the dog wants out - newbie project review
Bruce Hensley
Posts: 15
Hello all.· Help please!!
I could really use any suggestions about this project.· I've never built an electronic project, but understand basic electricity, can solder, have wired some of my house and home intranet, have moderate mechanical skills,·and have 20+ years professional programming experience in various BASICs and in research.
The dog now needs out hourly instead of just before and after work.· This is causing great stress, so we're looking for a fast solution (that's not too expensive).· Commercial units to let the dog open our sliding glass door range from $400 for one that won't fit our situation to $1600 that will do the job, but all need weeks to deliver. (Non-electric units aren't readily available or won't keep the rif-raf, i.e. cat,·out.)
Here's my idea ...
-·Use a junkyard·auto window lift to open and close slider; I seem to remember a unit that uses a ladder-shaped plastic chain or belt that I thought I could easily tie 1/16 inch cables to either end·of.
- Motor would open door, pause 2-5 seconds, and close.
- Mount motor at opposite side from slider opening;·cable end 1 goes to near bottom corner of slider to pull it open; cable end 2 goes up and around through 2 pulleys to top far side of door to pull it closed.
- Train dog to push foot switch on inside or outside to trigger open/close event (this is an extension of her sign language for going out); later, to help keep out the rif-raf, add capability for proximity switch ... while protecting against pooch holding the door open by sitting at door watching the squirrels.
- Put springs in cables to protect dog if she sits with tail in door; later,·add safety mechanism similar to elevator door.
- Microswitches to detect when completely closed and completely open.
- Use Dept of Ed carrier and one of the Stamps (which one?) that I might be able to find locally (Radio Shack), to reduce lead time, learning curve, and development time.
I'd like to begin building in a couple of days.· It seems like I could do this for under $200, maybe even under $150.
Any help would be greatly appreciated.· Am I even taking a reasonable approach?· Thanks!
I could really use any suggestions about this project.· I've never built an electronic project, but understand basic electricity, can solder, have wired some of my house and home intranet, have moderate mechanical skills,·and have 20+ years professional programming experience in various BASICs and in research.
The dog now needs out hourly instead of just before and after work.· This is causing great stress, so we're looking for a fast solution (that's not too expensive).· Commercial units to let the dog open our sliding glass door range from $400 for one that won't fit our situation to $1600 that will do the job, but all need weeks to deliver. (Non-electric units aren't readily available or won't keep the rif-raf, i.e. cat,·out.)
Here's my idea ...
-·Use a junkyard·auto window lift to open and close slider; I seem to remember a unit that uses a ladder-shaped plastic chain or belt that I thought I could easily tie 1/16 inch cables to either end·of.
- Motor would open door, pause 2-5 seconds, and close.
- Mount motor at opposite side from slider opening;·cable end 1 goes to near bottom corner of slider to pull it open; cable end 2 goes up and around through 2 pulleys to top far side of door to pull it closed.
- Train dog to push foot switch on inside or outside to trigger open/close event (this is an extension of her sign language for going out); later, to help keep out the rif-raf, add capability for proximity switch ... while protecting against pooch holding the door open by sitting at door watching the squirrels.
- Put springs in cables to protect dog if she sits with tail in door; later,·add safety mechanism similar to elevator door.
- Microswitches to detect when completely closed and completely open.
- Use Dept of Ed carrier and one of the Stamps (which one?) that I might be able to find locally (Radio Shack), to reduce lead time, learning curve, and development time.
I'd like to begin building in a couple of days.· It seems like I could do this for under $200, maybe even under $150.
Any help would be greatly appreciated.· Am I even taking a reasonable approach?· Thanks!
Comments
Do you already have a doggy door the the pooch is trained to push through?· (Guess not if you're letting him out every hour).
You could build a doggy door and put in a solenoid lock that will release via command.· You could buy one of those keychain FOB's and do it from anywhere in the house.· Just another angle....
There's RF Identification too...but that's getting pricey and the dog has to be somewhat close to the reader.
You certainly don't want this to fail and either trap the dog or not let him out.· A small buzzer with a sensor reading where the dog usually sits might be an idea.
Anyhow....you're idea is a great one.· But certainly do it in steps.· If you haven't used the stamps before, then download the 'What's a Microcontroller' text from the website and go through the lessons.· I've only been using stamps a year or two and still go back to the lessons when I forget something!
