This is a challenge issued by friends and family. The goal here is to make a robot that can bring beverages from another area to wherever the people are , allow the person to get their drink , then leave. It does not need to be fully autonomous , even remote control is acceptable. It does not need to grab them from the fridge, preloaded beverages are acceptable. This is Ez Robot though, I believe we could do better than that! Well anyways to start the project off I am using a roomba base. I picked up a stainless steel trash can last night as a outside casing. I will show you the rest as we go
By jstarne1
— Last update
Other robots from Synthiam community
Buratino's Tappie
Tappie: 3d printed robot covered with fabric upholstery - 45cm with 5 axis of movement (arms - eyeballs). Runs on EZB4....
J's My Hexapod Project
This is what Im now working on now I finally have all the parts for it now and all my aluminum cut thanks to a local...
Lumpy's Lumpy's Wall-E
** More pics added 08/01/12 ** I spent the most of Saturday working on my Wall-E. I did the typical carve out the bottom...
you can draw inspiration from this model
my cheap choice for a body, 39.99, I would have done plastic but i have a ocd obsession with it being round if possible. Since it is metal i already know there is the potential reception issue so i plan to have a external sma antenna mounted somewhere.
thanks PJ fast reply thats really cool
This is the main chiller unit, Igloo says it can chill 40 cans or 4 liters of liquid up to 40 degree below ambient temperature. The bottom larger sink is the hot side and top is the cool side. there is a large solid aluminum block that serves as a thermal pipe to move the heat from the cool side to the hot side . since the block is very thick i can insulate the area where the beverages will be. The Peltier unit is sandwiched on the hot side. Both already have fans that are blower style. Fans usually work much better when pulling than pushing into a heatsink. That being said I believe I will use some larger 120mm High performance PC fans to make sure the air flows nicely. Also PC fans are made with low sound options which is a pet peave of mine when building any project.
Anyone have ideas of hot to make this chilling unit better? suggestions ? thanks
http://www.igloocoolers.com/faq2 igloo facts page for the thermoelectric unit i came by at auction
Thermoelectric coolers are designed to cool contents as much as 40 degrees Fahrenheit below the room or outdoor operating temperature (e.g. If outdoor temperature is 75 degrees Fahrenheit, the thermoelectric cooler will cool the contents down to 35 degrees Fahrenheit).
Igloo Thermoelectric coolers are designed to operate between 48 to 60 watts (depending on model) and draw an average of 4.2 - 5.0 amps on a continuous basis.
if you want to buy one... http://www.westmarine.com/webapp/wcs/stores/servlet/Product_11151_10001_1193911_-1?ci_src=14110944&ci_sku=1193911&cid=sc_googlepla&device=c&network=g&matchtype=&gclid=CND345jE77gCFSYV7AoduwYAFQ#.UgRw4JLrxfg 179.99
I have the same Idea in mind for my personal robot, but wanted to stay away from the traditional omnibot shells , etc. So I was thinking of a Roomba Vacuum as a base, along with a strip water cooler fridge or trash can as the body. originally Stainless steel was my first option but than again making it as light as possible is my number priority so i drifted away from it. I am not going to start until the release of the new revolution board, so lots of time to brain storm
Since I bought my Peltier element and heat sinks from the auction I wanted to make sure it worked. I powered it up without fans. It took about 2 minutes for the hotside to warm up. It appears it will work but I'm still considering buying a higher power Peltier. This one is a 45w to 50 watt.
Nice project Josh.
Possible upgrade over this smaller unit. T e current one is about 40 watts.
TEC1-12709 Thermoelectric Cooler Peltier 12V 90W Model number: TEC1-12709 Voltage(V): 12V Umax (V): 15.4V Imax (A): 9A QMax (W) : 138.6W Dimensions : 40mm x 40mm x 3.6mm
On Amazon it's 9.50 so they are very cheap.
http://www.amazon.com/gp/aw/d/B002UQ150Q/ref=mp_s_a_1_3?qid=1376093914&sr=8-3&pi=SL75
Hey Josh! Neat idea with the Peltier. I'll be following your build for sure.
Thanks Rgordon and Bret , after friends saw the Brewster video they started saying " so why don't you have a robot bring you a beer" "why doesn't your robots do something useful". Take in mind cleaning robots are not exactly what guys think is fantastic lol. So beer bot challenge it is! This one is keeping the KISS rule though , keep it simple stupid lol.
good luck with your beer bot, josh!
:-) Mel
I'm digging it Josh
Ok for the 29 dollars spent on this thing it proved to be constructed much better than I anticipated. After removing all screws the fit is very tight plus about 8 clips. Eventually I became frustrated and man handled it and put my foot right through the bottom. Booyah! As you can see there are two cans. A outside and inside smaller one. The smaller one is for the refrigerated area so it can be insulated with expanded foam or polyester Matt. I like the idea of expanded foam though because it stiffins and will reduce chances of denting the outer skin with support behind it.
Alright just checking on how much room I have here. Basically I would like the refrigerated area to be at least deep enough to store 12 bottles or a 24 pack of cans. I would like for it to be deep enough for a standard sized 750ml wine bottle of which it could hold three of. Here are the pics for reference.
The area I'm insulating is very deep and longer than the original straw. I used some duck tape and connected a drinking straw to extend how far I could get the foam in the crevice. It took two cans of great stuff expanding foam to fill the gap all the way around.
Josh, that's coming along pretty well so far. Here is something to consider. I have had a lot of those 12v coolers over time. They work pretty well, but the cold side does create condensation. You might want to incorporate some sort of removable drip tray in the bottom.
https://www.youtube.com/watch?v=jiyRDyGcEEM&feature=youtube_gdata_player
This is a ten minute video explaining how the insulation works and how you can do this too. I fill from half way up to the top with foam in this video. I explain the product used , how I am modeling the refrigerated area after commertial units ect.
@Danger - thanks for the tip , if it builds condensation too much I will use a small pump I have to spray water on the hot side making a evaporative booster.
30 min update , you can see that the two cans for sure is enough because it has continued to expand almost filling the entire body now. For the next 30 minutes it will expand a tiny bit more.
This stuff feels every crevice and screw hole. It looks like I will have some excess to clean up after it hardens later tonight or tommorow.
