How to detect noisy electrical environments?

T Chap

In a building where there are tons of 270V cable runs, ballasts, AC handling motors, etc. is there an instrument that is hand held designed to walk around and detect noise that would be potential cause for low voltage problems. For example, a Propeller has sensors and button inputs over 20' cable runs, with lots of false signals even though there are caps attempting to filter noise. I need something like one of those things you see on TV that they are using to detect ghosts, only I want to 'see' the noise or try to pin point cables in walls, motors, etc that are offenders.


As a thought for a crude device, what about a Prop with a wire attached that is not shielded, not filtered with caps, with a beeper. If the cable detects changes, make a beep. This would of course require a voltage higher than the trip point on the Prop, so an alternative for a more sensitive method might be an ADC, with user defined trip point going positive?

skylight

With that sort of equipment around the environment then a detector of whatever type would sense noise but the relevance of this reading would be if the sensitive piece of equipment was picking this up, it would be a case of making sure the sensitive stuff was well shielded.ie isolated supplies, shielded enclosures,cables etc

T Chap

Or a case of positioning able runs away from noisy areas, that is more of the idea if shielding fails. In some cases, shielding only goes so far, especially in a case where there is a capacitive touch sensor that is exposed and cannot be enclosed.

David B

Try Bob Pease's suggestion of walking around with an AM transistor radio, tuned to a quiet spot between stations. It can pick up quite a lot of radio noise.

Phil Pilgrim (PhiPi)

If you possibly can, use differential signaling (e.g. RS485) for your cable runs, properly terminated at both ends, to reduce noise pickup. That will buy you a whole lot more than caps and shielding.

-Phil

skylight

Attention to cable runs is also important, at work where it's sometimes hard to avoid sensitive lines such as microphone cables next to dimmer cables (very noisy) then a trick we use is not to run the cables parallel to each other and where they have to cross to make sure they cross at 90 degrees, this tends to minimise any interference which the audio people would pick up.

Phil Pilgrim (PhiPi)

One other caveat: if you use shielded cable, ground it at one end only. This will prevent the shielding from carrying current, which would nullify its utility as a shield.

-Phil

GordonMcComb

Getting back to the original question, what you describe is an EM meter. You can build one (it'll be uncalibrated, but it's cheap) or buy one in a nice package. Cost starts at about $100, upwards of $4K for the complete 3M unit. They look impressive in operation, though in my (non-paranormal investigation) experience, the results aren't always conclusive. You know there's a strong EM field coming from somewhere, but it's not a simple conclusion that it's actually causing any problems.

Anyway, do a Web search for EM meter and you'll find plenty, along with circuits to build your own. The "natural" or "background" ones are aimed at the ghost hunters. You don't really need or want a super-sensitive one.

-- Gordon

bill190

The problem can also be all the zillions of "wireless" devices in the area. You may be able to take your gizmo out to the parking lot and have the same problem!

Here is a story about people's car keyless remotes not working due to this electrical noise in the air...
http://www.washingtonpost.com/wp-dyn/articles/A28217-2004Jul4.html

Best to use shielded wire grounded at one end only as suggested above.

T Chap

Thanks for the input. Standard grounding for my wiring is always to connect the shield to GND on one end only. Some applications make RS485 impossible at the moment without redesign. In the more important case of touch sensitive panels in use(Qprox/Atmel device), the panels must be exposed to get used, and have large (PCB) sensor areas for capacitive touch use(2"x2" for example). They cannot be shielded. The idea is to determine if an area is noisier than another to avoid false reads. In terms of cables, that is easily solved with shielded CAT5. I would like to provide an inexpensive (loaner) device for clients to use in these cases of problematic rooms.