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
·
Steve
http://members.rogers.com/steve.brady
"Inside each and every one of us is our one, true authentic swing. Something we was born with. Something that's ours and ours alone. Something that can't be learned... something that's got to be remembered."
Dave
Dave
There are commercial units for $100 that can go in a regular door and use IR transmitters.
Unfortunately, we only have the slider.· I could put a panel in the slider to hold it, but it would be awkward to make useful for people.
The walls are over a foot thick, or I'd put a regular electronic dog door in the wall.
The slider is in an alcove, so the commercial slider opener I found won't fit.
Any experience with the car window lift motors?· Is this the right thing?· Do I need anything special to control it besides the Stamp and a relay?
Do I need any other electronic parts besides the Stamp, carrier board, and some microswitches?
Will I be able to get up to speed on the Stamp in a·matter of hours?
Thanks again!
Nate
- dog push switch
- switch makes door open
- some time later, door closes
- switch at either side of door to go in and out
- other switch inputs for proximity (no squirrel hunting) and anti-squish stop
I count 4 inputs, and 1 or 2·outputs here, that is 5 or 6·lines, and any Stamp would be suitable.
The controlling program could arbitrate between the switches and open and shut the door for a timed ammount. The 'hard' parts would be powering this thing (both 5 volts for the Stamp and the high current 12 volts foe the motor) and then driving the motor. Motor drive would probably be simplest & fastest (implying·not cheapest) to use 2 relays, one when·energised makes door close, other when energised makes door open. Engage one at a time for a defined interval (or add two more switches to indicate door fully open and fully closed).
If this still sounds like a GO for you and you'd like some other help (like picking the relays and making the relay drivers from RS parts) let me know.
(Note my clever avoidance of any mechanical door opening items... I'm gonna assume you know how to do that or you wouldn't have started this idea)
Yes ... that is how I envisioned it.
The proximity and anti-squish features would be added later, if needed and if my skills develop.
Timing the opening period, instead of detecting full open is probably OK, but it seems detecting fully closed is safer.· I thought sensing both would be conceptually simpler when I program it.· Does the·Stamp language·have much in the way of event-driven programming tools?
120V outlet is nearby ... I thought I'd just use some RS universal power adapter/chargers (if they can handle the load of the motor).· In fact, I thought one of the Dept of Ed kits came with one for the Stamp.· Off-the-shelf is fine if it's not too pricey.
I can handle cables and pulleys, but am not comfortable with gears and other linkages (as evidenced by the design).· I'm clueless about motors.
I'm not quite confident enough for a go yet, but getting there.· If I get enough feedback on all the aspects, and let it sink in for a little, I think I'll be there in a day or two.··I'll be sure and let you know, since I'll be a lot more confident with help.· I hope to gain some confidence from comp.robotics.misc too.
Thanks.
P.S.· Small children are not an issue.
·
Is the dog to let it self out because nobody is there?· Having the dog door panel in at that time is nice, as it's less of an inconvenience...but it does become a security issue.
I certainly like the idea of a gravity close device (regular doggy door) with an electronic strike (latch).
As far as working with motors....you need to know the load you're moving and the speed you want it to move.· Also, the range of movement (when moving arms/cams).
For the Load...you need to know what kind of resistance the door offers you and if the motor will handle it.· I know there's some equation, but I don't know what it is.· Is the door stiff?· How far does the door need to for the pooch?
You may need to put in some micro-switchs to determine when you are at the ends of your range.· Ever pressed the UP window button when the window is already up.· Your lights in the car will dim a bit.· So you know these things don't have built in limit switches.
Now what you could do is put in some of the bot whiskers for sensing obstacles...or go with some IR or Ultrasonic.· (Laser would be nice, but that's a different thread....although a laser on an opto cell would work for a optic safety switch.
As far as gearing goes....I can't help you there.· You'll have to play around a bit with what you get in the power window track assembly.
Might be an idea to fasten a plate to the door so that you can remove the assembly when you have lots of traffic (BBQ).
You'll certainly want to know when things are closed or open...so you can tell the motor to stop.
Now, is the door going to be usable to humans in this period....then the opening will have to be wide enough.
The stamp is definately capable of all this....but you'll be overwhelmed if you try to do it all at once.·
Read the instruction documents they have on the site and work up to it!· It'll also help you word your questions better when posting on here! [noparse];)[/noparse]
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
·
Steve
http://members.rogers.com/steve.brady
"Inside each and every one of us is our one, true authentic swing. Something we was born with. Something that's ours and ours alone. Something that can't be learned... something that's got to be remembered."
