Is there a 1-Wire(*) device that can detect the presence of mains? I suspect that the DS2413 might do it but ideally I'd like it to be galvanicly isolated so there is no chance of mains getting where it shouldn't and frying things.
The application is to monitor the various pumps and demands on the heating system here in conjunction with a scattering of 1-Wire temperature sensors.
(*) "1-Wire" as in using the same system as the cheap temperature sensors you can get from Dallas/Maxim.
Glue one of the temperature sensors you are already using to a suitably insulated resistor.
John
Volts or amps?
Only, volts on mains wiring is really hard to detect reliably with a high resistance isolation as it is so easy to pick up stray voltages.
How about something really simple, like a mains powered relay, an opto- isolator, or a home brew optoisolator using a photo-device pointing at a fully sealed mains LED unit?
would a mod of something like
or use a 1 wire a/d onverter and ldr - with a neon on the suppy you are monitoring - giving electrical isolation and cleanish switching e.g. DS2450
If you aren't doing fast switching use a mains relay (preferably with gold-plated contacts) feeding a DS2413. It's easier than having to filter the mains input and gives excellent isolation. You may still have to debounce the input in software though.
Presence ie is something switched on? Loads are around 50W for circulation pumps or oil boiler. I've been half looking at hall effect switches to feed into a DS2413 1-Wire dual channel PIO device. But I haven't a clue about the field strengths that hall effect switches need or those that you'll find around a mains cable carrying
200mA or so.I have a load of relays that could be liberated from the old control box but that just seems a bit over kill and "old fashioned" but does solve the isolation side. I can't help feeling that there must be something from the process control area for this but perhaps they don't come cheap...
Mains LED indiactor, opto switch into a DS2413 that's another way. Might be a fair bit of messing about with heatshrink to make light tight units though.
There are voltage and current sensors available, but you probably don't want to pay the cost - especially for the current sensors! There are also little hall effect sensors at a more reasonable price. The problem is that they give an AC output around zero (and need a negative supply). I once used a system of connecting an opto-isolator & resistor to watch the voltage across series diodes (actually a bridge rec with + and - shorted), but you get a 1.2v or so drop and a pulsating DC signal to deal with. A relay to watch just the voltage is a lot easier!
I retired from electronics 5 years ago, so questions such as this are of interest. I followed up this DS2413 via Mr Google - now I did expect to be able to understand quite what this device does, but I'm obviously already being left far behind as the description was just so much gobbledegook.
Could some kind person give me an idea of what it does in words on one syllable plain English please.
Rob
Looks like a very handy device, that...
The only way I can think of doing a reliable non contact volt-detection is to do something like slip the conductor through a small bit of brass tube (the detector), then oversleeve that with some plastic, then oversleeve again with bigger brass tube which is earthed to provide some local shielding from other stuff - but then you'd probably need to amplify the signal which leads to a problem of power for the amp circuit.
Unless the DS2413 was sensitive enough to cope directly.
It mentions it has 1M input impedance and a max voltage of 28V on the IO pins - wonder if a signal diode->tiny capacitor+clamp zener+high-ohms bleed resistor could clean up the induced mains hum on the pickup (ie turn it into a DC signal) and still provide enough oomph to the chip???
If it could be made to work reliably, it would be elegantly simple, intriniscally safe and the pickup assembly could be made split for slipping onto the conductor with a finishing turn of metal foil sticky tape to complete the earth shield.
It's a massive conjecture because I've always been crap at analogue stuff...
It's a digital i/o device with two pins, either of which can be used as an input or an output.
The way the outputs are set and inputs read is via serial communication. Unlike most serial communication, 1-wire devices (they actually use two wires) have a 0V line and a communications line - the comms line is held at 5V when not communicating, allowing the device to power itself from the comms line and store enough power to carry on working during communications. A number of devices can be daisychained together and each addressed by its unique 64-bit code, which is set during manufacture.
There's a whole lot more to it, but that's the basics.
SteveW
A Y rated capacitor voltage divider driving a fet would be far simpler and meet more legal requirements.
NT
That's effecitvely what I proposed, only the capacitor is formed between the wire's insulation and the plastic between the sensing electrode and and the earth shielding.
I was trying to meet the objective of total isolation, hence no direct contact with the conductor.
Interesting idea with the FET - that could help. Still need to convert the AC signal to a damped DC one though otherwise the chip is going to have a hard time deciding if the power is "on".
Cheers
Tim
Anything capacitive will be horribly subject to transmitting any noise, usually many orders of magnitude more effectively than the signal you're actually interested in.
Use an opto-isolator, which mostly give you 4kV isolation and no differential grounding problems. If you select a part with darlington output, you can switch it with as little as 1ma driving the LED side, which you can easily derive from the mains with a dropper resistor, and some rectification or a protection diode, and a fuse for safety. If the DS2413 has an internal pull-up or pull-down, you could connect the opto isolator directly. Otherwise, you'll need to handle providing a pull-up or pull-down and micopower from the
1-wire bus.
Thanks Steve - I should have said I've played with the temperature sensing versions of this system, but the wording on the Maxim website for the DS2413 left me absolutely cold. But now that I understand the basic aim, I'll go back and see if I can understand enough of the phraseology to see if I could use it too.
Rob
Your proposal lacked a Y rated dielectric or any approvals. Converting
50Hz to dc is fairly trivial, if necessary.
Are you saying that the conductor's insulation is untrustworthy?
4N25 is ancient and has poor gain by today's standards. I would look at something like TLP523 instead.
JEDEC covers case styles, and in some cases pinouts for common parts.
Looks interesting, but I've never used it.
I'm saying its required by law to be a class Y dielectric
NT
Coming back to common sense, I'd feel happier with applying a detector sleeve over inulation that iteself meets the voltage requirement, rather than directly connecting a class Y capacitor. It's not the capacitor that's the risk, but engineering out other problems like, say, condensation getting on the PCB where the capacitor is mounted.
If you were manufacturing these, you'd test the prototype and probably apply coatings or potting to avoid this problem.
For a DIY solution, I'd much rather have a non contact solution as proposed, or use a relay in a box with space.screw terminals where I can visually confirm the isolation.
Saying the cap must be class Y is one thing, but it's meaningless unless you can verify the installation of it.
HomeOwnersHub website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.