Ok so after it hardened I used a steak knife to saw away the excess below the level of the top plastic ring. Then I used a few ounces of epoxy that I filled the gap with all the way around. This is hardening. The last step in this area is to make a silicon rubber seal to make it air tight when the lid is closed.
The lid - the lid is going to be lifted by a servo with a IR sensor.
Ok all is going well , I started to trace out the bottom of the refrigerated section and then realized what time it was. Basically the next steps are to epoxy a bottom in place and cut out a grate from foodgrade white 1/4" thick plastic. Then I will use pattern drilling accross the grate so air passes through easily regardless of how the cans may be stacked. I will post some of the equipment I plan on using and also research possible upgrades.
Wow Josh - you seem to be never sleeping! How many projects have you started? =) What happened to Jarvis and that little floor-mopping fellow? Have you had a chance to mail out the spare parts again? Mike
@mike yup I have a couple projects in rotation , sometimes you need time away from the project for the creative process of building. I have your parts ready do go with new envelope, I had to wait till my local post office is back open to remail. If you've emailed I been having problems getting emails since my gmail upgraded to this new thing. I don't get ez-robot update emails now either so I'm trying to figure that out. Right now I have Jarvis my big bot and long term development project , squeegee , the automated fish tank controller, my little sister did the electronic cat tail , and there's the two airsoft robots which I only work on when my brother is around. It's a group project.
About squeegee , I descided I will update the sensors because radar worked to avoid objects but a cleaning machine needs to get closer so I'm going to change the setup to be more accurate. It's an overall experiment as to what works best. It's a fun project. Each project I learn stuff I apply to the others so they are kinda interconnected in that way.
Quick update -the first piece of the bottom is super glued in. I must cut two half circles and epoxy them over the two ledges in the bottom. Then cut a hole for the heatsink to pass through. The heatsink will be silicon in place and fan placed on top like seen here.
This is the bottom drip tray where the heatsink will be.
Here is my proposed cool side of the chiller.
Seen here is a 120mm fan over the chiller . I'm considering doing two fans but I think it may not make a difference since the space is small.
Here the bottom is super glued in. I will let it sit for about 30 minutes before doing anything else as it's difficult to get this level inside a big tube. Inserting a wine bottle for reference you can tell this party entertaining robot can hold plenty of beverages.https://www.youtube.com/watch?v=nZVUhd6MUus&feature=youtube_gdata_player
2nd video part one , attaching the bottom drip pan. Unfortunately my phone is only allowing to copy and past one YouTube link at a time so I apologize for multiple posts.
2nd video part two - backfilling the edges of the drip pan with polyester https://www.youtube.com/watch?v=KB7TkaiVk24&feature=youtube_gdata_player
Once this hardens which might take a few hours due to the rain outside I must cut the grate to cover the drip pan and pattern drill it with large 1/4" holes. The bottom will be covered with a sheet of styrofoam and the gaps filled with great stuff foam. I have silicon all the edges with a caulking gun.
It's getting late here and I don't think my drill is charged up but here is the tracing of the inside steel liner in the sheet of 1/4" food grade plastic. This will be screwed down to the drip pan we just installed so it cannot move and will be pattern drilled for excellent airflow. Also I am thinking that I will install a few LEDs somewhere in the bottom to light up the chilled area when grabbing a drink or even just for cool factor. Blue is the preferred color here.
On the lid I'm considering a clear top but solid also sounds good. Any suggestions? The lid ofcourse will be servo powered opening the lid .
Ok so I cut out about a 2" x 2" hole for the aluminum block to pass through. This is the conducter to pull heat from the chilled area. The chiller side of the heatsink I epoxied down because I don't have a second person to hold things in place to put them together.
Ok so I used a jigsaw to cut out the grate for the bottom of the chiller section that covers the fan. The plastic is very strong but while drilling I got to close to an edge and chipped it. Drives me nuts but it's fine for now I can always replace it if it bothers me later. I believe I will paint it black though which will make it blend into the bottom. You can see how far I made it on one charge for the drill. I counted 94 holes and there will be more than double this once I'm done. I'm waiting 30 min to get another charge on the drill. The holes are 3/8" so air should flow effortlessly.
Alright this is how far I got on the second charge , looks like it will take two more charges to complete so I won't finish tonight but tommorow the grate will be drilled. Also the space that chipped I'm actually thankful for because it's big enough to get a finger in so I can pull the grate out of the robot. It's too tight fitting to do it any other way. Lucky me! Lol
Oh man, this is gonna be awesome!
Thanks Niek! Nice updated avatar by the way.
Thanks! Just arrived back in the Netherlands by the way
Well the vacation is over but there is always next year !
@jstarne1 I really appreciate the level of detail and step by step method you use when you post your projects. I always learn a little something.
@antron007- Thankyou for the positive feedback antron , I try to take pics as I go along and more videos than when I first started. I want others to not only appreciate what I did but learn and apply it themselves.
I'm super happy my YouTube channel hit 7000 views this month.
Ok so lots of air holes here. After the holes were finished I wallowed them out like a venturi. This increases airflow to the fan. Basically the next step is cutting spacers and mounting the grate with a couple screws. Then hooking up the fan...
Alright after the appropriate adult beverage I started back work on the booze bot. I used a cutting disc to cut a line all the way around the interior liner . This must be removed because the attached metal has direct exposure to the chilled section. It would let heat in like a giant heat sink.
Materials
The 8 inch tall hot side of the liner removed and excess foam. This revealed something I was afraid of. The foam in the center never cured and eventually deflated into an uncured goop. The only thing I could think of was to expose it to warm water in the shower to speed up cure time.
You can see there is a gap all the way around the chiller section. I'm thinking I will fill this but I will do some internet research first of foam vs dead air space. I believe foam is the answer though.
