GPIO Change voltage

[jrich523]

Im looking to help automate my garage door and from what I can tell, it sits at 3.8V normally, but when the door button is pushed, it changes to 1-1.5V and when the light button is pushed it changes to 2v.

Is there a way I can use a GPIO pin to normally allow that to pass as is (let the physical buttons continue to work) but also allow me to digitally trigger that?

Also it seem that 3.3v is the max for the GPIO pins? which I would assume is fine for this since my guess is that its probably based around 1v, 2v, 3v aproximations?

I've done the stupid stuff like making lights change and what not, just not exactly sure how to tackle this problem, but it seems like it shouldnt be too bad to do?

Thanks

[gordon77]

It would help if you had a circuit diagram of the garage door system.

Could you connect a relay across the door button?

[boyoh]

Do as Gorden suggested a circuit diagram would be very helpful.
Just a word of warning, The automation must be fail to safe
If your intention is to operate it from a remote distance not
In view , this can be a danger to somebody near by
Sugest a external safety isolator

[jrich523]

I'll try to gather more info/pictures and some circuit diagram.

The reason Im using a Pi rather than Arduino is that it will be easier for me to program it as a "device" with in Domoticz (my home automation system) which I can make handle the "fail to safe". Thats sort of the whole reason for this. As I pull out of the drive I have on several occasions forgotten to close the door and dont realize it until im a bit too far to turn around. Mostly just looking for a way to remotely close it when I screw that up

[ElEscalador]

Use a relay in parallel with the manual button inside the garage. If you can also rig up an end switch that confirms when the door is closed, PM me and I'll send you code that confirms the whether the door is opened or closed before it triggers the door to move and will also tries again in case triggering open instead of closed (or closed instead of opened. It's actually two programs- one for open and one for close).

[jrich523]

Im not sure how a relay would work here. A relay is just a switch (on/off) where what i need here is voltage control.

To make this a tad easier, im going to use round voltages, but because i measured at the wall unit the distances causes a voltage drop, so im sure it will take some testing and screwing with, depending on how far the pi is from the motor control.

From the wall control its happily sitting at 3v
When I push and hold button 1 (door toggle) it changes the voltage to 1v
When that line is changed to 1v, it will make a door action depending on its current state. for each press it would do, open, hold, close.

so, outside of knowing its state (one problem at a time i need to sit inline with this controller, and be able to read the changes in voltage so that i can code the state tracking. From a control perspective, i need to change the voltage from the 3v down to 1v for lets say, 1 second (i'll need figure out the actual required time, but that shouldnt be hard)

The other button is for the light, it works in basically the same way but it changes the voltage to 2v to trigger an action

now I havent decided where I'll be putting the pi, so the specific voltages arent super important yet, other than, i need to be able to read the current voltage of the circuit, and be able to change it.

If i can get that part, the rest of it should be pretty easy for me to put together.


I had attempted to get some info on it but the manuals are pretty worthless and i couldnt find any part sites that provided anything useful in regards to a schematic.

Its a Genie PMX500IC/B ProMax and the wall control is the GPWC-BX (I think any Series 2 should be fine)

Thanks!

[drgeoff]

If the existing button is merely a switch then a normally open set of relay contacts across it is the safest option and is almost guaranteed to work.

Do not confuse 'cause' and 'effect' when you measure 3 volts changing to 1 volt.

[Burngate]

May I ask for some clarification?

you mention two buttons - the door button and the light button - and a controller and a garage door.
You also say it changes between 1v, 2v, & 3v

What is the "it" that changes? Is "it" a point on the controller where the buttons connect, or a point elsewhere?
And what reference are you using to measure "it" against? A ground connection of some sort?

[ElEscalador]

Somewhere, the door control comes down to a switch being closed for a moment. Many (most?) Garage door openers just require a pair of contacts to be momentarily connected to trigger. I'd bet a few $ that's all your button does at end of the day.

[jrich523]

Ok so im attempting to make a diagram for the switch now that i've taken it apart, and what i've learned is i have a lot to learn

here is a pick of the controller. the image of the back is flipped so left is left on both pics. Im still pretty confident that this is handling the events based on voltage control.




