0
\$\begingroup\$

I've noticed many motion control cards have optically-isolated inputs for limit switches? I can understand why you may do this with a signal which travels between two systems with separate power supplies, but what is the purpose when a shared power source is used?

Assume a control card with a 12v output for a NC limit switch. A voltage divider (using two resistors) lowers the voltage to an appropriate level determined by the MCU. A pull down resistor is used to tie the input pin to ground when the switch is open. In such a scenario, why would optically-isolating the input be a good idea? I can only think that a spike in the 12v output would produce an unacceptably high output after the voltage divider.

\$\endgroup\$
1
\$\begingroup\$

Motion control cards generally live in industrial environments, and such things as damage to cables and short circuits are not exactly uncommon.

Also, immunity to such things as arc welding (Including arc welding where there is a common impedance between the welding supply and the controller), plasma cutting, induction heating and all the other nasty stuff that sometimes happens in a factory.

When designing such a card, making it usable in the nasty, two systems (one of which is a welding robot) case is a small BOM increase for a lot fewer customer returns.

Isolated IO is very much a desirable feature in controller cards as it improves the systems immunity to muppet installers and also the immunity to high energy faults.

\$\endgroup\$
  • \$\begingroup\$ Couldn't I just get away with using a capacitor? \$\endgroup\$ – M-R May 9 '18 at 17:45
  • \$\begingroup\$ @M-R That don't help with DC shifts, but adding one is usually a good plan (Along with ferrites, a transorb of some sort and a fusible resistor). Doing a set of industrial muppet safe inputs and outputs to a modest micro or FPGA can easily double the BOM compared to some random dev board. \$\endgroup\$ – Dan Mills May 9 '18 at 19:42
  • \$\begingroup\$ Doesn't a capacitor smooth out any spikes in voltage? \$\endgroup\$ – M-R May 9 '18 at 22:10
0
\$\begingroup\$

If you are sharing the same DC power supply you can improve immunity by using a differential amplifier on the detector. But the main advantage of using an optical limit switch is with a small amount of hysteresis , you can achieve very precise control less than a millimetre of position without any mechanical switch bounce. Depending on your datum position accuracy needs whether it’s a millimetre or a micron optical switches tend to be better but mechanical switches chosen carefully can be made to work as well. Consider that most emotional and consumer service systems use optical for this reason such as DVDs Floppies FDD, and stepper motor CNC’s

\$\endgroup\$
0
\$\begingroup\$

Any awful lot of inexpensive Chinese Motion Control cards have opto-isolators installed in the such a way that they provide exactly zero benefit.

I'm talking about inexpensive CNC motion controllers originally intended to be connected to a PC's printer port. These controllers insert opto isolators into each signal line ** BUT ** use a common power supply on both sides of the isolation barrier.

The sad thing about those controllers is that it is significantly difficult to separate both sides of the isolation barrier. That is - both ground and Vcc hip-hop between both sides of the optos.

The general principle of using an isolation barrier is between the PC and machine being controlled is considered to be industry standard. But this requires true galvanic isolation.

Any controller that doesn't allow true isolation between the machine and the PC isn't up to snuff, even if it contains opto isolators on the signal lines.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.