PLC vs microcontroller

I know the difference between PLC and microcontroller. MY question is how Et200s, S7-300, s7-1500

and S7-400 stands compare to some microcontroller. For an example Atmega16 can work on a maximum frequency of 16MHz. Whats the CPU speed of the mentioned CPUs.



Why do I ask this? Well we have need(my company) to do something what normal logic says to use microcontrolell but since we are working mostly with PLCs andd in order to save time/money I'm actually wondering can we do same job with PLC since we are handling with PLC at the moment and certainly not with microcontrollers.



PLC/microcontroller would need to measure AC current and regulate electrical unit according that measurement.

I think that the main difference could be the regulations and standards for each solution. Many industries (to not mention all of them) wouldn't be accepting a non standard solution like a PLC can give you.
On the other hand, if you are using PLC's and you want to replace a "micro controller" with a PLC, I can't think about any down side.

https://cache.industry.siemens.com/dl/files/645/1117645/att_23136/v1/423op_e.pdf
0.03us seems to be the fastest execution time of any instruction for this particular controller, so 33MHz clock speed minimum.

Unlike a uC though, it has a whole operating system to process though, including comms etc.

S7-1518-4 seems to have a 1ns instruction execution time, so 1GHz+ CPU?
https://w3.siemens.com/mcms/program...0/cpu/Documents/s7-1500_techn_data_cpu_en.pdf

it is not just execution speed though, you have to look at response time to a change in input.

ADC conversion time?
Input scan rate?

Also reference this lengthy tome on calculating the time it takes from an input turning high to an output being energised.
http://www.beckhoff.com/english.asp?highlights/xfc/ethercat-plc-12.5-microseconds.htm

As a reference, Beckhoff brag about their 12.5us ScanTime as one of the best in the PLC industry.
http://www.beckhoff.com/english.asp?highlights/xfc/ethercat-plc-12.5-microseconds.htm

I hate when people mention Bubba and Otis troubleshooting it at 3AM... but it would apply in this case. Also, a microcontroller based system is possibly not easy to remote troubleshoot or reprogram from home.

Then there's all the good stuff that is already defined and done for you to configure through a nice IDE. What you would normally configure using a tickbox in a PLC IDE, is usually a lot of time for a programmer to develop.

Many time you can find a purpose built industrial solution for tasks that are not well suited for PLCs to handle directly. The main one that comes to mind is a closed loop hydraulic motion controller like the Delta RMC or Rexroth HMC.

Others could be strain gage conditioners, special analog converters.

If you can let us know exactly what you are trying to do, perhaps someone can point you to an industrial device that can work for you.

AustralIan said:

https://cache.industry.siemens.com/dl/files/645/1117645/att_23136/v1/423op_e.pdf
0.03us seems to be the fastest execution time of any instruction for this particular controller, so 33MHz clock speed minimum.

Unlike a uC though, it has a whole operating system to process though, including comms etc.

S7-1518-4 seems to have a 1ns instruction execution time, so 1GHz+ CPU?
https://w3.siemens.com/mcms/program...0/cpu/Documents/s7-1500_techn_data_cpu_en.pdf

it is not just execution speed though, you have to look at response time to a change in input.

ADC conversion time?
Input scan rate?

Also reference this lengthy tome on calculating the time it takes from an input turning high to an output being energised.
http://www.beckhoff.com/english.asp?highlights/xfc/ethercat-plc-12.5-microseconds.htm

As a reference, Beckhoff brag about their 12.5us ScanTime as one of the best in the PLC industry.
http://www.beckhoff.com/english.asp?highlights/xfc/ethercat-plc-12.5-microseconds.htm

Click to expand...

Thx will take a look so I will get deeper picture about all i have to be concerned.

ndzied1 said:

Many time you can find a purpose built industrial solution for tasks that are not well suited for PLCs to handle directly. The main one that comes to mind is a closed loop hydraulic motion controller like the Delta RMC or Rexroth HMC.

Others could be strain gage conditioners, special analog converters.

If you can let us know exactly what you are trying to do, perhaps someone can point you to an industrial device that can work for you.

Click to expand...

I need to control generator voltage. Generator voltage frequency is 50Hz in stabile grid/ power network. So I would say 20ms would be minimum.
But since AVR (automatic voltage regulator) needs to answer on some power grid peaks/disturbance it should be faster than 20ms I guess. For how much I dont know. But lets stay at 20ms.
Problem with PLC is how to measure governor unit, in this case generator current/voltage.
In a case I'm using Siemens measuring device SENTRON PAC4200 just for measurement and then via PROFINET/PROFIBUS to PLC, I think I'm over 20ms already
I'm also not sure whats sampling time of SENRON PAC4200, but Siemens 7UM62 generator protection device(taking measurements from this device is option too) it's about 30-40ms If I'm not wrong.

naturephoenix said:

I need to control generator voltage. Generator voltage frequency is 50Hz in stabile grid/ power network. So I would say 20ms would be minimum.
But since AVR (automatic voltage regulator) needs to answer on some power grid peaks/disturbance it should be faster than 20ms I guess. For how much I dont know. But lets stay at 20ms.
Problem with PLC is how to measure governor unit, in this case generator current/voltage.
In a case I'm using Siemens measuring device SENTRON PAC4200 just for measurement and then via PROFINET/PROFIBUS to PLC, I think I'm over 20ms already
I'm also not sure whats sampling time of SENRON PAC4200, but Siemens 7UM62 generator protection device(taking measurements from this device is option too) it's about 30-40ms If I'm not wrong.

Click to expand...

Nyquist says that you need twice the frequency of the signal to sample it... however, if you are planning on doing something with it, 10 times is a much better approach as you need to sample and react.

If I understand correctly, you are building an AVR? Or something to control an AVR?
If you're looking at creating an AVR, I think you'd be better off looking into something a bit faster than a regular microcontroller due to the wave form sampling, plus engine speed, comms, etc...

cardosocea said:

Nyquist says that you need twice the frequency of the signal to sample it... however, if you are planning on doing something with it, 10 times is a much better approach as you need to sample and react.

If I understand correctly, you are building an AVR? Or something to control an AVR?
If you're looking at creating an AVR, I think you'd be better off looking into something a bit faster than a regular microcontroller due to the wave form sampling, plus engine speed, comms, etc...

Click to expand...

creating an AVR is what I'm looking for.
At the moment I'm using PLC to enter references to AVR but
my goat would be to do that in PLC if it is possible.

I dont need AVR to do synchronization for me, I need it just to regulate voltage before synchronization and reactive power/PF after sychronzation(on the gridd)