This article started with one of my occasional visits down to the London meeting (at which I purchased an RFID starter kit). Well this article is one of the article options I had now that I have a starter kit (the other should be along next issue).
Having built, tested and boxed the starter kit I was ready to build a simple computer program to allow the tags to be read and some times to be measured. The program works by opening the serial port and waiting for a tag read to happen. When a tag is read the checksum is checked and a lookup is done to see if it is tag 1 or tag 2. You'll notice that my tag numbers appear to be missing the last 2 digits, these are the checksum and as such don't make up the number of the tag (so I don't include them in the number of the tag).
The textbox which acts as a log of events will gain a line for each relevant "stimulus" to the program and an indented line with calculated information. For example the line "Read FE4000F7BE True Tag 1" indicates that the tag with number FE4000F7BE was read, the checksum was good and it was identified as tag 1. After a complete pair has been seen (that's a Tag 1 and a Tag 2 in the correct order) the time between seeing both tags is displayed. As the tag separation is known this allows the speed of the coach passing the reader to be calculated. There is the potential of a timing error occurring as when a tag is read it is sent to the computer where it is put into a buffer and then retrieved by the application. This will be minimised by having as little as possible running in the background. Another potential error comes from the possibility of the tags emitting their numbers whilst in different positions relative to the reader, this I have no control over.
The experiment consists of fixing tag 1 (at the front) and tag 2 (at the rear) of a coach and seeing how close they can be at different speeds and still be read reliably. Reliably is taken as 5 "complete sets" for 5 laps of the track. The box with the reader will be placed beside the track, as you may have notice in the photo the reader sits in a cut-out made in the side of the box. The experiment will be done with the tags in three different orientations (along the coach, across the coach and up/down). The tags will pass as close to the centreline of the reader as possible, the gap between coach and reader is about 1 cm.
The graph shows the results for each of the three orientations plus the theoretical performance, taken from what appears to be the minimum time between reads of different tags (assuming the radio signals are only good as the tag is beside the reader). They would appear to show that having the tag up/down in my test (assuming the reader is under the track this would be across the coach/wagon) provides the best results (and closest to theoretical). The labels on the graph have been "rotated" so that they are correct for a reader below the track (see diagram).
Now for some maths:
1) The equation of the line is y = 2.3116x + 0.0311
2) We can assume that the y intercept element of this is a result of the errors mentioned above and simplify this to y = 2.3116x
3) Putting in some more useful names we get: Separation = Z / Max Speed
and Max Speed = Z / Separation.
Where Separation is measured in metres and Max Speed is in metres per second.
Adding in scaling gives us the following table and conversion to mph gives the following table.
|12"||0.02||Real metres per second|
|12"||Real miles per hour|
|4mm||Scale miles per hour|
|2mm||Scale miles per hour|