I just did more testing - mu Multimeter shows the voltage briefly dropping down to about 9 volts for a brief moment when the motor changes speed. It's hard to know for sure as the multimeter takes a bit to show a voltage stop.

I've drawn my circuit as it currently stands and added in the capacitors as I understand your instructions:
1) is this what you meant by across the resistors ?
2) Given I'm only using .39K resistors - would I still use a 1uf capacitor ?

The resistors at the GPIO end are as Numato instructed.

Thanks
Kent.View attachment 226334

1N4733 is a 5.1V zener diode. I had to look that up. There's a proper symbol for a zener diode, different from a Schottky diode. That's a good start, but you still need to limit the current to within the range the 4N35 diode will handle. And you need to watch for wattage ratings as well. I'm no expert on zener's by far. But I think you're reducing the voltage to the 4N35 by 5.1 volts. That is to say that 14V - 5.1V = 8.9V. And at 36V your 4N35 will see 30.9V.

Looking at A data sheet, the 1N4733 is a 5.1V Zener with a max current of 178mA. So your limiting resistor will need to dissipate, as you said, up to 36VDC. So 36 ÷ 0.178 = 202Ω at 6.4W. Your zener won't survive that much power. You can still get 14 to 36 volts dropped down to 5.1V using a much higher value resistor. At 2KΩ, (36 - 5.1) ÷ 2KΩ = 15.5 mA @ 556mW. And that's all just to protect the zener. 556mW is still going to blow the 4N35 diode. So you need a resistor on that line as well. (5V - 1.2Vf{4n35}) ÷ 10mA = 380Ω. However, since this is also subject to a reverse voltage, you still need the 1N4004 diode to protect against reverse voltage because the 4N35 reverse current is only 10µA (0.000,01A).

Now - remember; I said I'm not the expert on zener's so I'll let someone else point out the problems I may be introducing. The numbers I took were from generic data sheets just to get an idea of what you're looking at.

 

1N4733 is a 5.1V zener diode. That's a good start,

The 1N4733 was just a leftover from an older schematic. The TS was only referring to the output side of the opto in that drawing I hope.
Final circuit should be as below.

 

sghioto with your circuit, how fast does the GPIO rise? Won't R1 & C1 slow the rising edge? Will that be a problem?

 

Won't R1 & C1 slow the rising edge?

EDIT:
It will but that should not be an issue. The problem is the voltage on the motor is dropping out briefly during operation. The cap will hold the GPIO input high during those brief voltage interruptions.

 

The original intent of those 5 volt zeners was to limit the voltage that would be applied to the opto LED that had a series resistor chosen for 5 volt operation. It would also provide a bit of reverse voltage protection for the LED. Then the circit evolved and if the final circuit woks then the zeners are not required .

 

Well, I may be wrong, but if there's 36 volts present, doesn't the zener conduct everything above 5.1V? If so then you'd be pushing 30.9V through the opto. And on reverse current the 1N4733 will conduct just like a regular diode, thus exposing the 4N35 diode to 36 volts reverse bias. And it's only capable of around 6V reverse blocking. Above that and it's cooked.

It would be nice, and maybe they do, someone should build a type of zener that will only pass a set voltage despite the input voltage. Like I've said before, I'm not very familiar with zener's. That's why I pointed out what I think is wrong and what I think would be the correct way to address zener's as voltage regulators.

 

It depends on how you configure a zener, if you place them in series they drop a certain voltage, but you can also configure them as a clamp.

Plus other things like providing a reference voltage...etc.

 

I don't know what all this discussion is about using zener diodes.
The circuit does not require any zener or regular diodes as original posted in #41.

 

You can turn a 14-36V range into a 8 to 30v range. Subtract 6 volts. 36V pushes the limit of some approaches like the LM334 current regulator that I suggested. You can set it for 10-20mA to drive an opto-isolator.

 

Well, I may be wrong, but if there's 36 volts present, doesn't the zener conduct everything above 5.1V? If so then you'd be pushing 30.9V through the opto. And on reverse current the 1N4733 will conduct just like a regular diode, thus exposing the 4N35 diode to 36 volts reverse bias. And it's only capable of around 6V reverse blocking. Above that and it's cooked.

It would be nice, and maybe they do, someone should build a type of zener that will only pass a set voltage despite the input voltage. Like I've said before, I'm not very familiar with zener's. That's why I pointed out what I think is wrong and what I think would be the correct way to address zener's as voltage regulators.

