the order in which some switches are turned on

Hi all,

I have an array of N switches . Initially all are OFF.

Somebody turns them ON in some order. It is possible that more switches are turned ON in the same moment.

I need a device which shows me the order in which the switches were turned ON. For instance the device should give me: 4,3,1,5,2 (this is the order in which the switches were turned ON).

The way in which the output is shown in not important. It must be simple to read (by a human, computer, etc).

It is important that the device is able to handle the turned ON (in the same moment) of the multiple switches.

thanks, Laura

Reply to
laura
Loading thread data ...

If I have read your post correctly - there are a large number of possible solutions.

1) N "clocks", one of each attached to the output from the switches.

These could be simple counters with displays that start from zero each time they are powered up. The operator/computer has a "freeze" switch which simultaneously stops all the clocks.

The numbers shown on the displays show the order that the switches were operated - highest first -- lowest last. The "freeze" button makes them easier to read.

{They could, literally, be household clocks with hours, minutes and seconds hands. The "freeze" switch could be a multi-pole switch that disconnects all of them, simultaneously. However, they would need to be manually put back to zero after each "event").

2) The output from each switch is connected to an opto-isolator, which in turn is connected to one parallel port input line of a computer. The computer reads the ports then loops and reads again. A small software routine gives the output you want.

3) As (2) but using a PLC and ladder logic.

4) As (2) but using a PIC and software.

5) As (2) but using discrete logic chips.

The "simultaneous" operation of several switches is not a problem.

There is no such thing, for a start. Using clocks that can measure to a millionth of a second, the chance of two mechanical switches operating within the same microsecond is too small to worry about.

However, the system can be made to scan the switches sequentially. Even if they did go "on" simultaneously, the system will always show a sequence.

Software, of course, can be made to show a "tie" - when two or more switches have changed in the same sample interval.

-- Sue

Reply to
Palindrome

Are they in parallel or series. if in series you could have binary weighted resistance values connected across the switches so that when they are all off you would be measuring the combined resistance of them all in series...as the switches are closed in any combination they would render a unique value of resistance depending which switches have been closed which is immediately convertable to identify each switch closed.

Reply to
Danny

Welcome to the world of SCADA.

An RTU (Remote Terminal Unit)/PLC (Programmable Logic Controller) can be used to watch auxillary contacts on each switch and log thier state in real time (record that state change and time stamp when it happened). The information is usually sent via serial communications to a data logger or more usually a control system.

Some (most?) RTUs will keep a limited event history internally which can be read using a PC but generally they a not very user freindly.

The down site to this is its expensive and would require dual pole switches.

Real world SCADA usually includes being able to operate the switches remotley sometimes over large distances. Serial comms adds complex protocols (languages) to get the infomration from the site to the control system.

I do not know much about PLCs but compainies like Schnider, Remsdaq, Microsol, ABB and many others manufacture RTUs. Look for the number of Digital Inputs (DO) you need; a seperate DI is required for each switch. Check if the RTU has a readable interal event log if you don't want to get into communications and a PC based data logger/control system.

Reply to
GraemeM

Admittedly I don't know a lot about the details of what the OP is trying to do, but this reply sounds like making a mountain out of a mole hill. Wouldn't a $2 PIC or AVR microcontroller do the trick? Optoisolators can be used to monitor the ouput of the switches if they are not logic level, and the output can be displayed in any number of ways. LCD display, 7 segment LEDs, VFD, serial output to a PC or terminal, anything. Sounds like the whole project could be accomplished for under

20 bucks by anyone with a decent junkbox.
Reply to
James Sweet

My experience is SCADA and I agree its over the top at first glance. I don't know the capabilities of the simpler devices, can they store a time stamped event log? This functionality might not be possible in a PIC or similar device.

The last spec is also interesting, being able to detect simultaneous switching, something an full blown RTU can do across all its inputs but the spec does not state the number of monitored switches. Same question, I suspect the input capabilities on a cheaper device will be much more limited.

Reply to
GraemeM

You should play with some of the microcontrollers out now, with speeds in the tens of MHz the capabilities are astonishing. You can certainly store a time stamped log, many MCUs contain an onboard RTC, and cheap accurate external parts are available. They have onboard flash, but can be easily interfaced to off the shelf CF cards making gigabytes of storage a real possibility. A bank of shift registers could enable monitoring hundreds of switches. Most of these parts are just a few dollars even in single quantities.

If this were an industrial situation where lives or limbs depend on a thoroughly tested, certified, rock solid system then yes, I'd look into a commercial solution, but from the sound of the request something orders of magnitude cheaper would work just as well. More information would be helpful, define simultaneous switching, does it matter if two events a microsecond apart are detected as simultaneous? For some applications yes, for many others, no. You wouldn't buy a $200K CNC lathe capable of .0001" tolerance to turn table legs when a $600 unit with a .1" tolerance would do the job just fine. One has to know all the requirements before a solution can be chosen.

Reply to
James Sweet

GraemeM wrote:

This question was spewed all over Usenet by a lazy, cheating student.

formatting link
At this point, the OP already handed in other people's work as her own.

Reply to
JeffM

PolyTech Forum 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.