MCC & Cars

Posted by Hans Nouwens on 2010-02-28 01:28

Hi Jaap,

you cannot use a hall-sensor directly. The ucci scans all it's sensors by repeatingly switching them on. It senses any return current and figures out which switches are closed. This is done at a very fast tempo. It just looks like it is reading them all at the same time but it's not.

A Hall sensor needs a ground, power and returns a current is there is a magnetic field. It has three wires connected and has some logic processing. If you connect it to the ucci it is "restarting" every time it is beeing scanned by the ucci.
For example the (two wire!?!) MAX9921 provides an 80µs blanking time following Hall sensor power-up or restart http://www.maxim-ic.com/quick_view2.cfm/qv_pk/5410 Is this fast enough? How much time is spend reading each sensor Leon?
The startup time of your devices is not specified.

I myself have done some testing to get an idea about the sensitivity of hall sensors in general. I have not tested this specific model but in general the hall sensors are not that sensitive.
Secondly there are different models of hall-sensors. Picking the right one is easy if you know the subtle difference between a Uni-polar and Bi-polar. A reed-switch works whatever type and indifferent of you connect it.

Add to that the price of a hall sensor compared to a reed-switch and my preference is still the reed-switch.
I use the embedded ones, they are the same as the glass ones except for the plastic casing. For example Conrad Bestnr.: 503711 - 89. Conrad sells them for a unreasonable high price. 0,80 Euro is a normal price if you search at the right places. There is another post about a cheap source for reed-switches.

Here are some links with technical background information about reed-switches
http://www.1op87.nl/index.php/weblinks/42-techniek-achtergrond

I'm curious about your ideas about the hall-sensors. Please let us know!

Hans Nouwens.

Posted by Hans Nouwens on 2010-02-28 01:31

Here is the post about the switches from Farnell.

tiki-view_forum_thread.php?forumId=5&comments_parentId=4603&comments_per_page=1&thread_style=commentStyle_plain

Hans.

Posted by Jaap Knol on 2010-02-28 18:18

Hi Hans,

Thanks for your quick reply. However I feel that I have to explain my set-up of the track, as already done in another forum, in order to enable you to advise me on the possible use of Hall sensors.

Reed-contacts are installed on my test track currently. However in a different way than normally done. This is because of the use of magnetic tape instead of guide wire from Faller.
The magnet on the (floating) steering slider of car takes care for the proper guiding of the car, but will also switch the reed-contact. Unfortunately I'm unable to attach a drawing for clarity to this posting.

I'm now in the process of developing roads on my main train track and therefore I was wondering whether SMD type Hall-sensors could be applied. The reason is that they can be installed at the surface between two pieces of magnetic tape. The main advantage is that the installation can be done from the top side only.

I have made a few measurements with my multimeter on the Hall sensor type TLE 4905 G, which might help you to inform me on what kind of additional hardware would be required to make the sensor suitable for UCCI.
Please remember, I'm a mechanical engineer only, so pardon me if this info is not relevant for you.

Sensor without supply voltage and without magnetic field
Input voltage: 0 V
Output voltage: 0 V
Output resistance: 29 MOhm

Sensor with supply voltage and without magnetic field
Input voltage: 14.5 V =
Output voltage: 0.2 V =
Output resistance: indefinite

Sensor with supply voltage and with magnetic field
Input voltage: 14.5 V =
Output voltage: 0.2 V =
Output resistance: 650 Ohm

Hopefully this information will help you.

Best regards,

Jaap Knol

Posted by Hans Nouwens on 2010-03-01 01:50

Jaap,
I know the floating steering. It's a great system that I still want to try out myself. I searched for a cheap source of the magnetic strip but I have not found it yet.

It looks like there are two challenges with your solution:
1. if you place a hall-sensor between magnetic stripes, how do you prevent it not being activated all the time?
2. next to the electrical characteristics you found out time is of most importance. This is due to the scanning mechanism of the UCCI. Is it fast enough?

There is one user who installed the glass type reed-switches vertically. Just bend one leg and put is right into the magnetic strip. The magnetic field acts as a pre-tension. This results in a very sensitive sensor.

why do you want to install from the top only?

Hans.

Posted by Leon van Perlo on 2010-03-01 18:07

Hello Jaap,

Hall sensors (unipolar) may be used. However ...

Test's I've done myself showed that Hall sensors (the one you mention) are less sensitive than an average reed contact. It may be a difference if you use them in or on top of a magnetic strip.

I don't know about the consistancy between different Hall-sensors though. The TLE4905 has quite some spread in characteristics, so one device may work given a certain field strength, the next may not. There are devices out there with more consistent characteristics and higher sensitivity, but they are expensive.

