I've got a CR10X that seems to lose it's mind when I connect it to the pulse generators that measure my electric power consumption, and I suspect it's the long (250-foot?) CAT5 wire that leads out to the power transformer.
Is there a standard (CSI or other off-the-shelf) opto-isolator that would provide a pullup in one direction and a contact closure in the other? Yeah, it's a very simple function, and I _can_ built it myself, but with the power supply and the housing and the connectors and the breadboard and the soldering I'd much rather buy it if it's readily available...
Thanks!
What's the peak voltage of the pulse?
It's just a contact closure so I'll need some random pullup, probably 5 volts.
If it is just a contact closure, surely it is already isolated, as is generally the case in electricity meters?
Yes, but the isolation is 300 feet from the datalogger at the power transformer, and I want to isolate it in the building where the datalogger is.
I have a cable attached to an electric meter in the basement of a hotel and run it up to the 10th floor roof mechanical room. I do have one end attached to the logger +5. It works fine.
You could try this:
http://www.campbellsci.com/p7321-specifications
Connect this to 12 Vdc and your contact closure in the input side. I'm not sure if it's fast enough but I don't have one here to try. If not this, perhaps a standard 12 Vdc relay not an SSR.
http://www.radioshack.com/product/index.jsp?productId=3020762
* Last updated by: IslandMan on 2/15/2011 @ 4:07 AM *
IF you want to follow down the route of adding extra isolation you will need a source of power on the transformer side of the isolation. The 12V supply mentioned by Dave could not be from the datalogger but an independent supply. In addition the solid state relay suggested may not work as there is a recommended minimum load for those devices, so there may be enough leakage to make it look like "the switch is closed" all the time.
If the problem is not really isolation but a noise pickup issue, there are simpler solutions. If you are connecting the switch closure to a pulse counting channel, the internal bias to 5V is 100 kohms which can make noise pickup an issue. You can reduce this by connecting a resistor between 5V on the wiring panel and the pulse input. You could use a resistor as low as 1k to give 100X less sensitivity to noise.
If you are using a control port and the switch is connected between 5V and the control port, connect a 1k resistor between the control port and ground.
I'm almost certain the problem is (lightning-induced?) surges on the measurement lines, when the CR10X is connected to the power meters (300 feet horizontally, at the transformer pad)it loses it's mind within a few days or weeks and starts giving either fixed answers or nonsensical ones, but when not connected it runs just fine.
I've ordered some Opto-22 modules and (I had forgotten they make) a
http://www.opto22.com/site/pr_details.aspx?cid=4&item=G4PB4 so I can put something together without too much effort.
Thanks for the feedback!
I would advise making sure the datalogger is connected to a good ground as it sounds like the problem could be cause by either its ground reference bouncing around or current flowing between different grounds connected to the datalogger.
Also make sure the power supply connections to the logger have some lightning protection too.
Make sure any optical isolation you fit is designed to survive lightning surges as many opto isolators only isolate to 1500 V or so, whilst lighting surges can be tens of thousands of volts.