Open collector outputs and M1G zones

[apostolakisl]

I own one of these.

http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=170640310219&ssPageName=STRK:MEWNX:IT

But you will need 5v outputs from something else to control it (logic level ttl is all the juice you need, which is why I bought it, I use it to convert the CAI webcontrol outputs to full size relay closures).

It looks to me like you are trying to control Elk zone status using a computer? You would need a way to output 5v from your computer to control this relay board.

I think the gizmo you have already will work just fine. I think the doucmentation lacks. Just take an ohm meter and test the screw terminals see which ones are closed to ground. The turn them an on and see what closes to ground then. Then wire it accordingly. Elk zones are not going to be bothered by any slight noise or electrical "leak", especially if the only thing you have connected to it is your Elk and you use the Elk's 12v aux power to power the unit (check your total Elk amps if you do this)

 

[Lagerhead]

FWIW --

There is a Wikipedia article specifically on "open collector" (duh) -- /wiki/Open_collector

You can make a quick-and-dirty optoisolator with an IR LED and a compatible phototransistor pointed at one another inside a short length of opaque tube. This can sometimes be easier than working with a DIP package. Be sure to mark the leads with boh "input" and "output" and "+" and "-" at the time you make it.

 

[Ira]

As suggested, I did some continuity checking on the OC output device. For each OC output terminal pair, one terminal is connected to a shared ground. If I use an external relay (and an external PS for the relay), and connect the external PS ground to the OC output device's shared ground, I don't need to connect anything to the "extra" OC terminal to make it all work. That makes it look more like all the other OC output devices I've seen while researching this that only have one terminal per OC output (and a shared ground).

I will probably use relays between the OC output device and the M1G zones, even though it will probably work without them. I can get a DPDT relay and DIN mount for about $10. Not cheap when I need 16 of them, but not too bad. Plus with DPDT, I can switch something else at the same time, like maybe an LED mounted somewhere in plain site. Anyone know of an off-the-shelf, DIN mounted optoisolated relay that is no more than 1/2" wide, and costs about $10 or less?

Thanks,
Ira

 

[Lagerhead]

Anyone know of an off-the-shelf, DIN mounted optoisolated relay that is no more than 1/2" wide, and costs about $10 or less?

Consider skipping the relays. You can get four optoisolators in a 16-pin DIP package for $1-3 per chip. Total under $15 including sockets for sixteen channels, and every output is individually isolated. One example is the Liteon LTV-847, 79c at Jameco, but there are many other options. Check Digi-Key, Mouser, Jameco, etc.

The downside vs. relays is the small current capability, but they are perfectly satisfactory for triggering security zones.

 

[Ira]

I did some searching and couldn't find a socket that could be DIN rail mounted. I like the idea of the optoisolator, and the "4 in 1" sounds good, but I really don't want to rely on my soldering skills. Also, wouldn't the optoisolators require something like a 1.5Vdc power supply? If so, those look like they are hard to come by in a wall wart form factor.

Thanks,
Ira

 

[gatchel]

They make the same version of that board with relays, (I think).....It's much more pricey.

 

[Lagerhead]

I did some searching and couldn't find a socket that could be DIN rail mounted. I like the idea of the optoisolator, and the "4 in 1" sounds good, but I really don't want to rely on my soldering skills. Also, wouldn't the optoisolators require something like a 1.5Vdc power supply? If so, those look like they are hard to come by in a wall wart form factor.

Thanks,
Ira

You can power the optoisolators from the same power supply as your controller board. This makes sense as they will have ground in common with your controller anyway. Voltage doesn't really matter -- currrent is everything with LEDs, and the trick is to use suitable dropping resistors once you decide on a voltage. Probably one resistor per IC package would be adequate to provide the current limiting.

The sockets I referred to are DIP sockets for the optoisolator ICs. These sockets would go on a perf board or similar. The DIN mount would be a different matter entirely, as the last stage of the build for final mounting.

Soldering is not mandatory -- you could wire wrap instead. But I would solder to avoid the "thickness" that would result from the long wire-wrap leads, and either cluster the chips on the board, or maybe line them up with a long parellel row of on-board connector terminals.

It's really not difficult but I guess it helps to have done some small circuit assembly to have confidence here. And I acknowledge that my cost estimate does rely upon already having the tools and spare blank boards and a parts box with miscellaneous other hardware etc.

 

[Ira]

I asked the manufacturer about the "extra" ground terminal for each OC output. It's basically just a convenience feature. His example was using multiple relays, LED's, etc. with the board, each having its own PS, and each PS being a different voltage. The device has 16 OC outputs, so tying 16 grounds together at one place would be a little cumbersome, and if you wanted to add/remove one device and its PS, that would be cumbersome, too. Having ground terminals for each PS used makes it easier to do this.

 

[apostolakisl]

Ira,

Just wondering what your root problem is? What problem are you trying to solve by having a device that changes the status of Elk zones?

Lou

 

[Ira]

I'm using their wireless mirror relays for a few different projects. The main one is to physically isolate my lawn irrigation system valves/wiring from the controllers for lightning protection. For each zone valve, instead of just opening up the valve, the irrigation controller energizes a relay which causes a contact closure on the wireless transmitter. The transmitter sends a wireless signal to a wireless relay controller that energizes a relay on its end. That relay closure results in 24Vac going to the zone valve and opening it. It can do this for each of the zone valves in my system.

One of the features of the mirror relay devices is that there is two-way communication between them. So when the transmitter tells the relay controller to energize relay 3 (which should result in zone valve 3 opening), the relay controller will send signals back telling the transmitter (which is also a receiver) that it energized relay 3. Another device that is hardwired to the main transmitter takes the returned signal and turns on the OC output for zone 3. That OC output will ultimately close a M1G zone so that I can tell that the valve should have opened.