There are a couple of things I was confused about. One of which was the use of the switch.

numbered as 6-A-5 (A came after i realized it was important) and i checked the resistence on it (basically what is opened or closed)

Here is what i came back with when I test two points in both positions of the switch. The switch is basically a "lock" so in the Unlocked and Locked state are listed below

| Pin | Unlocked | Locked |
| 6 - 5 | closed | open |
| 6 - A | closed | open |
| 5 - A | open | open |

so this seems like in the "locked" state everything is open, which means its probably bypassing the reisisters, keeping the voltage at 3v no matter what button is pushed.

The only path that works while its "unlocked" is the 5-A, which appears to just be connected anyways via the... solder route? sorry i suck at this, still trying to get up to speed :-/

Also, when screwing with the buttons (was wondering if it was always on and went off on press, which its not) but i did notice that depending on which of the 4 nodes i touched it seemed to leak a bit.. but if i recall correctly thats normal?

also from what i can tell of the resistors, they are, on the left side 10.1k Ohms 1% and on the right 8.25 Ohms 1%. I had a hard time identifying some of the colors.. silver vs gray etc, but im fairly confident its correct

which, since i suck at the rest i wont attempt the math, but im guessing if you look at those at 3v's those resisters drop the line to either ~2v or ~1v

One other note, and i've havent screwed with this a lot... but the light goes off (maybe just really low) when the button is pressed, which i would assume is indicating that its using voltage control here, provided the light is dimming and not actually going out.. with the light in the middle like that, if it is going dim its unlikely it would produce enough light to be seen on the sides

hopefully this can shed some more light on it..

Thanks!

[jrich523]

May I ask for some clarification?

you mention two buttons - the door button and the light button - and a controller and a garage door.
You also say it changes between 1v, 2v, & 3v

What is the "it" that changes? Is "it" a point on the controller where the buttons connect, or a point elsewhere?
And what reference are you using to measure "it" against? A ground connection of some sort?

Sorry, it would be the remote control. I guess i had originally considered basically tossing the pi next to the wall control, but the more i think about it, its probably better to have it with the motor part. I think the only difference will be perceived voltage (~3v at wall, ~5v at motor)

[drgeoff]

I reiterate what I said above. If the push buttons are 'open' and 'close' for the garage door then use a couple of relays. Put the normally open contacts across the button switches and program the RPi to operate each relay to emulate you operating the push button. You do not need to remove the button switches. Either the button switches or the relays can be used to operate the door.

[jrich523]

the buttons are not 'open' and 'close', the are 'change door state' and 'turn on light'

button 1 - Open/Hold/Close (depends on state of door)
button 2 - Turn on light

tap button 1
> open door
tap button 1 (while opening)
> hold door in current position
tap button 1 (when completed opening, or in holding state)
> close door

tap button 2
> light turns on for 5 min

no other actions are handled by this, it might extend the time if you press it again but i doubt it, and dont really care too much, no real need for that light on the motor, the garage has a ZWave controlled light switch for the main lights so i'll never need to automate anything with button 2.

[drgeoff]

the buttons are not 'open' and 'close', the are 'change door state' and 'turn on light'

button 1 - Open/Hold/Close (depends on state of door)
button 2 - Turn on light

tap button 1
> open door
tap button 1 (while opening)
> hold door in current position
tap button 1 (when completed opening, or in holding state)
> close door

tap button 2
> light turns on for 5 min

no other actions are handled by this, it might extend the time if you press it again but i doubt it, and dont really care too much, no real need for that light on the motor, the garage has a ZWave controlled light switch for the main lights so i'll never need to automate anything with button 2.

OK, just use a relay driven by RPi to emulate a momentary press of button 1. That will have exactly the same effect on the door as you tapping the button.

