OpenServo.com

[ginge]

looking forward to this!

As an aside, I want OE to be included in the V4 firmware by default. We have a nice shiny I2C master hardware module spare that can be used. Any thoughts/ideas you know where to post
_________________
http://www.headfuzz.co.uk/
http://www.robotfuzz.co.uk/

[jharvey]

I got the SLA shafts from RobotJay today. At a first glance they look great. It appears the nub is far better centered than I can do with my shaddy lathe attempts. I can't wait to try this out in an OE. I think it will make the process much easer to assemble.

[BasicFox]

very nice indeed

I also have found a handy tool to solder smd (with hot air)
http://www.tme.eu/katalog_pics/9/d/8/9d8caa2a93063823de10abef497ef23e/nimrod%20kit.jpg (i didn't buy it from that site, the image is just larger there :p)

now i can solder even the smallest ones :p

[jharvey]

I completely forgot my butane Radio Shack iron has that hot air torch option. I now have a re-flow toaster oven, so I don't have a lot of need for a hot air tool, unless I want to do some rework. So far I've done OK with the rework dept.

I've had good success with a rough assembly of the parts. These SLA shafts look like they will be much better than my prior lathe theory. I still need to find some fishing twine or similar item to make the shaft a compression fit.

[jharvey]

Hey, there's some good news. The MG995 is very similar to the MG996. It appears to use the same POT, and the new circuit board appears to be very similar to open servo.

[BasicFox]

lol to see the atmega8 on their pcb :p I wonder if it's protected, otherwise it would be re-useable :p how to tweak your MG996 to an OpenServo

[robotjay]

I am glad to see that they stopped sanding off the top of the MCU. If you're inclined, you should try removing the mega8 and slapping a mega168 in there. If you're NOT inclined, you should send some of the boards to me, and I'll see what I can do. . Either way, nice vids. It looks like the OE works like a dream inside the MG995/996 servo. Talk to you soon.

-Jay
_________________
"Nothing is fool-proof; For we fools are ingenious and will find a way."

[jharvey]

I put the OE board up on BatchPCB's market place. If anyone wants to purchase these, they can simply go there and pick it up.

http://batchpcb.com/index.php/Products/23010

[robotjay]

Jared,

I ordered some of the boards from BatchPCB a couple weeks ago, and they shipped today. I noticed that you added a BOM to the OE Construction tutorial, thanks. There is a discrepancy between the value of C2 in your BOM and the capacitance of the part you reference. Is this supposed to be a 10uF part, or a 0.1uF part? Thanks.

-Jay
_________________
"Nothing is fool-proof; For we fools are ingenious and will find a way."

[jharvey]

Sorry about that. It's a bulk buffer cap. On page 9 of the data sheet I was looking at, it shows that cap between 2.2uF and 10uF. I just changed the web page to a 2.2uF, however, the .1 has worked just fine for the ones I've made. There is a good chance that upstream capacitance has taken care of bulk issues. That 100nF is for spikes that need a fast response.

Also thanks for voting for the OE board in Batch. Once I saw that voting thing, I figured out how to vote. Last I checked, and with my self vote, OE was the most voted for. It's kind of cool to know I made a board that's on top of the charts with BatchPCB.

[jharvey]

It's been a little while since I got to play with OE. I just got an entire 30 minutes to play a bit more. I tried the fishing twine theory. It shows promise, but needs some more work. The twine I have is from a friends kids fishing rod. I put the magnet on the steal desk I was working on, then I placed a small piece of the twine over the magnet hole, then I pushed the SLA shaft with pot body onto the magnet. It pressed on fairly easy, but also would rotate on the shaft fairly easy. So I think a larger twine is order. The one I had measures via calipers at .011 inches diameter. So I'll have to stop by a local fishing supply store, and see if I can find one that's a pinch bigger. I'd bet .015 would be just right.

Other form and fit tests have worked out well. There is a bit of a residue on these shafts that has to be cleaned off, but once that is cleaned off and the runner nubs are cut off, these fit nice and tight in the gear slot. This mild interference fit means the magnet should be an exact measurement of the output gear. I could see how the original brass shaft could have a mild backlash. So things are looking good.

[BasicFox]

This maybe an easy and simple solution to the alignment problem of the magnet.
I found a nail that perfectly fits the whole in the magnet (very cheap!)
Now all there has to be done is drill a hole in the existing shaft (using a lathe) and use some glue to hold it together.

[BasicFox]

Found the perfect drill to mount the magnet onto the shaft



_________________
Do try this at home!

[jharvey]

Cool approach. You must have a lathe. How close to on center were you able to keep it? The spec sheet wants something like .005in to .010 in on center.

When I mounted the board(s), it was important to be really close to concentric, as I used super glue to hold the board onto the housing. If I was off center just slightly, it would stress the glue.

I have around 100 of those SLA shafts, let me know if your interested in them. They have worked well for the 15 or so that I've made so far. I also have something like 20 magnets around here as well.

[RasmusB]

Is there an stl file available of the shafts?
_________________
/ Rasmus