Instead of detecting full closed with a switch, which might not be an airtight closure, could I detect the motor stall instead? Or should I just run the motor a couple of seconds longer than needed to close it?
The first part is the tougher problem as you probably realize. If you solve this problem then a simple coupling might be to turn the door feature on/off via X10 commands based on input sensors (the stamp can output X10 commands via an attached module). That at least decouples the two "systems".
If it were me I'd rather just teach the dog to pee in the toilet somehow!
Harry
The QPROX line of sensors are an excellent choice for touch applications. Here is a link to their web site: http://www.qprox.com/
Regards,
Bruce Bates
I'm not positive, but I think in a vehicle, the window motors are all on a 10Amp fuse (if not 20)....this is for 4windows and there may be other things added on.
If you're just using one motor, I'd say a 3A supply should be plenty.· However, you'll gobble up more current when you try to close an already closed door.· Just be aware that if you lose control of the motor you may burn it out or the supply (if you don't break the window altogether!).
This is where it's nice to have these safety switches.· However, you should have a deadmans switch (can't remember the right term for them)....but you don't want the thing to run at all if something fails.· So using some switches that are not 'sensed' and then something controlled is what you might look in to.
I work on radars and we have limit switches at our extreme operating ranges.· BUT, we do monitor angles as well, but sometimes things get goofy and they'll run on their own.· so, we have some 'sensed' switches where a computer (in this case the stamp) will read them and do some function related to them being triggered)....we also have direct wired switches...this is just in case the dish keeps moving past the first set of switches and then hits these.· These last switches actually trigger a relay that reverses the wiring on the motor and reverses the direction (stopping the dish from banging in to itself).· We have some status lines that go thru an extra pole on the relays so we know when they've been tripped.
Parallax has these little bumpers for their 'bots ( http://www.parallax.com/detail.asp?product_id=27312·)...also, they have this: http://www.parallax.com/detail.asp?product_id=30056·.· It's a flexisensor that will give a reading when something presses on it.· (been wanting one for a while).· This would be good to put in the rail and detect when the door is closed 'enough' OR you could use a simple magnetic reed switch on the door frame with a magnet on the door itself and you'll know when the door is closed when the switch trips.
You could buy those little 'entry' sensors as well.· Radio shack has them along the front entrance of their mall entry.· gives a little ding-dong when you break the beam.· Put one of those on both sides of the door (outside and inside) and if either one is broken, DON'T close the door.
Sorry guys.....I type entirely too much sometimes!!
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
·
Steve
http://members.rogers.com/steve.brady
"Inside each and every one of us is our one, true authentic swing. Something we was born with. Something that's ours and ours alone. Something that can't be learned... something that's got to be remembered."
I made the decision to go with the monster slider.· I bought the "What's a Microcontroller" kit (90005), with HomeWork Board, cable, and on board BS2-1C, so there's no going back now.
Tommorrow noon I'm headed to the wrecking yard to pick up either a seat motor or window lift unless I hear advice to the contrary.· Any reason I should choose one or the other?
It seems that when I close my windows in my car, the motor continues to be powered as long as I hold the switch, even if it is fully closed or open.· So, from this, is it reasonable to assume that they are designed with a high tolerance for stalling?· If so, it seems that timing the closure, with an extra second added might guarantee an airtight, weatherproof seal where any placement of sensors to detect complete closure would likely fail to get a tight seal.· Any thoughts or experience with the motors?
I tested the force needed to open and close the door, and got 4-12 lb depending upon acceleration and whether the door was in the part of the track with the most friction.· Does this seem within the capacity of the motors?
Will the 3A supply be more than enough or just enough?· Or do I need to test the motors to know?
Thanks again.
Bruce
I remember as a kid, having my had stuck in our cars window cuz my brother was being a knob!!· It hurt....so I certainly think i was pushing more than 10lbs.· However, a seat motor moves me and I'm a heffer to say the least!· Seat motors are heavy geared so they have to turn a lot to move the seat a bit.·
Certainly if you're getting these from a scrap yard they'll probably be somewhat cheap.· So, running in to the stops (the 1sec extra), even though it will reduce the life of the motor, won't affect the pocket book too much.· It's the current draw on the supply that becomes the issue.
