Monitor fan speed using optic sensors. Advise

[MVidal]

Dear Sirs:

I'm new to DAQ and currently working on an implementation based on FieldPoint.

Basically I have to monitor the speed of 12 fans on a machine. Each fan is about 80cm diameter and runs at no more than 1800 RPM. Looking at its axis, it seems there is no way to place a disc to implement a magnetic sensor.

I guess that my alternative is to place an optic sensor. And I should mention that I don't have much knowledge on this matter.

That is why I'm asking any of your kind help that guides me to choose the appropiate solution.

Here there are the requeriments that the system must consider:

* 1800 RPM max. (It can seem I would need to take 1800/60 = 30 samples per second. Optic sensors can read this frecuen
cy?)
* I think I can place a reflective tape on one of the six fan'wings which by the way are plastic and black.
* The system has a temp of 40oC and some dust can be present. Cleaning can be necessary.
* As I would use a FieldPoint DI-301 to get the signals, I guess I should get PNP sensors.
* I have access to a power supply for 12VDC, 24VDC, 220AC, and others if necessary.
* My major restriction is cost: Each machine has 12 fans to monitor, and there are 24 machines on the field!

I will appreciate any help on this matter. What other considerations should be taken? Which is a good source to find the sensors on the market?

Thanks in advance.

Best regards,

Mauricio Vidal

[Russell]

Hello,

If you must take the route of an IR optical encoder solution, you might look around the web for some examples. I found the following link that you might find helpful.

http://www.embedded.com/design_library/esd/hi/OEG20030617S0025

After the optical transistor, you'll probably want to have a comparator or a digital inverter/buffer in order to ensure sharp edge transitions to the FieldPoint module. (The FP module will need to receive sharp edges) Also, instead of a FP-DI-301 module, I would suggest the FP-QUAD-510 encoder module. This module is made for specifically for such an application. With it you can have it return data in terms of velocity directly. (With the FP-DI-301 this
would be much more complicated and would require a lot of polling on the module...probably more than you can really do.)

I hope this helps!

Russell
Applications Engineer
National Instruments
http://www.ni.com/support

[Buechsenschuetz]

A rotation speed of 1800 rpm will pose no problems for an optical sensor, even if you have more than one signal per turn. However, if you want to use a ready-made unit such as optical displacement sensors used in machine automatization you should check its maximum frequency - sometimes these sensors are speed limited on purpose to avoid any signal glitches due to momentary and unwanted reflections.

The temperature of 40 deg C is rather high but most semiconductors are specified for operation up to 70 deg. C. When using optical sensors, dust can be a problem if there is any chance that it accumulates on either the light source, the reflective area or the sensor area.

For dusty environments magnetic or inductive sensors are a better choice. But when us
ing these you will have to place some ferro-magnetic material on the rotating element, and if this is running at rather high speeds (such as in your case) of course this ferro-magnetic object (a tab or a slotted disk) has to be well balanced (which can be achieved by using a disk with several slots evenly distributed over the circumference, thus giving several signals for one turn). The sensors mentioned are available both as a single component without any additional circuitry (which you would have to design by yourself) and without any mechanical interface other than the solder pins, but inductive and magnetic sensors are also available as ready-to-use units.