I have an op-amp where the non-inverting input is connected to a pot and a free floating wire connected to the other input. If I touch the wire or even just get close to it without touching it, the op-amp responses as if I were connected to the ground and had current. Why does this happen? Is it because I do produce some electricity and the op-amp is picking up on it?

 

I have an op-amp where the non-inverting input is connected to a pot and a free floating wire connected to the other input. If I touch the wire or even just get close to it without touching it, the op-amp responses as if I were connected to the ground and had current. Why does this happen? Is it because I do produce some electricity and the op-amp is picking up on it?

Your body is acting like an antenna and picking up power line emissions.

 

Your body is acting like an antenna and picking up power line emissions.

I was hoping to use this effect to make a contact detector, however, I wanted the project to detect when one large metal object was in contact with another (and I'm unable to run electricity through the moving object.) It seems like this won't work though.

 

Something like this inexpensive capacitive sensor may work for you.

 

Something like this inexpensive capacitive sensor may work for you.

I want to make a sensor for an old CNC machine to warn me if a cutter broke. My thought was that I'd set the sensor to the metal CNC table, and the capacitance would change when the cutter is in contact with the piece. If the cutter broke, there would be a long period of time when the two don't make contact. Do you think that'd work, using a capacitive sensor?

 

Do you think that'd work, using a capacitive sensor?

I have doubts that the capacitance would change significantly if the cutter broke.

Here's another thought.
Could you perhaps sense the vibration or noise of the cutter making contact with the work?
If the vibration/noise suddenly stops than that's an indication that the cutter has broken.

 

Here's another thought.
Could you perhaps sense the vibration or noise of the cutter making contact with the work?
If the vibration/noise suddenly stops than that's an indication that the cutter has broken.

That would need very sophisticated software to filter out all of the other vibrations going on. The spindle moves in the Z axis with a lead counterweight like an elevator while the table is moving back and forth in the XY axis'. It could work when doing heavy cutting, but there are times when you're cutting carbon using a .01 of an inch diameter cutter. I'd be surprised if a sensor could detect that vibration.

Starting to feel like a pipe dream.

 

That would need very sophisticated software to filter out all of the other vibrations going on. The spindle moves in the Z axis with a lead counterweight like an elevator while the table is moving back and forth in the XY axis'. It could work when doing heavy cutting, but there are times when you're cutting carbon using a .01 of an inch diameter cutter. I'd be surprised if a sensor could detect that vibration.

Starting to feel like a pipe dream.

Another option would be an optical sensor working through optical fibers. The fibers would be positioned across the cutter, one fiber would transmit light, the other one would go to the sensor. This of course, requires additional space and hardware very close to the cutting tool.
Question, how is the material being cut held to the the cutting table?

 

This op-amp thing you describe is driven by what we used to call, "finger hum". Finger hum was used as a cheap signal generator for testing audio amplifiers 60 years ago. Many old timers have used j-fets to detect touch because all you need is a high impedance input and an antenna, commonly referred to as a "human". The op-amp has the high input impedance, but this trick can be done much more simply. You can adjust the sensitivity by adding a resistor from the input to ground, usually in the range of 100K to a few million ohms.

This might be useless for your CNC request, but at least you know what's going on, and it might come in handy in the future.

 

Question, how is the material being cut held to the the cutting table?

The table has 1/2 threaded holes. Usually, we bolt a vise or two onto the table, and put the workpiece in the vice. Other times, set the material on the table directly and pinch it down.

 

The table has 1/2 threaded holes. Usually, we bolt a vise or two onto the table, and put the workpiece in the vice. Other times, set the material on the table directly and pinch it down.

How thick is the material you usually cut? Is it normally non-conductive?

 

The optical sensor is also a good idea, but a little problematic when it comes to placement.

I think what I'll have to do, if I don't scrap the idea entirely, is just make a button, and program the machine to push the button at the end of a program before picking up the next cutter. (Not an original idea.) If the cutter is broken, it won't hit the button. I don't like this idea, though, because I don't have an easy way of judging at what time the machine should hit the button. So the program would really be telling me that the cutter isn't broken, instead of when it is...

 

How thick is the material you usually cut? Is it normally non-conductive?

It is almost always contuctive. We cut steel and carbon (to be used as electrodes in die sinker electrical discharge machines.) Material thickness can vary. Usually no more than a couple inches.

 

Conductive* (The carbon is conductive by the way.)

 

It is almost always contuctive. We cut steel and carbon (to be used as electrodes in die sinker electrical discharge machines.) Material thickness can vary. Usually no more than a couple inches.

I think I might have the answer ... last question ... what range of tool diameters do you use, and what's the rated HP of the machine's spindle, or router? Does the spindle run on ordinary 120/220VAC, or is it DC?

 

.750 - .005 inch diameter cutters. I want to say if runs on 220AC. The HP is written on the machine but I can't remember. 5HP?

 

After the cut, but before tool change(or at anytime really) .......is it possible to position the tool against a test distance difference or standard, to confirm tool length? You could even roll circumference of tool in the same way, to confirm tool chips.

 

Ok, here's my take. And there are two possibilities.

1. Monitor the amount of current flowing through the spindle. When it cuts, it consumes more current, and that can be detected with a proper sensor. The problem is that as the tool gets smaller, so does the current needed to cut the the material. So an extremely well-tuned sensor has to be used. If the tool breaks, then the current going through the spindle will not change.
2. Monitor the sound that the tool makes while cutting the material. For this, you could use a circuit with a tunable audio filter (like an equalizer) that could be adjusted to monitor the specific frequency, or characteristic sound, that the tool makes while it's performing the cut.

Option #2 is the one I like best when cutting thin materials with very small tools. Although it would need to be adjusted to tune it to the specific tool and material being cut every time you use the machine. But option #1 is better when cutting thicker materials with higher diameter tools, since it's a lot easier to design and implement.

 

Another thought... one could build a circuit tuned to the characteristic sound that the spindle makes when it's not cutting... so you will know if it is indeed cutting if it starts making any other sound than that. That would make it a lot simpler.
What remains to be answered is if the spindle makes (more or less) the same sound when its spinning freely with a small or a large cutting tool.