It's not a perfect feedback solution, but it's good enough for the application. There are points of failure, but the chances are a little slim. The main one being that the response from the relay controller is basically an ack. If the relay failed, or the valve solenoid failed, etc., the ack would still be sent and the M1G zone would still be closed, implying that the zone valve is open. However, I have other ways to detect those types of failures if I want to. For example, I'm on a private water well system. I get an M1G zone state change whenever the well pump starts and stops. The way my irrigation system is set up, the well pump will always start within two minutes of the irrigation system starting (maximum storage tank draw down). It will run continuously until about one minute after the zone valve closes (except when another zone valve opens) to fill the storage tank. So if a zone valve is supposedly open (according to the M1G zone state) for more than a minute, and the well pump isn't running, something is wrong. I also plan on putting a flow switch on the main irrigation line, along with a water meter, to give me more/better information.

The "regular" relays that cause the mirror transmitter to see a contact closure are DPDT relays. One set of contacts does the contact closure for the mirror transmitter. The other contacts on each relay are unused right now, but if I really want to go overboard, I've thought about having those contacts also connect to M1G zones. That way, the M1G knows whenever one of the irrigation controllers says to open up a zone.

I say controllers (plural) because I can actually control the irrigation system from three different physical devices (albeit no more than one device is "in charge" at a time). I can control zones from the original Hunter Pro-C controller. I can control them from a Rain8NetPro using McsSprinklers (or via the zone buttons on the Rain8 itself). And I can control it from M1G outputs using M1G rules, CQC custom written rules, or the CQC irrigation "driver".

There's absolutely no doubt in my mind that I've gone completely crazy on this thing, but it was/is a learning experience for lots of different hardware and software that I want to become proficient with. The system is still in the testing stage, but the mirror relays were the last (expensive) piece of the puzzle. I should be switching over the new system in a week or two if no problems crop up.

Ira

 

[apostolakisl]

When your main transmitter recieves the ack, what does it do? Does it turn on a voltage output? If so, what are the specs?

What voltage does the transmitter board use to power itself and how many amps?

 

[Ira]

The OC output device is connected to the "master" transmitter via a supplied ribbon cable. When the master gets the ack from the remote relay controller, it signals the OC output device. The OC output device turns on (or off) an onboard LED for visual feedback and turns on a OC output.

An important thing to know is that the ack is merely and acknowledgment that the command to open/close a relay was received. It does not mean the relay actually opened or closed. According to the manual, a relay could be closed, but if the ack was missed, the OC output device would not show it as closed. There is a "refresh" command that supposedly will catch such a scenario, but I haven't tested it yet. It's unclear in the manual, but it looks like when a contact closure on the master causes a command to be sent out, the state of all relays is returned. So you can dedicate one of the master inputs to doing refreshes at given intervals if it is important to the application.

The boards I have all require 12Vdc regulated power. No idea what the power consumption is and couldn't find it on the website. I asked the manufacturer for the info and will post back his response.

In case it matters to someone, these boards are supposedly all "made to order" in Missouri at the time the order is received.

Ira

 

[apostolakisl]

So the OC is integrated to the rf realy board by the manufacturer's specific design.?

The reason I ask is that if they all run on 12vdc and they don't exceed the amp specs on the elk aux out, you have no need to separate your electric grounds at all. If all of the components live off of the Elk's 12vdc, separation is irrelevant. Perhaps you can put your multimeter on the board and directly measure the current?

I suppose an internal failure of the OC or the rf relay transmitter could damage your elk, but somewhere in there you just have trust that you aren't that unlucky.

And short of having another set of rf relays where the relay closing on the sprinkler zone also closes a seoncd set of rf transmitters sending a signal back to a second set of relays connected to a zone on the Elk, I see no way to be absolutely sure that the zone relay is closed.

However, even if you do this, you still won't know that your solonoid or some other part of the irrigation has failed.

Hat's off to you here with your super anal approach :rockon: . Personally, I use the "walk around the yard every now and then" approach looking for stuff that seems to be lacking in h20 and escalate the investigation if need be.

 

[Ira]

Yes, the OC output board is proprietary. It can only be used with their master transmitter. It is a different board, but is connected to the master with a ribbon cable.

The board manufacturer recommends separating their boards from other devices' PS's, unless you put in other stuff to control transients, surges, etc. I guess they consider their stuff to be on the "sensitive" side.

My final line of defense against zone valve problems (being open or closed when it shouldn't) will be a combination of M1G zones knowing when a controller is calling for an open valve, when the well pump is running, and when there is flow thru the main irrigation line. CQC will consolidate the info and make a decision as to whether something is wrong or not (mainly that water is flowing when it shouldn't), and if necessary, shut off power to the well pump. Knowing things like the well pump should turn on within two minutes of a zone valve opening and staying on as long as a zone is open, and that the well pump should almost never run for more than two minutes unless a zone valve is open (i.e., during normal household usage), helps make it possible to identify problem scenarios.

 

[apostolakisl]

The board manufacturer recommends separating their boards from other devices' PS's, unless you put in other stuff to control transients, surges, etc. I guess they consider their stuff to be on the "sensitive" side.

I guess it boils down to what are you worried about damaging what. I think the Elk has a very high quality power supply that is relatively well protected, but if a transient originated in the Elk, it might enter your OCO. Essentially, your OCO would become like all of the other boards on your Elk as far as vulnerability from a transient that originates in the Elk network. As far as the OCO damaging the Elk, wiring it this way makes the risk so low that I think you could consider it zero for all practical purposes. Considering that the OCO costs less than $100, I would consider about $10 to be the most I would want to spend protecting it.