining: 3.20000005mAh, Learning Battery Capacity: Yes, Battery Capacity: 1000

Servo plugs fitted. Me and hot air and plastic connectors are not friends. :frowning:

RPL servo driving code is running nicely, now I just need my servos to start testing… :frowning: