Creating work for myself

[Doug Griffith]

I just thought I'd share one of my latest laser projects. I think it's kind of interesting and figured others would as well.

A while back, a customer came to where I consult with a prototype hydroponic plant rotating machine. It was essentially a 4 foot diameter planetary gear that was water-jet cut from 1/2 inch thick acrylic. The goal was to manufacture the machine as a vacuum formed product and the company I was with didn't know how to approach such an endeavor. So I stared at it for a while and told them I could do it. That was almost a year ago.

Below is a shot of some of the tooling while being cleaned up. After doing the CAD/CAM, I machined it out of layers of MDF glued together. It was then sent to a foundry to be cast out of aluminum. There are 3 other tools more complicated than these.
http://www.dogcollarlabor.com/smc/IMAGES/gear_tooling.jpg

So the tooling was almost complete and my next mission was to rotate the outer ring exactly 2 times per day. That is VERY slow. A variable speed gear motor was cost prohibitive so I decided to get the cheapest and slowest gear motor I could find and build a drive system around it. That is where the laser came in.

First, I needed to source readily available components such as the drive wheel and motor. Then I needed to do the math and figure out the gear ratios required to move it that slowly. I wrote an online calculator to help with that.
http://www.dogcollarlabor.com/smc/ratio_calculator.html

A major goal I kept in mind was designing with as many common components as possible. Just in case they get injected molded some day. For example, all the gears are the same. So are the swing arms. I also designed it to fit a sheet size that fit in my laser and also divided evenly into a 4 x 8 sheet. The size of just under 12" x 16" worked fine. Once everything worked, I made further tweaks to the design so parts nested together and shared common cut paths. This made for quicker cutting.

Below is prototype #2. The odd cut lines are due to sharing paths.
http://www.dogcollarlabor.com/smc/IMAGES/proto1.jpg
http://www.dogcollarlabor.com/smc/IMAGES/proto2.jpg
http://www.dogcollarlabor.com/smc/IMAGES/proto3.jpg

I know some of you are thinking acrylic is not good to make gears from and that I should use Delrin. The thing is, these gears move so slow you can't even see it. There's barely any friction to worry about. Plus Delrin is costly.

In the end, each drive system takes about 25 minutes to laser cut. The first run is 800 units with supposed thousands following.

Cheers,
Doug

[Dee Gallo]

Jeez, Doug, my brain hurts just thinking about YOU thinking about this... amazing what an inventive person can do when the right job comes their way. Would you really want to make hundreds or thousands of these?

:) dee

[Doug Griffith]

Thanks Dee. I had a few painful moments. But in all honesty, the tooling hurt the most because it's 3D while the laser cut part is 2D - well almost.

As for making hundreds or even thousands of them, "money talks". Plus I can program while it's cutting.

Cheers

[Frank Corker]

Jeepers Doug, I'm with Dee on that, you have a machine for a brain. Brilliant!

[Rodne Gold]

a prototype hydroponic plant rotating machine??

For weed?
LOL

[Dan Hintz]

<chuckle> I was thinking the same thing as Rodney ;)

Doug, why not use a stepper motor (like on half of our lasers) and a small micro? No kidding, the micro and stepper controller board would be little more than about 1" square, cost <$10 in parts, and you can program it to run at any speed you desire (maybe even settable with a button or dial). No need for any special, slow-running motor that sucks down power running all of the time... the stepper is pulsed once for every move, so power drain would be minimal. That's a huge geartrain reduction, so the motor could be ultra-small since it doesn't need much torque (prior project I did had a 400:1 reduction and lifted 15 pounds off of the floor at around 1"/sec using a motor 1" long and 1/2" diameter). I don't know what your final gear ratio is, but a stepper motor would probably only need to step once per 30-60 seconds.

If you'd like some more info (or would be interested in subbing out that portion of the project), you know where to find me :D

[Scott Shepherd]

The customer swore it was for "Hemp" :)

[Scott Shepherd]

They don't want to pay in snack cakes, do they? :D

[Doug Griffith]

Thanks Frank


a prototype hydroponic plant rotating machine??

For weed?
LOL

I like to say it's for tomatoes! It's a cheaper solution than rotating the lights. Plus it takes far less space. So they say!

[Doug Griffith]

<chuckle> I was thinking the same thing as Rodney ;)

Doug, why not use a stepper motor (like on half of our lasers) and a small micro? No kidding, the micro and stepper controller board would be little more than about 1" square, cost <$10 in parts, and you can program it to run at any speed you desire (maybe even settable with a button or dial). No need for any special, slow-running motor that sucks down power running all of the time... the stepper is pulsed once for every move, so power drain would be minimal. That's a huge geartrain reduction, so the motor could be ultra-small since it doesn't need much torque (prior project I did had a 400:1 reduction and lifted 15 pounds off of the floor at around 1"/sec using a motor 1" long and 1/2" diameter). I don't know what your final gear ratio is, but a stepper motor would probably only need to step once per 30-60 seconds.

If you'd like some more info (or would be interested in subbing out that portion of the project), you know where to find me :D

I DO know where to find you and may take you up on that. I built it this way for a few reasons. First, no matter what motor system I used, I still needed to build a friction drive mechanism so why not gears as well. Second, I could not find stepper motors anywhere near the $10 range you mention. Nor the controller boards. And last, I created an ongoing job for my laser.

[Doug Griffith]

They don't want to pay in snack cakes, do they? :D

That's funny. I'm going to guess the green stuff. You know, cash. Plus I worked in some points on the total unit.

[Dan Hintz]

I DO know where to find you and may take you up on that. I built it this way for a few reasons. First, no matter what motor system I used, I still needed to build a friction drive mechanism so why not gears as well. Second, I could not find stepper motors anywhere near the $10 range you mention. Nor the controller boards. And last, I created an ongoing job for my laser.
Sorry, I wasn't including the motor itself in my <$10, that was just the micro, PCB, and stepper driver. :o I don't know what you're paying for your current motor, but if it's always running, the power savings going to a stepper over hundreds of plants could possibly pay back the extra investment within the first year (or less). The micro could be programmed to wake up at the specified time, pulse the motor, and then go back to sleep. If you're doing that once a minute, you'd be pulling micro-Amps for all but the 50 milliseconds the micro turns on to pulse the motor (which would pull probably 50-100mA for those few short moments, too). Averaged over a minute you're talking about 0.1-0.2Whr (Watt-Hour) per plant... a constantly running motor may be more along the lines of 7-10 Whr per plant. With electricity prices the way they are, well, you get the idea. (NOTE: these are off-the-cuffs estimations and may be way off base, I would need to run the numbers, but the ratio should be realistically close).

[Doug Griffith]

Sorry, I wasn't including the motor itself in my <$10, that was just the micro, PCB, and stepper driver. :o I don't know what you're paying for your current motor, but if it's always running, the power savings going to a stepper over hundreds of plants could possibly pay back the extra investment within the first year (or less). The micro could be programmed to wake up at the specified time, pulse the motor, and then go back to sleep. If you're doing that once a minute, you'd be pulling micro-Amps for all but the 50 milliseconds the micro turns on to pulse the motor (which would pull probably 50-100mA for those few short moments, too). Averaged over a minute you're talking about 0.1-0.2Whr (Watt-Hour) per plant... a constantly running motor may be more along the lines of 7-10 Whr per plant. With electricity prices the way they are, well, you get the idea. (NOTE: these are off-the-cuffs estimations and may be way off base, I would need to run the numbers, but the ratio should be realistically close).

I'm paying app $40 per motor. Offering a version with a stepper shouldn't be too difficult. I'd need to add a few extra holes and modify the top swing arm pivot point. I'll look into motors.

[David Harvey]

What an excellent project! You did very good.

I agree with Dan, a small stepper motor and micro would be your optimum selection. I have designed a unit incorporating a stepper and micro for one of my products and it has worked extremely well.

The motor specs: 12vdc, unipole, 200 ohm, 7.5 deg., 1:85.25 gear ratio, holding torque >600 gf.cm.
The circuit I designed and programmed and can handle steppers with currents up to 1 amp draw. The same circuit I've programmed for use with a DC geared motor also...with no hardware change whatsoever, other than a different motor. The circuit operates on a WDT cycle (~8 sec's) and consumes ~5 ma during no motor movement and ~80-90ma during movement. Very economical on power consumption.

That's the flexibility of a micro...

I've included a picture so you can get an idea of the size.

Ohh....Doug..I forgot to finish the point. In addition to the power savings as Dan talked about ,a big savings in parts can be had too. My unit in moderate quantities costs me ~15.00../including stepper. You should seriously consider going to a stepper and I'm sure Dan can help you out admirably if you should decide to do so. Good luck! A great project you have there that shows tremendous creativity on your part!

David

[Dan Hintz]

I ran some back-of-the-napkin numbers...

To give you a sense of scale, the average household uses about 8-10,000kWhrs per year, or somewhere in the neighborhood of 20-30kWhr/day. Assuming about 50mA of power draw @12V, a stepper motor turned on once a minute to move would consume around 2mWhrs (that's milliWatt-hours), and a standard motor run all of the time would consume about 1Whr per day. At 1,000 plants, that's the difference between 2Whrs for the steppers and 1kWhr for the normal motors, about a 500:1 difference in power draw. Even if the motors take more power than what I have assumed, the ratio between the two should remain about the same. At those draws, the steppers use 48Whrs/day, but the standar motors use 24kWhrs/day. At $0.25/kWhr they would save about $2,200 in electricity costs... that should pay for the extra minimal cost in stepper motors (not to mention having to run more heavy-duty wiring).

One of my products, the Footwell Lights, uses the WDT and the average current draw is <50uA (about 100 times less than Harvey's design, but pointing to the fact that you can all but ignore the power draw when nothing is moving).

[Doug Griffith]

I ran some back-of-the-napkin numbers...

To give you a sense of scale, the average household uses about 8-10,000kWhrs per year, or somewhere in the neighborhood of 20-30kWhr/day. Assuming about 50mA of power draw @12V, a stepper motor turned on once a minute to move would consume around 2mWhrs (that's milliWatt-hours), and a standard motor run all of the time would consume about 1Whr per day. At 1,000 plants, that's the difference between 2Whrs for the steppers and 1kWhr for the normal motors, about a 500:1 difference in power draw. Even if the motors take more power than what I have assumed, the ratio between the two should remain about the same. At those draws, the steppers use 48Whrs/day, but the standar motors use 24kWhrs/day. At $0.25/kWhr they would save about $2,200 in electricity costs... that should pay for the extra minimal cost in stepper motors (not to mention having to run more heavy-duty wiring).

One of my products, the Footwell Lights, uses the WDT and the average current draw is <50uA (about 100 times less than Harvey's design, but pointing to the fact that you can all but ignore the power draw when nothing is moving).

I just met with the client and offered a stepper upgrade. We may possibly do it in the future. He feels the amount of power to move the machine is insignificant compared to the lighting, 2 hydroponic pumps and the aerator. In current systems, the lighting is moved which draws more power than the small motor I'm using. Plus, as long as it works, saving a few dollars isn't an issue. This is not your typical industry!

[Dan Hintz]

Plus, as long as it works, saving a few dollars isn't an issue. This is not your typical industry!
Are you sure this isn't weed? :D

[Scott Shepherd]

Did you get the urge to stop by Taco Bell and order 6 tacos, some nachos, and a large soda on the way home? :D

[Dan Hintz]

If they offer you brownies, send some my way.

[Doug Griffith]

The spoken word is tomatoes but we all know what it's for. It also has the cover reserved for the leading "tomato" magazine. January issue if I remember correctly.

Personally, I like to keep my processor running at full speed. Except the occasional vino or stout.

[Dan Hintz]

Just found this:
http://www.basketrotator.com/Page.asp?PageRef=22

A solar-power hanging basket rotator... something like this might be less expensive to recreate, and due to the slow speed may not even need to be connected to the mains. that would really save them some dough over the short and long haul.

http://www.basketrotator.com/uploads/images/6.gif

http://www.basketrotator.com/uploads/images/20.jpg

[Doug Griffith]

Thanks Dan. The unit I designed is a hydroponic unit and each of the 4 reservoirs will weigh up to 50 lbs. It is more or less complete so a redesign is out of the question. I know it seems odd, but they are also not concerned about saving money in the long haul. They just want it to work.

[George Brown]

I know it seems odd, but they are also not concerned about saving money in the long haul. They just want it to work.

Weed, mon! :D:D

[Doug Griffith]

I just thought I'd update this old post showing progress. This shot shows a group of the first one hundred units. The gears are .125" acetal and the housings are .187" rubber modified extruded acrylic. There is approximately 30 minutes cutting time per unit.

One of the first prototypes has been doing it's job without problems for almost a year now.

[AL Ursich]

I was thinking that it must be for the Washington State "Industry"....

Well I have the advantage of a Dairy Farming Background from my Childhood seeing the Automated Systems installed in 1964 to pipeline the Milk to the cooling tank. Then years of Tinkering, 20 years in the Navy in Radar and Computers and finally 8 years with Sony... 2 in Factory Automation making CRT's in the mid 90's, 2 in the Calibration Lab and 4 at a Service Center.....

So my first thing that I would change is get rid of the Electricity.... That will be a big Liability with all the water.... Could bankrupt you in the very first Opportunist and slick Lawyer.... So I would remove the Electricity from the actual moving base.

Next is Scale Factor.... You are faced with moving some pretty serious weight or something pretty light... It would be a challenge to make this in different weight versions.


So here is my answer to the problem of moving a flower POT.... Or my Design concept... Feel free to use it.

The Answer here is AIR.... LP AIR.... Used as the moving force....

I envision 2 disks of say plastic... Something Cheep.... Made into a Lazy Susan type device in many sizes and weight classes. Mainly Plastic rollers with stainless steel shafts to support the parameter.

The Movement Devices would be plastic linear solenoid that could have as little travel as 1 inch. A Spring loaded return would bring the solenoid back after extending. Around the parameter of the top disk would be teeth facing in one direction. Sure you could use common available metal solenoid devices that I used on Omron Factory Automation operated by a PLC or a Program Logic Controller. I am thinking Plastic due to the moisture or stainless....

So my system would use a Stepping Action and with a slow build up in pressure in the solenoid so it would not shake the plant. Your Analog Continuous Movement would be a shake free and better.... Using 360 Teeth would let you easily setup the rotation speed...

Each plant could have it's own assigned PLC Output.... Only air tubing would need to be routed to the wet areas keeping the 24 Volt DC Air Valves high and dry...

I bet most serious green house operations already have PLC's operating the Watering, Temperature and Air Movement Systems...

So all we need to do is design Air Operated Rotating - Linear Movement Devices. AOR-LMD's:eek:

Your AOR-LMD Turn Table would be a universal device that only needs a air input. That would let different brands of PLC's and Air Control Solenoids interface with your product. The Product would pulse the air making the movement with a locking device preventing free wheeling of the Table between control pulses.

For the budget GROW you could link every AOR-LMD Turn Table to a Common Air Line and a Cam Operated Switch and a simple Air Compressor is all you need.... Or a Laptop with a USB Driver.... So many Options So little time.

And YES.... I do have a little too much time on my hands..... and a Vivid Imigination....

AL

[AL Ursich]

Be sure to look at my Air Operated version in another branch.....

AL

[Doug Griffith]

Air is a good idea and definitely safer. A type of ratcheting action.

I'm thinking it will evolve into a low voltage system. Probably when drive production makes it's way to China. Then there will need to be a retrofit kit to convert the older units.

Currently all the electronics are within an enclosed box above the water line. Everything hooks up to a pricey GFCI unit. Pumps are already submersible. The drive motor is outside the box but beneath a protective cover. Not ideal but how the first release operates.

Re scale factor: the little motor is geared down so low that it has a ridiculous amount of torque. It easily turns 200 pounds when the real world requirement is more like 20.

Oh, and the inventor is against any type of rapid movements. It puts the plant into shock or something like that. He knows more than I.

I attached an illustration I just created for the instruction manual.

[Martin Boekers]

WOW and we aspire to put names on plaques!:eek:

It refreshing to see some great work with lasers and design that isn't just awards!

Incredible!


Marty

[Doug Griffith]

...and here's a pic of an actual unit. The drive unit mounts in one of the corners.

[Brian Fiore]

Thanks Frank



I like to say it's for tomatoes! It's a cheaper solution than rotating the lights. Plus it takes far less space. So they say!


I don't know much about growing anything, but why not just use multiple lights in the room, on timers? Are the lights the costly part? :)

[Doug Griffith]

I don't know much about growing anything, but why not just use multiple lights in the room, on timers? Are the lights the costly part? :)

From the end user's standpoint, cost isn't a real factor. It's reducing electrical usage and limited real estate... and the lights are expensive.

[Frank Corker]

You still amaze me, I think I must be from the stupid planet.... great work and fab design.

[Robert Walters]

Doug,

With such low speed needed, you could use solar panels for each unit as hydroponic "tomato" lighting is typically 1000W @ 100 lm/W. That should give new meaning to "Think Green!". lol

Also, don't forget to ask for samples of the "tomatoes" that have been used on your turntable.

[Darryl Jacobs]

I am impressed!! That is some great imagengineering!! Makes me think I need to get that CNC Mill or Router purchased!!

Curious, for gear design do you "fake" it our use a program?

Cheers!

Darryl

[Michael Simpson Virgina]

If cost is not an issue and you want ultra low power usage. I would look at a Dynamixel actuator. These are continous rotation servos with a brain. They use very little power and you can get all kinds of info from them. Position, Load, Temp, Speed. You can control them with a micro controler with no need for a stepper controller.

They have several sizes The low end ones I believe can be purchased for around $40. The large I think go for $300+


I use them for some magazine articles I wrote a while back. Here are some links from one of my web sites.


kronosrobotics.com

[Doug Griffith]

If cost is not an issue and you want ultra low power usage. I would look at a Dynamixel actuator. These are continous rotation servos with a brain. They use very little power and you can get all kinds of info from them. Position, Load, Temp, Speed. You can control them with a micro controler with no need for a stepper controller.

They have several sizes The low end ones I believe can be purchased for around $40. The large I think go for $300+


I use them for some magazine articles I wrote a while back. Here are some links from one of my web sites.


kronosrobotics.com

I'll check them out. Thanks for the link.

[Doug Griffith]

I am impressed!! That is some great imagengineering!! Makes me think I need to get that CNC Mill or Router purchased!!

Curious, for gear design do you "fake" it our use a program?

Cheers!

Darryl

Thanks. For laser cut spur gears, I usually design with "the real thing", compensating for the laser's kerf times 2. Then create I a single smoothed out bezier curve tooth along the centerline of the overlap (the kerf offsets outwards and there is overlap where the gears mesh in the drawing). Then I replicate the single tooth around the axis. This makes for a far smoother path that still maintains contact close to that of a real spur gear. I hope that made sense.

[Darryl Jacobs]

That makes good sense. Thanks. I will have to try it that way. I usually fake the adjustment to get the thust face angles close.

Cheers!

Darryl

[Doug Griffith]

That makes good sense. Thanks. I will have to try it that way. I usually fake the adjustment to get the thust face angles close.

Cheers!

Darryl

Attached is a pic of what I do for teeth. No corners makes for smoother cutting and more distributed stress. Less control points makes the gears more manageable.

[Darryl Jacobs]

Makes good sense!! I printed a business card sample with the gears as shown elsewhere here and it took a long time to cut because of the exactness of the details in the gear design.

I will convert it to you design and report back.

Cheers!

Darryl