This is a ver highly rated pc fan Arctic F12. The specs are at 12 volts but I'm considering overvolting to 16 -24 volts to increase airflow.Ok foam won. I filled the lower end of the sides and bottom of the beer robot body with another can of great stuff foam insulation from DOW. That's 10 dollars worth or three cans if your counting. I was going to just use a solid foam sheet to cover the bottom but there's no point in more work as this is made to do the job I'm asking of it. I made a video and posted to YouTube. I meant to say water speeds up the process and drops cure time in half but I said it doubles it on accident at first in the video. Sorry about the gaff!
https://www.youtube.com/watch?v=FfM6iF8WoYg&feature=youtube_gdata_player
Tiny update , this stuff expanded a great deal. I'm only concerned I put in too much at once because I don't want the inside to deflate and be mush. If I must I will buy one more can for touch ups. I poked lots off holes with a long knife with hopes it would allow air into spots that otherwise would not have hardened. What I've learned is it's probably best to avoid getting it deeper than 4 or 5 inches then dry and add more.
At ten min afterwards...
45 min afterwards.. the outside couple inches have dried it's deep down I'm concerned with. We will see tomorrow when I get to trim it back if it worked.Lastly tonight I used the silicone caulk and sealed up the edge all the way around the chiller heatsink and around the drip pan edges. I don't want this to leak condensation in any way unless it's on purpose. My phone was dead so I couldn't do a video for that. To touch in the video I talked about the fans im using and considering raising the voltage to them to at least 16 volts if not up to 24 volts. The pc fans are 12 volts from the manufacturer. At 12 volts the fans produce about 18 db which unless your ear is 5 inches away is practically silent. This is due to the larger 120mm size and sickle blade style which produces the most volume with the least noise.
After sitting 6 hours it appears to have hardened up. I believe the holes poked allowed air to get past the outside skin so the inside could cure. Next is to use the dremel and a carbide drum bit to clear out the excess foam because it expands way past where I want insulation to stop. Hopefully all the foam hardened and everything is peachy. Worse case scenario I cut it back and need to make touch ups. As far as the base I'm getting a Roomba 4xx series I believe from Fred /robotmaker. He is doing a battery mod on it so I can run it from a regular 12 volt battery instead of needing a special roomba battery. Hopefully he will be able to get this out today as the time to start fitting the case onto a motorized base is right around the corner. Other things I can do if the base is not here yet.
Ok update on the bottom half of the insulation. I waited all day till I was off work to insure the 12 inches thick of foam was fully hardened. I used a steak knife to saw out the big sections like I was cutting a round cake. Once I had it sliced down to the last 5 inches I carefully started trimming away like a human cnc one line at a time. A hour later I had it cleared away to a ideal thickness which is about 3 inches thick solid accross the bottom.
Man, super progress Josh!
Thank you bret! It means a lot to hear that. I appreciate it
https://www.youtube.com/watch?v=FiH5Z6RGphs&feature=youtube_gdata_player
This is a video showing what I did tonight and the next steps in mounting a couple cooling fans. Also how I intend on sealing the lid as well.
https://www.facebook.com/photo.php?v=670209156330666 I edited update video with music I made for Facebook but I think it's a good 30 second summary of the different aspects of the projects progress
I picked up a refrigerator thermometer to stick inside the chiller. This way when I test it I can get an idea of how fast it cools and how low I can get the temp in this real world test.
Ok great news , tonight is the first test for the chiller. I'm running it empty first for minimum load and checking that the system can handle pumping enough heat out to hit 34 - 36 degrees F is the target within 3 hours. I took a short video to show the setup so far. I'm using a pc power supply modded to be a 12v bench supply to ensure I have a consistent voltage to the chiller. I estimate that the system draws at the most 6 amps at peak. The original fans are not being used. I am using 120mm F12 fans which are the highest output silent running fan I've ever purchased. There is no bottom ventilation fan to push air in or out of the body and also I did not install the blower in the top yet. Those things will increase efficiency but I want benchmark data , lets walk before we run!
Three things I needed for tonight's segment , arctic silver thermopaste , hookup wire and double sided tape to ensure the fans didn't vibrate.you can see the 120mm fan covers the hot side nicelyhttps://www.youtube.com/watch?v=k2lm2p84Dwk&feature=youtube_gdata_player
Video showing the conditions set for the first test of the chiller. If I must I will swap to a higher wattage peltier and/or up the voltage from 12 volts to 17 volts to get more power to the heat pump and fans.
I'm going to let it run overnight , which in my case is about 6 hours , lets hope nothing goes wrong. Honestly though they are pretty safe to begin with.
The highest Peltier replacement I could find was 154 watts that was 40mm x 40mm in size Tec1-12710 154watt Peltier @15.4 volts
Reverse the polarity to get ice cold in minutes or heat to boiling Used for many applications from CPU coolers to alternate power sources Great for your own custom car drink warmer/cooler as well
http://www.amazon.com/TEC1-12710-Thermoelectric-Cooler-Peltier-Plate/dp/B00ATWGN9G/ref=sr_1_23?s=electronics&ie=UTF8&qid=1376971238&sr=1-23&keywords=thermo+electric
Ok first test is completed. The thermometer does not show a significant difference in internal temp verses outside. I'm going to do some research here but my first impression is the Peltier is too low power or bad. It does get one size warm but not hot like you would expect something that's 38 degrees hotter than the surrounding temp. I will find some examples and maybe videos and see others experience and choose the next step from here.
The first option for replacement is this 154 watt model at 15.4 volts for 10 dollars or less however I found another unit that is 50mm x 50mm just slightly larger but is a 254 watt model at 15.4 volts. It is $45 http://www.frozencpu.com/products/2408/exp-01/245W_Potted_Peltier.html?tl=g30c105&id=dXF2NVas
Some measurement is needed to see if the 50mm version would fit. If it would I may go that route. Otherwise I will stick with the 154 watt version. I looked up different DIY videos for electric coolers around the same volume which were all plastic and foam based with a large heatsinks. The example videos showing the inside and outside temp were working very well so I believe my Peltier must be damaged. I'm also considering runing the whole thing at 17 volts from the -5 and 12+ rail paralleled to get the most out of the fans. That's 1.6 volts over the the rating but there is always some voltage drop under load so it shouldn't be much more load on the Peltier.
That's too bad about the peltier. I know you'll get it done by hook or crook.
When I get home I'm going to unscrew the heat transfer block to make sure it was pressed together. It was a tight fit but I'm just going to double check. I also must add a ventalation fan or two in the back for the heat to escape. Then I will test again. I'm convinced moving up to a nicer Peltier is the way to go though. Thanks for the support troy , it's super frustrating when equipment gives you a problem.
Tell me about it Josh! LOL You always figure it out in the end though.
Sorry for the lack of updates , I picked up a infection from a wisdom tooth that broke. So I been in agony. In my boredom after the meds kicked in I started going over the setup and flaws. One of the things I noticed is though the hotside fan worked the bottom on the chiller didn't have a good way for the hot air to escape. So I'm adding an exhaust fan in the back 120mm silent fan just like the first two. I really didn't like how drilling looked to make a vent so I repurposed a speaker grill as a large van vent. Once the fan is in place it should have no problem pushing the hot air out the back.
Second problem is that the lid was not air tight or insulated. So those two will be addressed. Reading in forums I came accross info stating the arctic silver thermo paste I was using won't work on the cold side because it needs heat from a cpu to cure. So I may swap that for thermal grease that does not need to cure.
Lastly it wouldn't hurt to up the power on the chiller with a newer unit which I posted previously.
My mouth and head is throbbing so I'm done for the night.
Sorry to hear that buddy. Get well soon.
Get well soon, Josh. I remember when that happened to me. Not fun at all.
God Bless You!
;)
Thanks guys , I appreciate the thoughts. I went by radioshack while waiting on food to be cooked and picked up silicon based headsink compound. I got lucky they had a buy one get one free deal. I believe today I will order a higher power Peltier unit and maybe a control board to control temp. With a control board I can make it much more energy efficient by the Peltier and fans shutting off or at least slowing down once the optimal temp is reached. This will also reduce the chance of the beverages freezing if the newer unit has that much power. I will post the items purchased. The roomba will arrive monday in the mail.
OK so I placed a order on amazon to beef up this setup.
Voltage: 12V Umax(V): 15.4V Imax(A): 10A Qmax(W): 154W Dimensions: 40mm x 40mm x 3.2mm
Peltier Cooler Foam Washers
This insulates the edges of the peltier from moisture and heat as well, this is very importantHeat sink , x 2 98 x 40 x 20mm
I plan to use these to bolt onto the sides of the current heatsink which will increase cooling capacity
This is very useful as I can have something to rate the inside temp of the chiller in comparison to the ambient air temperature around the robot. This will be a nice tool to "tune" the chiller later on so that I get good power conservation and internal temperature regulation.Now I basically need a arduino compatible temp sensor. three of them. One for the hot side of the heatsink one for cold and one for ambient surrounding air. The ezb will eventually be able to control temperature. I will have a manual switch to turn on the chiller without a connection to ezb as there will be times the robot is not running around that the chiller will be useful for beverages.
OK test number 2, I added the fan and turned the unit upside down just for this test. I want to see if the additional ventilation made a significant difference. In any case I have a much stronger unit come that should have around 3 times the heat pumping power as this original equipment unit. I will let this run and check it at 1 hour, two hours and three hours will be the max run time. The OEM manufacturer claims a 36 degree difference so if that is the case the inside temp should be 36-38 degrees. I places the thermometer in the chiller as a solid reference to check each time.
http://www.youtube.com/watch?v=OHg3HGSwpfc
Update - Goodies have arrived! Pictures tell all. Today my tooth was extracted and I feel so much better I could do cartwheels. The doctor says no cartwheels for 24 hours though.
Props to Robotmaker for providing the roomba , I appreciate it. This is my first roomba and buying a new or even broken one looked like a lot or money or risk online. I have two stronger Peltier units still on the way. They will be here by thursday. If one unit is not enough I am prepared to move the current one over some and have two run in tandem. Now I can wait for the higher power peltier chillers and if I want I can start playing with the roomba.
Nice!
@bret thanks man! Ever find that laptop? Email me if you did.
I need to make a way to mount the top section into the roomba. Second thoughts are that I need to install at least one caster in the rear because right now it's riding on three wheels. This would make the robot more stable and not fall backwards.
Yes, I did find it.
Hi Josh, thanks for the goody bag! It got to my home adress (mom picked it up at the local post office) but I havent had a chance to look at all the parts yet. Thanks, man!
Josh, if you could figure a way to put one of the peltier units in the cover it might cool more evenly. Cold air sinks. If it is generated in the top it will flow down over the beer.
@ mike your welcome man , enjoy
@Danger - I'm trying to keep as much discreetly mounted as possible. Most the commertial coolers use the bottom or back of a cooler. Because the idea is the lid can raise by a servo I must keep the lid light.
Ok so in order to make this a more efficient cooler I researched the successes of others for a while. One thing I noticed is peltiers that were water cooled worked very well. I'm considering getting a waterblock for the peltiers , a radiator, small pump and some hose and making a small water cooled system to keep the "Hot side" of the Peltier as close to room temp as possible. Also I have two peltiers so I'm considering using both peltiers. The peltiers should arrive today in the mail. I did two doodles to illustrate the idea if two peltiers were used.
Very simple , same as what is used to cool a cpu.
To ensure airflow I was thinking of having a 120mm fan on both the front and back of the radiator.I will check online to find any systems done like this.
Hey Josh, I didn't have the keyboard but I have this adapter...do you think it will work?
I can ship it with the laptop....
Bret
I don't think the usb adapter will work because I'm betting there is no wires just a usb to ps2 converter pcb. I will email you the address.
The PS2 cables are available at GoodWill on the keyboard or mouse for $1.99 each.
My first day with my new team at work my cubicle neighbor brought me a never used OEM mouse. Oh OEM how I love you.
Like a good geek I cut the mouses tail off and stripped the outer insulation to see what's inside. It appears there are 4 individual wires , a ground and a shield. Is this all there is suppose to be? I imagined there would be 6 wires and a ground.hello updates on progress here, next problem is powering the roomba because i don't have a suitable battery i need to wait till i can buy one on amazon. I am waiting on the peltiers which feel like they are taking forever to get here. i believe they will arrive by friday. I want the beer butler to be able to operate in at least two modes with minimal human interaction.
here are ideas i would like to implement in order of importance:
Mode 1 "stationary" - being it must cool the drinks, monitor temp and open the lid when a hand waves over it. This should require no human interaction at all besides the hand wave trigger.
Mode 2 "stand and deliver" - Robot must remove itself from the base and follow a prescript to a point in the room where it will await for a few seconds for someone to wave their hand triggering the lid to open. once the drinks are retrieved the lid closes and robot returns to a spot in front of the base and triggers roombas docking feature to reconnect
mode 3"come to me" voice command trigger and robot follows heat PIR sensor and stops when it is close enough. when done it stays in place unless is it is told " Go Home" and then roomba home docking is activated
Mode 4 "beer butler" to be determined but i have heard of other drink serving robots that can roam around and serve a party, though it sounds cool this is clearly going to be the least used feature
examples of roombas docking feature on my particular model
http://www.youtube.com/watch?v=77BrSbFdNQU
http://www.youtube.com/watch?v=qcBs-fj4ORY this is not my model but just here for reference
OK so what i learned with the auto dock is the roomba consistently lines up to a dock and then slows down to connect. it looks accurate with less than an inch play. That being said the pins on the roomba provide 17 volts at about 2 amps to charge the battery. This is too much voltage for the 12 volt peltier and far too little current (amps) , at max they can handle 15.4 volts so 17 is a little too high of a stretch. My idea is to place a second set of contacts about a inch away from the original ones and those two contacts can be the 12 volt source for the constant peltier cooling. I will probably post to ask a question to gather opinions but this appears to be the most practical way for the robot to connect itself to the 12 volt power for the peltiers in the chiller without human assistance.
Any suggestions are welcome , I'm just using open thought process to develop a solution.
I'm showing the contacts here. Any good ideas for adding pads? Seems like modding this caster and trimming the sides off might make room for two more contact plates.
Tiny update - I checked the tracking for my Peltier units the TEC- 12710 which draw 10 amps of current at 15 volts.
I took off the hot side heatsink and tested it agian! I read 12.04 volts and 2.8 amps. This is basically enough to cool a soft drink in a few minutes from room temp but it won't do a thing with a 12 pack of beer. The wattage of this Peltier I have calculated to be 34 watts so my assumption that it was a 40 watt Peltier is probably correct. Here are pics during the test...
Maybe I will keep this peltier since it seems to work even though it is very low power.Using some basic math here you can see that the new Peltier is 3 to 5 times more powerful. Even with this I'm concerned it's not enough for what I want. I will figure it out though
I know you will Josh, you always do. I'll ship your laptop soon, had a family emergency and just got back. But I found this sweet base for my next robot. The soda can is for scale, it is 3 feet long and 17 inches wide.
it uses a 12 volt battery like the ones I used for Mini-BWow that's going to be a really big base , hopefully a outside robot.
Actually it is going to be my big indoor bot/butler and surveillance bot. I found a little grill at Goodwill that is going on the top. It will house my paintball gun. This thing already uses a 12V 8ah battery. Excited to start working on it, even though I have several others that aren't finished yet. I think I'm taking a page from the Josh Starnes Bot Building Book...
Ok so I spend part of the day getting basic materials like 1/8" aluminum bar stock that's is just about the same width as the pelteirs by about 38mm wide. I removed the original Peltier and see it is a 12706 , the last number indicates the aperage at full on. Notice the previous tests showed it drew 2.8 amps at the most. That tells me most likely something was wrong with the Peltier or the thermosister governing it.
Ok so both 12710 peltiers are hooked up with ample heatsink thermal paste. I will run it for a while and see if the heatsink handles the heat dissipation without getting much warmer than room temp. This is what you want afterwall. You want the peltieri to move as much power asppossible without warming the heatsink up. The 35 to 40 degree difference is dependent on this. If the heatsink was 100 degrees the cold side is 65 degrees which would not work at all! I plan to test this and see how the performance is and go from there. These tests will determine if I move to a larger cooling setup for chilling or if I'm sitting pretty.
8 pm Ok so first 30 min weren't so good . I checked it and thi inside chiller was really hot and the outside which should be hot was barely warm. So that sucks. Apparently the hot side was different so o reversed voltage and promptly smoked my cooling fan. Ugh ! So I have it switched bit I bet it will take just as long to move the heat to the correct side. I will offer updates every fee minutes.
850 pm , I descided to just turn it off and drop heatsinks to room temp. I'm not sure what was going on. My worse fear would be I habe two peltiers in there pumping heat in a circle. I will need to turn them on individually and check the direction of heat. I should have done this to begin with but I assumed they would be the same from the same manufacturing company.
930 update , I pulled the Peltiers out and they were in backwards. I tested them and then put them back in the unit. I now have it powered back on with the Peltiers in the correct direction. On a side now apparently my air conditioning went out and it's 75 degrees inside.
On the bright side, if it cools good with a ambient temp of 75 you know it's gonna work in a more desired environment. On the dark side, now you're stuck being in a 75 degree environment.
Lol thanks. I don't know if these peltiers are bad or what but they work great at heating lol. The first 30 min I'm concerned I broke them because the heatsink was HOT! It didn't burn be but it was a good 130 to 140 degrees. I'm concerned that something about running two in tandem is causing a problems. Hey I guess there's always the option of buying a mini fridge and gutting it lol.
Alright it's official , I'm tired of messing with this so so stuff. I'm considering an actual ac compressor to cool this bad boy. From what I have seen there are some small ones out there for mini fridges that I could easily retrofit.if I had done that to begin with I feel I would already have some frosty beers.
Reading your last post gave me an idea. Instead of mini fridge, use the parts from a small deep freezer. My theory is that if you use a slightly weak voltage it won't freeze. I'm still learning a lot of this stuff so my logic may be flawed but from what I think I know it sounds like a plausible theory.
Hey Josh, any luck with those peltier coolers?
Not really lol , I got them to work on a small scale where they could cool a single drink but when I turned up the power to even half capacity of one Peltier they produced tons of heat on one side and cool on the other. They seem to work best on a very small scale and work 100 times better as a heater than a cooler
I've found it may be better to use a real refrigerator compressor from a mini fridge of some sort as a much more robust cooling solution. I'm looking for a cheap fridge that is in good working order to rob the compressor and thermostat from
I've found then for 20 to 40 used or 60 to 130 new.
I am sorry, Josh. You worked so hard. Good Luck on another solution.
Here's a DC powered 10 Liter Chiller
You might be able to either take it apart and use the innards or just mount it on the Roomba.