[boyoh]

The lack of electronics knowledge is stopping you from completing this project
You will only except one answered, that is the answer you want to hear.
I don't think you will get that answer, The safety factors in corporate , in
Automatic doors are very strict, Do you have safety pressure sensors in
In the control system incase of door obstruction. My advice is stick to local
Visual control of your door, Better Safe than Sorry

[jrich523]

I dont think that im expecting "one answer", or rather, maybe i do, but i dont have any preconception of what that answer should be. the question is a much more high level one than what you guys are assuming. It is everyone else that is looking to provide one answer, the answer to how they would solve this problem. Which i've very much enjoyed hearing, however where i really want to go with this is, "how does it work". I would like to be able to prove that it is using voltage control to operate the door. not because i feel im going to solve it that way, but as you pointed out, my experience level is making this much more difficult and knowledge, of any kind, will help that..

so approaching it like a teacher, rather than an engineer, is really what im looking for. I made an attempt to figure out what its doing, i've told my methods and their results. It would be nice if someone could maybe fill in the gaps about what i've done wrong, any assumptions i've made... that sort of thing...

on top of that, whats wrong with taking the voltage control approach if that is how it works? the advantage there is that i dont have to modify the wall unit by soldering to its buttons, i could just wire it in to the middle of the wall controller and the motor. Or maybe im going to build custom wall controls in the end and need to simulate exactly what its doing now. I'd prefer to have options first, and then decide rather than just pick the first method tossed at me.

I dont doubt that tossing a relay in there and having the pi wired in to the buttons, rather than inline will work, and might be a fine solution in the end, that doesnt mean i dont want to know whats going on here. Another note on that, I had been talking about putting the pi inline, meaning have it site between the wall unit and the motor unit, so when people say 'toss a relay on it' my mind is still on the inline track and it didnt initially occur to me that the relay method had to do with soldering leads directly to the buttons, so you can see how i could be confused there.

And as far as your comments around safety, thats WAY OFF.. i dont know what sort of garage door openers you guys have in the UK, but here in the US all the safety features are on the door itself and what im doing would i no way disable any of those safety features. If anything what im doing is because I want more safety, I want to be damn sure that the door isnt left open when no one is home, or even perhaps over night, where someone might use that open door to enter the house... If you feel that you dont want to provide any help, thats fine, but telling me that im seeking one answer and am ignoring safety is WAY off.

Its like me asking "hey who left the door open" and you responding "you just need to close the door! why do you keep asking!"

[mattmiller]

Im looking to help automate my garage door and from what I can tell, it sits at 3.8V normally, but when the door button is pushed, it changes to 1-1.5V and when the light button is pushed it changes to 2v.

What people are telling you is that those voltages are most probably the EFFECT of your existing control system and not the CAUSE of getting the door to operate.

That is why you are being advised to connect relays to directly emulate actually pushing the buttons.

Note: This is 99% the best way to interface to ANY low voltage existing control systems such as your garage door.

[RaTTuS]

And as far as your comments around safety, thats WAY OFF.. i dont know what sort of garage door openers you guys have in the UK, but here in the US all the safety features are on the door itself and what im doing would i no way disable any of those safety features. If anything what im doing is because I want more safety, I want to be damn sure that the door isnt left open when no one is home, or even perhaps over night, where someone might use that open door to enter the house... If you feel that you dont want to provide any help, thats fine, but telling me that im seeking one answer and am ignoring safety is WAY off.

Its like me asking "hey who left the door open" and you responding "you just need to close the door! why do you keep asking!"

the saftey issue is not about your home safety it is about you killing / injuring someone when you cannot see what is happening
a small child , dog , postman , could be in the way of the door when you trigger it crushing them to a massive amount of litigation, and or jail time -

[pcmanbob]

Hi jrich523

I will try to explain what you would need to do if you did not modify your existing controller simply and why.

1. you would have to replicated your existing controller but replacing the switches with relays that the Pi can operated.
2. having looked at you picture it does no look like you can daisy chain controllers ( that is add one after another) so you would have to remove you existing controller and replace it with the one you built.
The reason for this is if you normally have 3v on your control wire you cant have another controller trying to put 1v on the same wire it will have no effect.
so you would loose the option of operating you garage door via the existing buttons.

Now if you do as had been suggested and just connect relays across the existing switches
1. its easier than making a copy of the existing controller
2. it allows remote ( via the Pi ) and local ( via the switches) operation.
3. if you change your mind and want to remove the remote operation function it just a simple unsolder job to remove the relay connections.

