Using an Optical Mouse instead of a Rotary Encoder

If you have to focus on the mat and recognize the pattern when it moves you need a lot more computer than PicAX. A small ARM running Linux at 100 MHz will do. If you stick to simple ANSII C.

If you have line on the mat you can sense you can do it with almost any automotive CPU that has the high resolution timer to measure the time between marks. You probably don't need a camera or a lens just a light sensor or two set to go high or low when the line passes it. You do that with lighting and filters and save a bunch of money.

Explosion proof enclosures are too hard. Just make sure the window seals and can be replaced. That means no Silicone glue. It won't stick to its self the second time.

Gordon

Thanks Gordon, I thought that may be the case, that the clever stuff is done by the computer software. Unfortunately the pattern on the sarking changes with different batches. So maybe it's back to the Rotary Encoder. Tony

I'm not at all sure which technology is used by optical mice to detect the motion, but I strongly suspect that any data processing is done within the mouse, so the output of the mouse is just ∆x and ∆y data. This must be the case in order for them to be interchangeable with with rotary encoder mice, especially since they are hot swappable.

I suspect that some investigation into USB standards might reveal the data format used. I can't believe that much computing power is assigned to the mouse input. Certainly not pattern recognition!

Tony

Over the years programing computers for an Ag Engineering Department in a University I have seen people work on a lot of problems like these. You have the luxury of faster computers, code in the can and inexpensive IDEs. That make things easier but the basic problems solving is still the same. It isn't simple or easy to covert real world problems to pulsed you can count that are the same every time. Getting it right in hardware is the important part so you don't have to patch it up in software later.

If you are planning to work in this field now I would take a hard looks at using ARM computers and gcc cpp as a compiler as the compiler is portable to all platforms and the ARM Chip starts at $3.00 bucks and will make a laptop that is not too far off the top offerings form the majors. You can buy ready made boards for 29.95 up www.newmicros.com. Working with family of CPUs and one compiler that works on everything is a real time saver. I would try to make a deal with my employer to retain ownership to the basic libraries I wrote giving them lifetime license and source code. So I could use them over and over and over building on them over time. I always had that understanding with my contract customers by pointing out that if they didn't want to do that they would have to pay me to write them over for them. It wasn't a problem at the university under my employment contract that was written for a different kind of job than I had.

There are many ways to use optics to see patterns by using slits, screens, etc. The wider your knowledge of optics the more ways you can see to solve this problems.

Get some samples of the material and glue it on a belt or drum. Run it by various sensors and watch the output on a recoding oscilloscope, computer or both and count the times it triggers as each kind goes by. For a mouse just power it up and count the pulses it puts out.

Anything that you can predict the number of pules per inch on before it goes by has potential to be used. If you already have a method that measures the material as it goes by get that signal. Once your have a sensor or two or three to test combine it with the data from your optical sensor for a month or two and see how good your sensor is before putting it into operation.

Problems like this may take 2 or 3 years to work out or be solved in week. There is no way to guess unless I am there looking at the line you run and have some sensors to test. There may be a lot of problems that have to be worked out as you go or it could be a snap. A great deal depends on the material and how clever you are. And probably most important how quickly you can realize your on the wrong track or if your on the right one and need to stay with a year or two to see it though. Committing a year or two to hopeless cause or one you can't pull off is not good for you professionally or for your mental health. Pulling one out after a year, month or week of hard work that everyone else things is impossible is great for your future and self esteem. In fact if it is the right problem at the right time it can get you name recognition in a major company almost over night if you come up with solution to something that has been a thorn in their side long enough. It can also make a lot of people mad at you as well.

So if you get lucky and a mouse works mark it up to good fortune and hope you don't have to pay it back on the next project.

Gordon

A pulse could be sent to a count down counter that opens the circuit to your feed motor.

Using a hall effect sensor you can create a pulse for measurement:

The gear can be connected to a shaft that has a toothed roller on the other end of it that stays in contact with the material to be cut. There would probably also have to be a corresponding roller on the other side of the material that is spring loaded to provide tension to prevent slippage. I would strongly recommend the shear as opposed to a saw blade as well. This setup should be relatively inexpensive and accurate.

Regards,

Keywalker

Hi Tonymech,

About a year ago I worked with a mechanical engineer on coming up with a solution to a similar problem. We also needed a way to measure the length of material coming out of a machine in a a class 2, div 2 hazardous environment. In our case, the material is coming out of a hydraulic press, so there is nowhere that an encoder could be used.

The ME that I worked with looked into ways to use an optical mouse as you described, but was unsucessful. However, in his research he found this company: http://www.betalasermike.com/ which manufactures and sells a laser sensor that can be used to track length and speed of a moving material with high accuracy.

Thank you EElectrician, I'm following up on that, they have agents in Australia. Tony

Another option that would allow you to use a standard encoder in a hazardous environment would be to purchase an intrinsically safe barrier. These devices are commonly used in refinaries and other places classed as hazardous environments.

An intrisically safe barrier uses solid state circuits to limit the amount of current that can be present in a circuit to a few milliamps. This means that even in the event of an electrical short caused by damage to the device or wiring the current present would be so little and for such short duration that it could not cause ignition of the gas.

Here is a page from Pepperl Fuch's Engineering Guide:

Hello there, I am working on a similar project and o wonder if any one can help me with it, the project is based on using an USB optical mouse to measure the speed of the media coming into a printer that makes labels.

ii have got an optical mouse and i made sure that it works on the media that i have, then i connected the USB mouse to the USB analyzer and i got data that by trying many times i got to know which group represent the change of x direction and which one is for the y direction. that is all that i know so far, it seems that i need a software or an application that can translate my data to a form that represent the speed. can any one help me with that please

thanks a lot