Yup that's what I pulled the guts from , a 12v electric cooler from igloo , it's cool I will tear apart a perfectly good fridge and see what I can do lol.
Might be expensive but I wonder if one of those 12 vdc powered RV refrigerators might work ?
@robot doc , they sure would but as you pointed out the small 3 cu ft ones start at 500 and that's out of my budget (as if I ever have one lol) because the plan is to run the compressor only on home base 110 volts can easily be used on two separate pins on the dock. I could have the 110 v on a relay that only trips when the 12v pins close loop.
I picked up a small fridge on craig'slist a hour ago in great condition , but you know me , I tear something perfectly good apart to make it better. Cooling the beverage AND bringing it to me is definitely better! Here are pics of my 15 dollar find !
What's cool is the Kenmore fridge even has specs sheet taped on the back for the compressor , nice. It appears the compressor is about 8 inches wide. Given the current width is 12.5 inches it may be a tight fit due to the connections.
Alright time to test! I need to make sure the fridge works well or using the parts for the robot are pointless. First thing I noticed from visual inspection is there is no external radiator... Hmm maybe it's inside? Maybe it's small enough it doesn't need one. Anyways I'm running it for 20 minutes and one hour marks to measure temperature.
Here is the start mark
20 min mark temp
1 hour mark holy cow! 26 degrees ( the mercury began to move soon as I picked it up) this was measured from the farthest point at the bottom on the fridge. I'm now going to move it to the top which I bet is colder.
Checking the specs the wattage is 96 watts based on the back plate. It indicates 120v x .8 amps . Now I have 110 volts household current so we know the actual wattage being used is lower. Basically this uses 3/4 the power of one of the 12710 Peltiers. This means the setup is far more effective but there will be a slight weight tradeoff. I'm sure the compressor weights more than the Peltiers and heatsinks.
Looks like the project is coming along great. Were you planning on just using the battery power for locomotion? When its on its docking/charging station it will engage the cooling system to make the most of the battery life.
Yes that's the idea. I am open to using the battery and small 120 watt inverter to power it at a party or cookout but 99 percent of the time the cooling could be done by the dock connections.
update checked the thermometer at the top by the evaporator. About 5 degrees! Now we are talking. Also take in mind the mercury started moving up the moment I picked it up and I had to wipe it on my pants to read it clearly from the frost. I'm sure that warmed it up. This has plenty of power to do what I want.... Cool a 24 pack of beer! Or 4 wine bottles or 12 long necks.
I'm super happy with today's find! I technically have not started back building but this find and the tests were very important for the future success and wow factor of the project.
Awesome news Josh!
Videos for installing a compressor. Lots of brazing which is really manly soldering
https://www.youtube.com/watch?v=5dUdYqAqkZk&feature=youtube_gdata_player
After watching the process for charging the system I may consult a professional with the tools to make sure I do the same quality job you would expect in a household appliance and most importantly reliability and safety! The next steps are gutting this perfect fridge and building mounts for it under the chiller section. This will require a frame I'm sure so I will take my time next weekend on it. Just for now I get to enjoy tearing it apart. I will post a video.
Here's a guy who did this really quick and easy , he used a automotive adapter and a freon R134a can to fill the compressor. My system as it is only holds 2.3 ounces but I may need to adjust that based on pressure if the volume of the lines is reduced significantly when I install it in the robot
https://www.youtube.com/watch?v=SsF4gqbcSV4&feature=youtube_gdata_player
Nice on the fly adjustment and modification. I can't wait to see this thing in action. Hopefully I can get back to my bots soon, but it has been a crazy summer. First the move, then I had to put my dog down, then my father-in-law died, then my daughter's fiancee committed suicide by jumping in front of a semi truck on the freeway. I haven't had time to think. It has been so nice for me to come on this forum and see what everyone else is doing. It keeps me in touch at least.
Anyway - good luck on the new refrigeration unit.
Omg death by truck that's brutal , well maybe you NEED robots to take your mind off stuff.
. Thanks man I have a project I don't always make as many videos as I want but I do take enough time to get things working beyond just being a novelty. I'm so glad I prefermed tests instead of getting this guy rolling around and turned outthe thermocouple elements didn't work. I was super bummed that the Peltiers didn't do this trick but those tests led to much much better things! Soon I will get him moving around I just need to do some measuring and framing to mount the compressor under the bottom. I will sketch some ideas. Oh and I need to get a battery for the Roomba.
I know you can get those batteries on ebay. Is the roomba gonna be able to handle the weight? maybe you want to take all of the innards from the roomba and put them on a custom chassis with bigger motors?
@ Brett very very good point. Since AAMERALIS did some serious upgrade to his designs perhaps he wouldn't mind selling some of the Roomba motors he bought and I make something more custom. You know I'm not afraid to fabricate something lol. Fred says they can handle up to 70 pounds I will find info to source that.
Man that's a tight fit
I build custom packs for the Roomba. They are 4500 mAH packs, email me if your are interested. They ain't cheap though..... e_vickery(at)BellSouth(dot)Net
@Josh That is snug indeed. From the vids you posted it appears those tubes bend pretty well so I think you of all people can pull it off.
@antron. Yes it's a tight fit. I'm going to go shopping for copper tubing today and see if I can find a small radius bender for 1/4" copper tubing to make some tight bends needed to make this fit properly. I am considering a spiral condenser I've seen people do for refrigerators that don't have one on the back.
I also see you have a tailors tape. I was coincidentally given one about a week ago and find it very useful and easy to work with. I'm digging it.
Oh yea I realized that measuring things with curves tailor tape is s great idea.
Alright I pulled out the dremel and cut the compressor out , it was very easy. The hard part will be recovering the thermostat and that wiring. I need to grab a sharp chisel to get through the steel outer casing. I want to use as much original equipment as possible
As you can see the compressor is not very large. It's really the wiring and tubes that make it a tight fit but I have a plan that with give me about 4 inches more room all the way around. I will draw it but cannot make the case improvments until I get paid this weekend.
Look very hard to fit the compressor, the heat dissipator and the freeze element inside the little can you have, you also will have to reconnect all the tubing and bring new freon inside this unit, you take a very hard road jstarne1 !
I look forward to see your progress, I am interrested in seeing the batteries duration. Do not stop, go forward !