THAT is why the zener was originally proposed to SHUNT the LED and it's current limiting resistor, and then have it's own current limiting series resistor to limit the zener current. That means that across the voltage to be detected would be a resistor and 5 volt zener in series, and then across the zener only would be the LED in the opto-isolator and it's current limiting resistor. Sorry for coming up with such a complex circuit that nobody was able to visualize it. Or maybe it was not an adequate description on my part.

 

That's OK Sarge. I already admitted I don't know all there is to know about zener's and how to use them. But at least I've drawn a diagram of something - be it right or wrong. Don't you have any drawing software? I used to (and sometimes still do) use Paintbrush. I think there's a drawing tool in Excel as well. I now have CAD drawing software and that's what I've been using lately.

 

A direct lightning hit on my antenna tower a while back, before I found this site, took out the computer that has the cad software that I had been using. And that was an XP machine, and now XP is quite obsolete, and since the cad was from a download and that computer is no longer recognized I would need to purchase a current version which is rather horribly expensive now for one who is retired, and can no longer get it as a business expense of the company that had me as a licensed "cad seat".
I can still draw sketches and scan them and probably attach them as cad files but they are not nearly as good looking, plus being quite a bit more work. Perhaps some time I will get some decent cad software that has a similar interface.
I did get that free circuit board design software but producing a schematic diagram with that is so very different that I avoid using it.

 

I draw all my schematics with Windows Paint. I keep a file with a bunch of component symbols I can quickly copy and paste.

 

I draw all my schematics with Windows Paint. I keep a file with a bunch of component symbols I can quickly copy and paste.

Yeah, I did that too, keep a file. Now I have a TurboCAD file with components drawn. When I come across something I don't have I create it then copy and paste it into my components file. Makes for much faster drawing.

 

Hi @sghioto and others,

An update on where I got to today once I was able to get to JayCar and purchase the 27K resistors (I already had 10K and 2.2K resistors).

Have now wired up circuit per latest circuit diagram and all appears to be working as I'm wanting coming into my GPIO board - and with the 10uf capacitor in place I'm not getting any unwanted dropouts while the door is opening / closing. I now have four inputs for open and closed (push switched) as well as opening and closing (using new circuits from the motor power supply).

Thank you once again for your assistance (and patience with all my questions). This is the first time I've used this site and it's been a great asset.

Thanks
Kent.

 

OK, the TS now has the system functioning as intended.
It sounds like an interesting project and I would be most interested in an explanation from the TS as to what the project does.

 

Hello,

I draw all my schematics with Windows Paint. I keep a file with a bunch of component symbols I can quickly copy and paste.

I also do something simelar with Colourpaint in Linux, using most times this sheet of symbols:

Bertus

 

OK, the TS now has the system functioning as intended.
It sounds like an interesting project and I would be most interested in an explanation from the TS as to what the project does.

Hi @MisterBill2,

In short - as a hobby I'm trying to extend my 'Home' as much as I can. I have some devices (eg. Phillips Hue lights) which work natively with Apple's HomeKit - but other devices that don't, so I'm seeing if I can add them myself.

Apple Home includes support for a Garage Door, but I needed to know if Open or Closed and when Opening / Closing to make it work properly - plus also the ability to 'press' the door open button. I've already coded the HomeKit side of the equation - and now that I can detect each state all I need to do now is connect the last remaining dots together.

Cheers,
Kent.


Question:
• A Zener Diode is not something I've heard before. I took that it limited the output to the diodes voltage - but a posting earlier today implied that it doesn't let current through until a minimum voltage is reached - did I read this wrong ?

 

Hello,

TS stands for thread starter, wich is given in a started thread by you at your user name.
In AAC Classic (blue) style:

In the default (orange) style:

Bertus

 

The function of a zener diode is similar to diode reverse breakdown, except that it is intentional and calibrated, so that it can serve as a voltage reference. In the forward direction the voltage drop would be similar to a normal diode, while the reverse breakdown voltage of the 1N4733 is nominally 5.1 volts at some specific current. The plan was to have the excess current all flow through the zener diode so that a fairly constant voltage would be available for a second resistor in series with the LED could hold a fairly constant current.
I had not anticipated that there were so many unfamiliar with the operation of a zener diode, but were quite willing to comment anyway.