kproche: (Default)
Here's video of the very cool mechanized Gold Medal ThinBot was awarded at Robogames 2012 last night.

kproche: (Default)
Here are a couple of photos of ThinBot after installing the full-length food-grade Norprene tubing.

photos under the cut )

I also did a quick calibration test of one pump, it delivered almost exactly the 100ml requested. The new hoses also maintain better vacuum back to the pump, so the nozzle drips almost not at all when dispensing ends.

Next step (this weekend): build the bell ringer and glass detectors, which are not necessary for ThinBot to perform properly but will definitely add some zing.
kproche: (Default)
Having identified the interference problem with the pump motors, I took a break and built a new cable for the control unit.

pix under the cut )


I also mocked up the design for the new mounting panels:

more pix )


Satisfied that the new design would work, I set out to build it, and assemble the works:
building the new pump head assembly )

Next post: assembled with new hoses.
kproche: (Default)
The Anko pumps are much beefier (over a pound apiece), so I got a new set of panels and set out to rebuild the pump head portion of ThinBot, with blue LEDs behind each pump rotor.

Everything looked great until I tried to assemble the four panels into the head assembly and discovered that the pump motors collided with each other.

Along the way, I figured out a bunch of things for the second attempt, so the effort was not entirely time wasted.


Photos under the cut )

As I said, back to the drawing board.
kproche: (Default)
I just got notice that the better*, custom, peristaltic pumps I ordered for ThinBot in December have finally shipped from Florida, delivery tomorrow.

Too bad Barbot was 3 weeks ago. On the other hand, this means they can be installed for the actual competition at Robogames.

*Well, we hope better. Certainly designed for more demanding service than the aquarium dosing pumps I've used as backup.
kproche: (Default)
I have submitted my registration to enter ThinBot in the Robot Bartending category at the Robogames in San Mateo April 20-22.

More details as I know them.
kproche: (Default)
I haven't had time yet to write up the awesome (and intense) event that was BarBot 2012.

ThinBot performed very well, and got many kudos from the crowd and the other 'bot makers. Based on its performance Friday night, I did a quick update to the pouring software Saturday morning, and we had to field swap one pump head in the middle of Saturday night (15-20 mins downtime, max). Other than that, even with the "Emergency Holographic Pumps" it went pretty smoothly.

We were mentioned (with photos and video) by name in the Make Blog review of the event, so for the moment I'll point you at that:

http://blog.makezine.com/2012/03/06/drinking-in-at-barbot-sf-2012/

Scroll down and look for ThinBot!


(Note for the future: it doesn't matter how fab your 'bot is when all that's left is the cheap booze. The drinks won't taste as good)
kproche: (Default)
Since I know better than to have Friday night be the first real run with a new 'bot, we let ThinBot make Vespers this evening:
pix under the cut )

MMMMMMmmmmm, Vespers....

What's the beverage equivalent of nomnomnom?
kproche: (Default)
We used Andy's camera to shoot some better photos than I could get with my iPhone.

Presenting (sans fluids): ThinBot!

photos under the cut )
kproche: (Default)
We observed last year that it was important to elevate one's power supply above the wet components of a barbot, lest leaks or condensation find their way into the wiring (one bot was offline for a while until its power supply could be replaced).

I'd already mounted my power supply and control relays on a shelf above all the other components of ThinBot, but wanted a little more protection from spilled drinks. Besides, a dome would just sort of finish the look.

After a bit of shopping (we considerd the outer bowl from a salad spinner, and another from a microwave popcorn popper, but they were both translucent rather than transparent) I found a polycarbonate serving bowl that I could repurpose as a dome:

Photos under the cut )

I then reassembled everything so [livejournal.com profile] bovil could use a better camera for a final shot of the assembled ThinBot; I'll put that in a separate post.
kproche: (Default)
This morning I took the time to measure the flow delivered by each of the 8 peristaltic pumps and adjusted the software to take the differences into account (I have a way to adjust the values via a setup screen, but I might was well start with a table adjusted for reality). The pumps all work, and deliver between 66 and 88ml/min. They all also self-prime properly. Big sigh of relief.

The final mechanism to test on ThinBot was the circulation chiller system. Rather than deal with delivering ice and/or shaking/stirring, I'm chilling all the ingredients via chilled water circulation (an idea I happily admit stealing from the iDrink 'bot at last year's BarBot).

From a thermal reservoir (ice chest) of ice and water on the floor, the design calls for a fountain pump to fill the chiller/nozzle assembly, which will then drain into the upper bottle tank, which will drain in turn into the lower bottle tank, which then drains back into the reservoir. This will cool the spirits while adding visual interest as a table fountain. In theory, any way.

Satisfied that ThinBot could actually deliver cocktails, I moved the whole thing outside to test the circulation. If anything went horribly awry (or the aquarium sealant failed) I wanted it to happen someplace where it wouldn't cause damage.