The compressor will only run while docked is the plan. The battery will power ezb , lights and Roomba.
Ok payday arrived and that means getting parts
Ok so I need to salvage parts from the refrigerator. One of these things is the temperature controller. Second is I needed to remove the evaporator and measure it so I know about how large to make the new one. Take in mind this one is being made from tubing.
I measured each length of this little maze of tubing in the evaporator. I measured approx 168 inches. Wow that's a lot of surface area! it more than I expected but with it woven together like this it's possible. Here my pic I measured on.
I salvaged the thermostat and housing so I keep a professional look.
I intend of actually making the evaporator from a continuous 1/4" copper coil in the chiller. It will be coiled against the wall. So measuring I find it's just under 10.25" internal diameter. That is a circumference of 32.2" around the inside walls. To match the original minimum length of the original evaporator we need 168 inches of tubing in the chiller. Dividing the 168 by 32.2 circumference we get 5.21. This means I need the evaporator to wrap around the inside 5 to 6 times before going back out to the condensor.
Here is a neat video that visually shows the process
https://www.youtube.com/watch?v=nA-XbfxqgRo&feature=youtube_gdata_player
the only part I'm not sure about is the very small tubing but I will continue to salvage the rest of the parts from the fridge tommorow.
Ok I salvaged the last piece which no doubt is very important.
the required stages are as follows
Compress the refrigerant
Leaves the compressor from the discharge tube and raises the temperature to over 100 degrees
Discharge goes to radiator This cools the hot high pressure refrigerant
4.the refrigerant moves from the radiator to the drier or expansion tube.
The refrigerant is pushed through the small tube as a liquid.
The small tube opens up to the evaporator
The liquid expands into a gaseous state rapidly as it passes into the larger space of the evaporator.
After the evaporator the compressor vacuums the refrigerants into the compressor.
The thermostat looks like I wont have any problem fitting but I am considering mounting in the very bottoms as well. The advantage is the wiring would be compact and hidden under the grate. The disadvantage of being in the bottom is any condensation will run to the bottom as well and I don't want to risk a short or corrosion. For these reasons I'm leaning towards a higher mounting position.Ok so I did a rough doodle of what I am thinking. I researched more and found that the minifridge does not have an external condensor instead the copper line runs behind/under the outer sheet metal skin of the refrigerator and uses the actual case like a heatsink. since I don't have a large casing like that on the robot my best substitution would be wrapping it around the outside of the stainless steel body. Then the condensor forces the fluid through the capillary tube into the evaporator which is INSIDE the robot to cool the beverages.
Ok I decided to do the outside condenser first. The original fridge didn't really have one so this should be more than enough as far as surface area goes. The last coil is not afffixed because I need extra tubing to play with later to get everything connected.
Look I'm done! Ok not really.
Ok that's more like it!Nice engineering progress !
Thanks Robot-doc! I was a little frustrated that afternoon because the swaggin tool just didn't cut the mustard. However the new one already arrived! I couldn't wait to try it and it was perfect!
Ta-da! First try I was able to swagg this tube. So much better than the first tool. Now I can continue the work on this robot.
Nothing like the proper tool for the right result. Nice tool kit.
You are an air conditioning expert,too. Wow! you are knowledgable of many trades. That is good because you never know when you are going to need it.
P.S. If you are interested in that turtlebot, send me an email.
;)
What's your email again?
:)
Josh you're a jack of all trades!
yes the right tools...
looking good.
:)
j
Alright it's been a couple weeks playing the waiting game. The Roomba Home base and speed charger arrived. With that I also picked up "upgrade motors" for the Roomba in case the current ones don't have the power I need. Basic math tells me they will pull it I just don't want my project to seem like it's struggling. Last major piece I need to continue is a replacement Roomba battery. I may order one today from amazon.
Alright now that I finally have the charger I can order a battery. Fortunately for me the Amazon fulfillment center is in town so next day shipping is pretty cheap. It will go out monday and I will have it Tuesday.
Alright , great! Now that the Roomba control board has the appropriate battery and voltage I can run it without throwing the "dead battery" code and beeping. The 12v sla I had on hand are fine for many things but cannot maintain the minimum voltage to power the Roomba. Since the base can already charge this I will work on the project for the moment as it is. I noticed the Roomba battery from turnigy is bright yellow , I won't lose this thing! I charged it for about 3 hours directly plugged in , not through the home base like recommended.
Next I will wire up that ps2 connector and see if I can drive this guy around and initiate a auto dock from script. I will probably do a video for that.
Nice update. I look forward to a video.
Just remember that as long as the battery is installed inside the Roomba, there is a constant drain of about 20 to 30 ma due to the MCU always being powered and the IR receiver always on waiting for IR remote control commands. The battery will reach 12 volts in 4 to 5 days if not being charged.
Oh wow , that's a good tip. Thankyou for bringing that to my attention Robodoc. This is my first Roomba project with EZB. I didn't think of the battery going dead that quick. Well one assembled the project will be on a dock so that should keep the battery topped off. Somehow time flew by last night. I went to buy a black Christmas tree and apparently I was back after 10pm.
This project is shaping up beautifully! Keep it up Josh and we might have to build one for the EZ-Robot office using your plans
This a really neat project! I am anxious to see it in operation. Josh, my hat is off to you sir...you are a man of many talents. I will be fascinated to see the auto docking to charger. Videos...videos...videos...
Good job man!
@Jason , Thankyou , you guys absolutely could have a office beerbot. You can cheat too and not use a mini fridge compressor and just dump a bag of ice in it. I like the idea of it being a home appliance just as much as a robot too so I went the extra step. Thanks
@RGordon - Thankyou Rex , you really should plan for the next makerfaire and we can meet up agian. I will get videos.of this guy working once I have the majority of the body mounted. It's a careful process.
Ok update : I cut up a ps2 mouse and after testing with a meter I found that the pin I needed to have a wire soldered to was not connected to a wire . So I had to cut it up and after a few min I was able to get to the pins inside. I then soldered two jumper wires.
Ok when trying to do the initial connection the Roomba powered off and couldn't get it back on. I checked the battery and had 15.04 volts and the charger was 22 volts. After research I found the 15 volts means the battery is dead. So apparently robotdoc was very right about the battery draining thing. I reset the battery charger by holding the power button down 30 seconds and now it's charging again. I will continue tommorow.