Your maximum current draw will be when first moving the door (overcoming friction) and when you push past stall.
Do you have either of these in your own vehicle?· Even though they'll be different motors...it would still give you an idea in the current readings!
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
·
Steve
http://members.rogers.com/steve.brady
"Inside each and every one of us is our one, true authentic swing. Something we was born with. Something that's ours and ours alone. Something that can't be learned... something that's got to be remembered."
Be safe not sorry.
And keep that practice in mind during design.
73 spence
k4kep
Garage Door openers have a couple of ways of preventing injury.
1. An 'electric eye' at 1 foot height. When blocked during closure, will open the door.
2. A 'load checker'. If the door is 'stopped' when the door 'closed' switch is not closed, then the door will open.
'Stopped' here means -- well, I don't know how the closer knows the door is blocked. I've done it by hand, and the motor reacts when I'm holding up the ENTIRE weight of the door. Motor load check? Motor current check? I don't know.
3. It DOES have a limit switch so it knows when the door is closed. Proper adjustement of this switch is critical, and can result in quite a lot of 'holding' force.
You have the additional problem of closing on a small object like a dog's tail -- A system of LED emitter bounced off a reflector back to a reader may take care of this.
Kit had a clutch to allow the motor to safely stall in case the window got stuck on something... but the motor barely had enough UMPH to close the window. I wound up pinning the clutch so it never engaged, and to fully close the window it kinda helped to give it a few extra pushed of the button.
Stalling the motor never caused any heat or current problems. I'd never do it for more then a few secs anyway, but unit could tollerate the stall current possibly forever. Never blew a fuse, but the long power wires might have helped act as a resistor.
Bottom line: I'd guess that stalling a window motor should work just fine (as long as it ain't squishing your doggie at the time).
Thought B:· Sensing position can be as simple as a micro switch (RS still sells these, don't they?). Resistor (10K? 100K?) to +5V, switch to GND, common of switch and resistor to the Stamp. Window could have 3 such switched 1 for full up· 2 for full down· and 3 for squish
(squish sensed when squish true and full down false).
·
I went with a window lift (Mercedes 1990). The "transmission" is such that I need no gears, and interface it to my pulling cables as is! It gives me about 18 inches of opening, which is adequate for dogs and people.
My car windows are on a 40A circuit, so I'm still not certain about what power supply to buy. I can't test without one, so I'm in Catch-22 mode. Maybe I can measure my car window motor load ... but need an ammeter first ...
I know it sounds like cruelty, but squish detection will be added on later. There are no small children in the house. The dog's current behavior going through the door is to scoot clear quickly (probably due to experience with overanxious humans). We will observe behavior and play it by ear. I'm thinking electric eye might be the easiest to add on without interfering with the weather-tight seal. I hope to incorporate some spring in the system so that while squishing is possible, breakage is not.
Spence, I can't envision a dead-man's (or in this case a dead-dog's) switch that the pooch could operate to let herself through.
Ernie,
I'm going to sense position with a single microswitch. By placing shims at two locations along the base of the door, and orienting a roller microswitch so that the shims slide under it, and comparing to motor direction, I think I can logically detect the 3 critical locations ... fully open, entering the "safe zone" (the door is flush with the outside of the jamb and nothing can possibly be caught) during close, and fully closed (3/4 inch in from the start of the safe zone).
I'd like to post my code for a quick review when I get to that point.
Thanks again,
Bruce
RS has low end meters for 15 (analog) to 20 (basic digital) bucks. That and a pair of 1 ohm 10W resistors (also an RS item) gives you a 20 amp amm meter, each amp of current shows as half a volt. You could push it to higher currents as you wouldn't be using it continiously (just stop before the resistors glow cherry red). Reading 40 amps intermittantly should not be a problem.
Assuming window lifts have similar power requirements(maybe a bad assumption), I thought I could measure the load on the lift motor in my car.
Couldn't I measure the current drop across the car battery with and without the window operating and take the difference? If I don't start the engine, the only other things on might be some dash lights and running lights. I could even hang a 20lb weight on the window to simulate the extra load from a squished Fido.
P.S. I bought a multimeter. It indicates a top range of 10A (as all the models did). I assume this is adequate. Even though the window is on a 40A fuse, there's other stuff on the fuse, including the other window.
