RPM counter that stops machine after 10 cycles

[privas]

rpm counter pic

I have a machine that twist metal square tubes. It makes a spiral turns into the tubes to make them more decorative. Problem is that I have to count the turns myself and then manually turn off the machine. I would like to make or buy a device that would turn off the machine after 10 turns. The machine completes 14 turns in about 90 seconds. The machine operates on 110 volts and has a basic on/off switch.

Any ideas or suggestions?

 

[VVV]

rpm counter schematic

How about a sensor on the moving part of the machine and a counter that turns off the machine (using a relay, for instance), when the required # of pulses has been received from the sensor? The sensor could output more than 1 pulse per/ revolution, increasing accuracy.
If you want to get fancy, make it programmable, to allow you to vary the # of twists in the metal. In fact, that would be advisable, because at some point you may want to twist shorter pieces, which require fewer turns, for the same appearance.

 

[privas]

rpm counter uses

Okay, I like the idea. Now how do I make this device.

1) what type of sensor
2) how do I wire it to the relay
3) how does it keep count
4) how can I vary the number off turns before it turns off the machine
5) can you make a schematic, because I can't. I solder parts but I can't design anythng.

 

[VVV]

The only real problem here is the sensor. Everything else is fairly easy to do.

You can use a Hall effect sensor or a pickup coil. Both require small magnets installed somehow on the moving part of the machine. These magnets activate the sensors. You can install more than one magnet. I would try to glue them using epoxy.
The Hall sensors would be my preferred choice. You can re-use the sensors you find in the 5 ¼ inch floppy drives. Can you get such an oldie?
The sensor has to be really close to the magnet (and to the moving part), and installed securely (against vibrations) and protected by some sort of cover, for protection.

Optical sensors are also possible: basically a disk with holes around its edge can be installed on the moving part of the machine and you can use a photointerrupter, from an old printer for example, to "see" those holes. Again, mechanical precision and protection are essential.

If you think it is easier to install a switch somehow and some sort of cam on the moving part, that is also OK. Ideally, the switch should be closed/ open more than once/ revolution.

You have to decide which solution is more feasible, depending on what the actual machine looks like, what the available space it, etc.

Once you have that going, you can use a programmable counter and make it stop the machine after x turns.
If you prefer, you can use a microcontroller. Are you comfortable with PIC's?
Again, this is the easy part, I am not at all concerned about it.

 

[privas]

I like the switch and finger idea better because the work enviroment is dirty and hazardous. So once the finger hits the switch ten times the machine would turn off. The programmable counter would switch the relay from a normally closed position to a normally open position after the set number of turns has been reached. Then it should reset itself or have a reset button. Can you please make a simple sketch so that I can better understand how to wire everything correctly? Please list components needed.

 

[VVV]

OK.
I have to ask again, are you comfortable using a PIC?
Do you have access to a programmer?
This would be the easiest solution by far. Just one PIC12C508 would do (plus a 5V regulator, a relay, a transistor, binary switch and a few resistors/ capacitors).

Another solution would involve the following parts:
one 555
one 4093
one 4013
one 4029

plus the rest of the stuff (12V regulator, relay, transistor, binary switch, resistors/ caps).

Let me know if you can find these components.

 

[Gorgon]

Hi
Another way to do this is to use a timer relay and a stop switch. Since your process is relative slow, about 65 sec, you could just start the process and the timer relay. This is set to a few seconds less than full time. When the timer is out you connect the switch and when activated this stop the process, when activated the 10th time.

The motor is activated and held activated with a relay. The switch and the timer relay is placed in the hold circuit.
The switch is a NC type which opens when activated, and you just short it with the timer relay contacts.

When the timer is out, the activation of the switch will release the hold circuit and the motor stops!

The timer relay is a standard industrial type, or something you build yourself.

In addition to this you need to include some form of emergency stop switch (Big red button type).

TOK

 

[privas]

The timer relay is a good simple idea but it nots precise enough. I would need to create a chart with the different seconds each metal tube requires and also adjust for their different lengths. Adjusting the timer relay precisely would be diffcult.

With the finger/switch counter there would be much more precision.

I know nothing about PIC but I can find all the components you listed.

one 555
one 4093
one 4013
one 4029

plus the rest of the stuff (12V regulator, relay, transistor, binary switch, resistors/ caps).

 

[VVV]

Hi privas,

Here is the schematic I put together for you.
It may seem a little complicated at first, but it contains exactly the parts I mentioned. Let me know what you think, or if you have questions.
The binary switch programs the number of revolutions, from 1 to 15. Avoid setting it to 0, that is not a valid choice.
You can build a sturdy version of a binary switch yourself (I see you are mechanically inclined), or you can just hard-wire the inputs of the counter for 10 turns: connect J4 and J2 to +12V and J1 and J3 to ground.
You can add the switch later.

Note that the start switch MUST be normally closed (otherwise you need another inverter or transistor). If this is a problem, we can change that.

Meanwhile, I will think about a simpler version of this schematic.

You do not need to know that much about the PIC. You can always learn. That solution would simply involve the PIC and possibly the binary switch.
If you can get and program the PIC, that is still a very attractive option.

Regards,
VVV

 

[privas]

Your schematic is awesome. It's amazing you designed it so quick. I'm going to try to make this over the weekend so that I can give it a test run on Monday or Tuesday. I really appreciate your help.

Regards,
Privas

 

[VVV]

You're welcome.

You can actually test just the circuit, once you have built it, no need to actually connect it to the machine until you are sure it works.
The machine-mounted switch should be a snap-action type. There is debouncing built-in (about 20ms), but you cannot just use two wires touching, that would give you false pulses.

Let me know how it goes. If there is a mistake somewhere in the schematic, it was not intentional. But whatever the case, we'll get it to work.

 

[rituparnasaikia]

You can mail me with your requirment. It can be solved with a small embedded system. Can be used a 89c51 with a sensor.
rituparnasaikia@gmail.com