Full charge should be 17+ volts
https://www.youtube.com/watch?v=l5OWLAMniWo&feature=youtube_gdata_player
Unedited video of the first movements. I did sit the body onto to add weight and the roomba moved effortlessly like it wasn't there. I have a couple things that must be done at some point. A relay on the power wire , a relay to trigger the power button press to activate roomba so it can connect. Those can be added later when wiring is ran. I'm mostly concerned that I cannot trigger the dock feature. I must have "disable sensors" checked to use the Movement Panel but I believe this is causing me not to be able to use the autodock feature.
-Showing manual demonstration of how the Roomba needs to dock.
Ok here Im MANUALLY demonstrating what it should do if autodocking was working. I later figured out that if i press clean first it can automatically find the home base. SO that being said this video is not very important.
Manually getting the roomba to dock from three different directions, this is not command based , its manual control.
https://www.youtube.com/watch?v=FpRBsTIv6bI&feature=youtube_gdata_player
Roomba first auto dock , Who wants a treat !? Whos a good Roomba? I got some volts for ya
https://www.youtube.com/watch?v=6ifJbZP5HAY&feature=youtube_gdata_player
first success with auto docking next i will make sure i can repeat the success over and over.
https://www.youtube.com/watch?v=5x7O2bADOc4&feature=youtube_gdata_player
Repeated auto dock and initialization without human help. Very happy. Now I just need to figure out what the script commands are.
Josh, do you have any info on when the next makerfaire will be ....and where?
not off the top of my head but North Carolina where we met does one in that building / school every year
Next I need to create some scripts and I posted a question about how they should be utilized but here are all the terms I could find. I just dont know how to structure them to function.
Roomba Commands Init EnableSensors DisableSensors PowerOff SpotClean Clean MaxClean DisableAllBrushes MainBrushOn MainBrushOff SideBrushOn SideBrushOff VacuumOn VacuumOff SeekDockingStation
Important note that i explained in one of the most recent videos for repeated docking success, when undocking I noticed I needed to click the Disable sensors then the INIT agian and then Disable sensors agian to be able to remove it from the charging mode so it can come off the dock. Otherwise if you only press INIT and try to backup it doesn't do anything.
EDIT: This looks useful, looks like I need to start the commands with Roomba
Roomba( cmd ) Execute the specified command on a connected Roomba Vacuum on Port D0 and EZ-B 0. Look for the available Roomba commands further below in this document. You may also add the Roomba Movement Panel for graphical controls. Example: Roomba(init) Example: Roomba(SideBrushOn)
Ok some progress here ,it's time to form up a frame to connect the chiller section to the roomba base. Inbetween there will be a platform for the compressor which I will install once everything else is working correctly. I did a sketch a min ago...
I cut a wooden ring out by hand and used superglue to tack in place. Once that was in place I used tape to back off the seem so that I could fill any small gaps with epoxy resin. This is a filler and also an adhesive to keep the ring in place. I will also add 3 to 4 screws later which will be predrilled.
So that I can keep this project a reasonable height I may cut 3 to 5 inches off the bottom of the chiller section. It is basically unused and holds heat in the area where the compressor and dryer are going to be. I will cut this down tonight and then continue the framing.
This is the framing idea I have. Keeping it light as possible for being a wood frame I am thinking of three legs. Then there is a top ring, center ring, the platform for compressor and base.
Three legs
Top ring
Center ring
Compressor platform
Round base.
I may need to get another piece of wood to do it all but this is the simplest I believe it can be and still support the weight of everything. I may get a hole saw and drill large holes into the wood frame in a pattern to reduce the overall weight but we will see about that later.
Ok so I pulled out the jigsaw because I don't have a tablesaw. I used a 1 1/2" wide piece of aluminum as a template and traced the first leg out I carefully cut it out biting my tounge the whole way. The leg fits over the rings so the next step is knotching out the leg to fit over the ring I epoxied in place. Hopefully I will have all three legs cut either tonight or tomorrow.
I found a foam dome I could use as a form for a topper on this robot. What does everyone think? Dome topper or flat?
I like the dome. It adds a R2/Jetsons vibe to it.
Nickname it R2 C1.(R2 chiller 1) This looks awesome!
Dome is a better plan, that way no-one will attempt to use it as a portable table.
Having a dome does give me open space to put sensors like the ceiling camera to read QR codes.
Ok so I did test some QR codes on the ceiling. The small QR code would read with a illuminator indoors at 2-2.5 ft reliably. The enlarged 400 x 400 line QR code was able to be read at 6-7 ft but it was only consistant every 2 seconds when at 5 ft distance. The largest QR code was approx 600 lines by 600 lines and read consistently in any direction at 7 ft 6 inches away with a 1 watt led illuminator. The next steps are testing if a IR illuminator will react the same and building it. I have 100 850 mm IR less for the next tests. Unless the camera control can determine angle of orientation on the QR code I will need to use a Compass module to determine orientation of the QR code landmarks.
So I needed to do small modifications to the case I bought for the JARVIS hardware while he is still being built and wired ect. The usb connectors and audio and Mic were all 2 inches to short to reach the plug and then they must navigate around or over the video card. I used a bundle of jumper wires to extend the cables about 6 inches in length. It was about 70 soldered connections but I got them extended and turned on the first time. The case is pretty for sure
. While I was soldering these connections my sister hijacked my phone to post on my Facebook " how pretty she was". Yup I really should put a lock on that thing. Lol
Hey at least the leds look awesome on this case. It. Would be better if the case was red though. Just saying
As for the QR code tests I now need to made a IR illuminator that highlights the ceiling to make the QR codes easy to read.
That is a really good looking case, Josh.
Thanks troy!
https://www.youtube.com/watch?v=c7CzLF8z7uY&feature=youtube_gdata_player
I tested the Camera reading QR codes from 7 ft 4 inches away and it does even better than with a bright white 1 watt led pointed at it. To make the camera read the IR light I had to remove the IR filter over the image sensor. This was very easy and increased the overall light sensitivity in low light like many homes are. Now I need to figure out how to read the angle that the QR code was read at so this can be formed into a simple landmark IR navigation system.