Here's video of the first test:



Good news: it worked (after adjusting the length of the chiller/nozzle tank drain line slightly), and the fountain pump is sized perfectly (it's rated to lift a column of water 4 feet, which is perfect when ThinBot is on a standard-height table). Bad news -- the stream draining from the upper bottle tank to the lower tank splashed enough to spatter the table significantly.

On to try two -- at [livejournal.com profile] bovil's suggestion, I suspended a length of ball chain through the upper tank drain to just above the water height in the lower tank. It worked beautifully, quieting the splashing but still delivering the falling water sound.



In the home stretch now!
kproche: (Default)
This is my last weekend to get ThinBot ready for BarBot next Friday.
Jobs to do: finish installing the lighting harness (requires some disassembly to do so), confirm the aquarium dosing pumps can do the job, and mount the pumps, power supplies and relay board.

I had purchased some OLS multicolor LED accent lighting to try replacing some problematic lights in the living room, they were inadequate to that but I discovered the packaged strips (1 foot long) slid handily into the grooves in the 80/20 extrusions I'm using as the frame for ThinBot. I bought a bigger set to give me a total of 10 1' long strips. Unfortunately, the connectors on the as-sold wiring harness do not slide into the grooves well, and I was going to have to create a fan-out in any case, so that was step one this weekend.

Harness video embed under the cut )

I did a quick test of one of the pumps; they work well but seem to pump closer to 100ml/min than 60. I'll do a full calibration tomorrow, but the good news was that the pump had no problem lifting a water column over two feet in the air and self-primed properly! (hurrah!).

Having established that there was actual hope this could work, I proceeded with the remaining main assembly, and then turned it on and waited for smoke. None happened (again, hurrah!).

smoke test pix under the cut )

And finally, a quick video of the lights running. What remains is putting in the rest of the pumping tubing, and testing the chiller/fountain circulation. That's tomorrow's job.

Big Finale under the cut )
kproche: (Default)
I'm happy to report that the pumps I ordered from eBay were on hand to pick up as soon as we got back from Gallifrey. I've got the mounting hardware all done, but no new pictures yet.

This weekend is the final build and (finally!) wet test for ThinBot. Let's hope I have something other than an expensive piece of water-art to display on March 2nd and 3rd. Photos once I have them.

If you want to attend either or both nights of BarBot 2012, you can buy tickets in advance via the web page. Buying in advance saves you $5 over at the door!
kproche: (Default)
Originally, I designed a lovely outlet for the nozzle on Thinbot to swirl all the ingredients together as they hit the glass, ready to print on a 3d printer.

Then I found out it would cost 200 bucks to print it. Hence the redesign to the chiller/nozzle with the stainless tubes passing through the cocktail-shaker chilling tank.

I did, however, want to keep the swirling-together functionality, which I'm accomplishing with short bent pieces of tubing on the outlet end of the nozzle.

photos under the cut )

I got word that one set of pumps (the emergency backup aquarium dosing pumps, shipping from Tel Aviv) has shipped. For now, though, we are off to Gallifrey later this afternoon!
kproche: (Default)
Some highlight photos of the work completed this weekend (I have detail photos, but not time to write the descriptions yet).

I've packaged the touchscreen and Arduino controller in a nice box.
The silicone-sealed chiller-nozzle passed its leak test with flying colors today.
The support structure for the chiller nozzle and the glass platform are done.

Photos under the cut! )

In other news, I got Tiki Dalek Kraziness (TDK) repacked and prepped for shipping down to Gallifrey. We're sending him to a nearby FedEx office rather than paying the (to me) outrageous receiving ($5/box) and storage ($10/box) fees the hotel would ask. Those would double the cost of shipping him!
kproche: (Default)
Small amounts of progress. I've begun machining the ports into the box that will hold the Arduino Mega 2560 controller and touchscreen (photos of that when its done).

I also did assembled and leak-tested the chiller/nozzle. Alas, the compression surfaces around the O-rings were too small to seal properly and it dripped, so I had to go to the alternate plan (the original plan, actually; the o-rings were a long shot) of using silicone aquarium sealant.

lots of sealant under the cut )
kproche: (Default)
Today I disassembled the chiller/nozzle to clean up and polish all the stainless tubing, so I can assemble and leak-test it tomorrow after I've run the polished tubing through the ultrasonic cleaning system at work.

I also did a test assembly of the frame with the lucite base and shelves and the polycarbonate tanks.

Thinbot frame pix behind the cut )
kproche: (Default)
Work proceeds apace on my entry for BarBot 2012...

Sunday we picked up bottle to soak clean from St. George, and today I did a test assembly of the chiller/nozzle after modifying the plastic cocktail shaker and custom-bending a bunch of stainless tubing.

Pictures under the cut )
kproche: (Default)
No time for a full post, but under the cut you can see some of the parts that have arrived.

parts under the cut )

March 2016

S M T W T F S
  12345
6789101112
13 141516171819
20212223242526
2728293031  

Syndicate

RSS Atom

Most Popular Tags

Style Credit

Expand Cut Tags

No cut tags
Page generated Jul. 24th, 2017 08:27 pm
Powered by Dreamwidth Studios