The 10A rating might not be enough on start-up or heavy loading.·
You could put in a heavy wattage 1ohm resistor and measure the voltage drop across that (being that your meter is probably rated to 600Volts).
The idea is:
you could measure the voltage in relation to the current through the resistor as ohm's law says: Voltage = Current x Resistance.
I need a kick in the melon here guys....
which end do I start with.... 10A across a 1ohm resistor is a 10Volt drop.· Which would say there's not enough voltage to kick the motor.
on the other hand....12volts across a 1ohm resistor is 12A....
I always had an issue with where to start to determine drops and all....
Is the series resistor a feasible idea?!
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
·
Steve
http://members.rogers.com/steve.brady
"Inside each and every one of us is our one, true authentic swing. Something we was born with. Something that's ours and ours alone. Something that can't be learned... something that's got to be remembered."
A series resistor is a very good way to read current *IF* you size it properly. My intent on offering the· two RS 1 ohm resistors as a current device might work up to 20-40 amps... but has .5 ohms of series resistance.
That's a 20 volt drop at 40 amps. No way is that possible with a 20A source.
the .1 ohm paper clip idea is a clever workable idea, assuming you can accurately measure the .1 ohm (not really something low wne DMMs can do). You could assume .1 ohms and go on from there. 40A in that is still 4 volts, a big loss.
If I was there with my DMM I'd try the 10A DMM scale, worst case the internal fuse would pop (maybe make sure it has a real internal fuse and not some diodes or such). But even a DMM needs some shunt resistance, may run into the same problem.
Another fast cheap simple power source idea·is a car battery charger. The output is a pulsing DC, but the motor wouldn't mind that for a short time. I used to run my model trains on pulsed power, made slow speed running very smooth.
·
I'm using, and now I need an affordable power supply.
I attempted to determine the power requirements with my own car ...
I removed the 40A fuse and put my ammeter in line, then turned on
the ignition switch. I got a drain of about 1.5 to 2A. Then I ran
the window lift and it ran up to 7 to 9A, so I guess it's drawing
about 7A without major load. Then I kept the window switch
activated after the window was closed, and the meter over-ranged at
its 10A max. So, I'm thinking 20A should be enough.
I can buy a 12V/25A power supply from Radioshack for $100, but I'm wondering
if there's a more cost effective solution.
For example, I can get auto jumpstarters for $30 to $60 that are
rated at 300A/7AHr to 600A/22AHr for starting, but have a 12V jack
for powering accessories and a 110V charger to charge them. Would
these be a reasonable alternative? Could I leave one of these
plugged into 110V all the time?
How about an auto battery speed charger rated at 25A for $60?
I've seen a 6A 110VAC/12VDC converter for $25. It's been too long
since I did these kinds of calcs ... would that be ...
6A x 110V = 660W = 12V x 55A?
Or would it be 6A at any voltage with the wattage varying?
Any advice or alternatives would be appreciated.
Also, how do I find the right relay for the Stamp to drive this motor?
Thanks,
Bruce
Most power supplies are rated in output voltage by output current. So a 6 Amp 12 volt supply will put out 6 amps, not take in 6 amps. Most 'Charger' circuits are designed for short duty-cycles -- which your application might meet. If you over use them, I think they go into thermal shutdown until they cool -- not a good thing for your application.
Summary: The RadioShack 12V/25Amp may be your most reliable bet.
The TIP120 will switch up to 60 volts, and 6 amps -- which is way below what you're trying to control. I wouldn't try to switch 20 Amps with a relay if I could avoid it -- I would think it would wear out fairly quickly. What did the car use to control the motor -- a direct switch connection I am guessing. I'm sure there's a device that would do this, though.
But to be honest, this all sounds like it's getting too big.
If you were to change your gearing then you wouldn't need so much torque off the motor but it would take longer to open/close the door.
Also, if you dog is going out every hour (with the possibility of that frequency increasing) then you'll end up leaving the door open as the contactor will just get too annoying with all the snapping.
Did you think of pneumatics?!· or a linear actuator (for those that watch Monster Garage will know)....basically you apply a voltage and they will push a rod out and pull it back in.· Again, you're getting in to prices, but as opposed to hydraulics that make a mess when they leak it's a bit better.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
·
Steve
http://members.rogers.com/steve.brady
"Inside each and every one of us is our one, true authentic swing. Something we was born with. Something that's ours and ours alone. Something that can't be learned... something that's got to be remembered."
Jim