I hope this has explained it in a way you can understand.

[boyoh]

HSE advise that those in control of the maintenance of electric gates to regularly review their risk assessments, taking account of or any changes to the operating conditions or environment, and that they have taken appropriate steps to address them. If the general public can access the ‘access device’ then additional protections should be in place and those maintaining these devices have a real duty to ensure they take their responsibilities seriously to make sure that anti-crushing, shearing and trapping safety protection devices are correctly set and maintained.



So to adhere to the European Safety Standard, you must make sure you carry out a full risk assessment of the automated access installation you have. This will highlight any modifications required to comply with and that are set out in the BS EN 12445:2001, BS EN 12453:2001 and BS EN 41-1 and the EU Machinery Directive 98/37/EEC. After they have been evaluated it is your duty to make sure that levels of protection are installed and working to ensure the maximum level of safety to prevent serious injury.

[drgeoff]

To emphasise a point already made. If you do not use the relay across existing push button method you will need to remove the existing control button circuitry. That means your new method (RPi or whatever doing your "voltage control") will be the only means of controlling the door. If the RPi fails or crashes you will have no control. The relay method enables you to keep the existing buttons in situ and fully operational. Even if the RPi goes haywire and starts randomly opening and closing the door, the simple and quick action of unplugging the PSU to the RPi will effectively disconnect it from the door and give you control through the buttons.

[jrich523]

@mattmiller - I guess i dont understand the difference between triggering the 'cause' is any different than simulating the effect? I guess the risk is that i incorrectly simulate the effect and I think thats a fair statement and agree with you there, but with something this simple i wasnt as worried about that? Also since there are only a handful of results for that one button I had assumed it would be easy enough to prove if simulating the effect was enough. But again, completely agree with your point and that is a great reason to lean towards the relay

@rattus & @boyoh - The safety features you're talking about are handled by a laser to look for objects in the way, which is wired directly back to the motor unit and would not be impacted by what the remote is doing. The other feature is a pressure stop, which I believe is handled by the motor unit itself. As I said, the safety features would not be impacted by simulating the remote. On top of that, that would be something I would validate after adding something in to the mix. I have not forgotten the importance of safety here, I just havent brought it up as a topic because I believe I have that area addressed.

@drgeoff - That is more so what I was looking for, if i do go down that approach what limitations do i have, which do your point, it would basically conflict with each other. The real thing I was wondering if it was possible to add the RPI is series and have it be able to read if the voltage changed and just pass it on through, and then also be able to change the voltage. I guess basically what you're telling me is it wouldnt be that simple and would increase the number of things that could go wrong.

Thanks guys, that we helpful and I think I have the info I need now. The information and assistance is much appreciated.

[boyoh]

One last Idea to push the button is to use a small 12v Solenoid, This would be non intrusive on the door electronics
Suggest a proximity sensor be used covering the door entrance,

This is only a Idea if you persist with your project

[4LULZ]

Is this the manual for your garage door motor? http://www.geniecompany.com/data/produc ... 535556.pdf

If so have a look at page 20, Fig. 4-1a (or b) - the wall switch that you have currently is creating a temporary connection between the terminals labelled 1 & 2, so you can just connect the NO and COM terminals from a 5v relay to those and the control the relay using gpio write commands.

I setup a garage door opener based on this instructable a while ago and it worked really well, maybe it'll help - http://www.instructables.com/id/Raspber ... or-Opener/

[jrich523]

@boyoh - thats an interesting idea, i'll look at that more, thanks

@4LULZ - Its a PMX500IC, but the front page of that guide says it supports Series II controllers (which is what my wall controller is)

Im not sure how you can tell its based on being open/close from that, but I did just test it with the wires and a simple tap of the wires will get it to open/close but it will only do the fully open, fully close, what i mean is, i cant issue the "hold here" with that method.

I'll look at that method a bit more, I cant really think of a reason I would need to "hold" the door remotely so this might be good enough but will def require some more digging and testing.

Thanks