As Hans pointed out, Hall sensors need power supply. Consumption is about 3,5mA each, so if you have 100 sensors, you draw 350mA, Nothing to worry about, but you cannot simply "steal" it from the signal wires. It requires additional wiring, although it may run through the same flatcable as the signals.

You need to do something in terms of multiplexing, otherwise you need quite some cable between the road and the electronics. This can be done, but it requires additional electronics. Could fit on a PCB like the SWdec, but it needs te be developed, which costs time an money.

I'm willing to think of such a circuit (expecting nobody else does) but not before RAIL2010 in less than 2 weeks. Further we'll need consistency figures for Hall sensors. How exactly shall they be positioned to get a reliable and consinstent activation by the Faller steering magnets? This could be tested by putting some 10 sensors in a test-road, each having a LED to show it being acivated. The LED can simple be put between the sensor output and Vs with a series resistor in between.

Kind regards,
Leon

Posted by Jaap Knol on 2010-03-01 19:54

Hi Hans and Leon,

Many thanks both your replies.

@ Hans

''I know the floating steering. It's a great system that I still want to try out myself. I searched for a cheap source of the magnetic strip but I have not found it yet.

The magnet strips can be obtained at http://www.bau187pkw.de

''It looks like there are two challenges with your solution:
1. if you place a hall-sensor between magnetic stripes, how do you prevent it not being activated all the time?
2. next to the electrical characteristics you found out time is of most importance. This is due to the scanning mechanism of the UCCI. Is it fast enough?''

The first challenge can be eliminated. The Hall sensor will only be activated by a perpendicular magnetic field and that is what the magnet of the steering slider is doing.
I have tested the sensor in different positions and it works perfectly.

The second challenge is the big question and the proper answer can only come from you or Leon.

There is one user who installed the glass type reed-switches vertically. Just bend one leg and put is right into the magnetic strip. The magnetic field acts as a pre-tension. This results in a very sensitive sensor.

The user, who installed reed sensors vertically, is probably me. I mailed Leon last year a sketch, showing how to install reed-switches in combination with a magnetic tape. The same sketch was posted on Beneluxspoor.net in Nov. last year.

why do you want to install from the top only?

Part of the roads will not be accessible anymore from underneath, as they are on top of a ??shadow?? station.
Moreover I hate working underneath the track.

@ Leon

I believe that things are going to move. You made some relevant comments, which of course need to be studied by meself first, before I can react, but I feel that we are going in the right direction.
Will refer back a.s.a.p.

Best regards,

Jaap Knol

Posted by Jaap Knol on 2010-03-08 21:43

Hello Leon,

Hall sensors (unipolar) may be used. However ...
The Hall sensor which I used for testing is of the unipolar type, however the polarity is wrong compared to the magnet of the steering slider and magnetic strip on my test track. Of course I can turn the magnet and the strip, but I still try find the proper sensor, that will match with the existing polarity. Or may bipolar sensors be used.

Test's I've done myself showed that Hall sensors (the one you mention) are less sensitive than an average reed contact. It may be a difference if you use them in or on top of a magnetic strip.
My intention is to cut the strips approx. every 15 cm and put the sensor in between the ends. The top of the sensor will be at the same level as the top of the strip.

As Hans pointed out, Hall sensors need power supply. Consumption is about 3,5mA each, so if you have 100 sensors, you draw 350mA, Nothing to worry about, but you cannot simply "steal" it from the signal wires. It requires additional wiring, although it may run through the same flatcable as the signals.
Why not use an external power source to drive the Hall sensors and connect it to the SWDEC.

You need to do something in terms of multiplexing, otherwise you need quite some cable between the road and the electronics. This can be done, but it requires additional electronics. Could fit on a PCB like the SWdec, but it needs te be developed, which costs time an money.
We may not need additional circuitry to multiplex the signals. I've discussed this matter with my son and he thinks that even an better option may be to include the optocouplers, a voltage clamp and some resistors in a redesigned version of the SWDEC. He is planning to test this on a breadboard.

Best regards,

Jaap Knol

Posted by Fokko Dusseljee on 2013-01-21 22:46

Hello all,
It was a quite time ago that this issues, about using Hall-sensors, came across this forum and I like to know if there were some conclusions whether Hall-sensors could or couldn’t be used of is everybody using SWdec and reed contacts instead?
Was it possible or has someone develop a new idea or some PCB to use Hall-sensors? Was the investigation an success?

Thanks in advance for every feedback on this matter!

Greetings,
Fokko Dusseljee