3 Phase motor Speed Controller

I need to make the speed of a 3 Phase motor adjustable using a PLC . I thought that the best way is to use the PWM, but I realized that my PLC type does not support this function.... Is there any other solution?

for starters a PLC cannot run a 3 phase motor directly your solution is to
install a VFD then send the speed reference to the VFD

The cheapest way would be to use a VFD to control the motor speed and use the PLC's analog output to control the VFD.

Shohadawy said:

I need to make the speed of a 3 Phase motor adjustable using a PLC . I thought that the best way is to use the PWM, but I realized that my PLC type does not support this function.... Is there any other solution?

Click to expand...

Here we go again !!

The cheapest way is to use PWM pulsetrain from PLC to control the VFD speed.

Maybe it's us, obviously we didn't convince him last time that VFDs were the best way to go!

See this thread posted by the same OP


http://www.plctalk.net/qanda/showthread.php?t=41742

Analogue 0-10V or 4-20ma out to VSD

Shohadawy said:

I need to make the speed of a 3 Phase motor adjustable using a PLC . I thought that the best way is to use the PWM, but I realized that my PLC type does not support this function.... Is there any other solution?

Click to expand...

If you ever find a PLC capable of doing that and not burning out in the first 5 mins, please do let us know.

Your best bet is a Veriable Frequency Drive. VFD for short.

Hitachi
KEB
Schneider Electric

These are only a few suppliers that can do what you want.

Somehow Peter's last rant thread comes to mind atm.........

The PLC does not have to support PWM function- it may be software emulated.
With transistor output, of course.

I have never used a PWM signal as a drive reference. However, to keep the command bandwidth useable I suspect there is come carrier frequency you need to be above. This carrier frequncy and your scan time will define the speed command resolution you can achieve.

So, Sergei, can you attach some numbers to these frequency and resolution questions?

But as others have said, most plcs are not capable of solving the control variables needed to perform even volts/hertz commutation in the time you have available, regardelss of whether it supports PWM or not. You will need a VSD. And as others have also said, the velocity command interface flavor is pretty much up to you, the drive supplier and the plc supplier.

Keith

kamenges said:

I have never used a PWM signal as a drive reference.

Click to expand...

Then why are you arguing with one who did it on hundreds of commercial projects?

kamenges said:

... to keep the command bandwidth useable I suspect there is come carrier frequency you need to be above. This carrier frequncy and your scan time will define the speed command resolution you can achieve.

Click to expand...

Absolutely correct. Embedded PWM is a high-speed function, as well as the software emulated PWM output must be.

kamenges said:

So, Sergei, can you attach some numbers to these frequency and resolution questions?

Click to expand...

Speed control resolution depends mostly on the PWM resolution and stability.
The responce is anyways faster than typical acceleration and deceleration settings in the VFD itself (0.2 sec minimum, typically 0.5 sec).

kamenges said:

But as others have said, most plcs are not capable of solving the control variables needed to perform even volts/hertz commutation in the time you have available, regardelss of whether it supports PWM or not.

Click to expand...

Who and where said this in this topic?

kamenges said:

And as others have also said, the velocity command interface flavor is pretty much up to you, the drive supplier and the plc supplier.

Click to expand...

Neither combination of VFD with speed reference voltage input and PLC with transistor output excludes speed control by PWM.

Who's arguing? I'm looking for information. I know there are drives on the market that will accept a PWM signal as either a velocity or a torque reference. There are also drive that will accept PWM signals directly to their power sections. Which of these are you suggesting the OP is capable of doing?

The jist of the OP as well as the thread referenced in post 7 seem to indicate the OP wants to directly control the power delivery to an AC motor from a plc using a PWM signal. This infers the last case I listed (direct PWM control of a drive power section, performing electronic commutation). Sergei, if you can do that out of a 'plc' using a personally derived software algorithm with no specialty cards you are using something that most of us wouldn't call a plc. I have no reason to doubt that you can write the algorithm. I have every reason to doubt that a plc can execute that algorithm in a timely manner and get the signals out of standard plc hardware.

So that leaves us with the other two cases, namely a PWM velocity or torque reference. I have never used a drive with this type of interface so I don't know the signal constraints. But since no constraints were given let's start with an illogical extreme and work backward. We'll set our PWM period to 2 seconds (0.5 Hz). This is a completely unreasonable value but since none other was provided we will use this. Assuming a typical plc scan time of 10 msec we have the plc software resolution capability of 200 counts. So we can change the command in 0.5% of full scale steps. That's not very good and that assumes an unreasonable PWM period.

This discounts the drive input, which is designed to accept this stpe of signal. It prpobably doesn't care what the period is; it will decode it correctly no matter what. The period just plays to response.

So how is this done? Numbers and links are always welcome.

Keith

Well, first of all I am talking about AC inverter (a.k.a. VFD), configured for speed reference by analog voltage signal, not "direct PWM control of a drive power section, performing electronic commutation".

The typical conection to PLC is shown on the attached drawing.

The resistor should be equal or just above the drive input resistance (see the drive specs) and the rectifying capacitor converts PWM to voltage.
Starting guideline is 10kOhm/1mcF.

With PWM instruction at 100Hz typically it is possible to control the inverter speed from 3Hz at 1%PWM to 120Hz at 100%PWM. The step value is different, from ~2Hz/% at the lower speeds to ~0.5Hz/% at top speed.

Anyways, control resolution over 100 steps or higher linearity is out of the budget control method scope.

Concerning emulated PWM, its count is to be done not in every PLC scan, but in high speed routine with the output immediate refresh. This may be a timer interrupt or even a regular subroutine being called many times per scan.
With 1ms period and 50 steps it results in 20Hz PWM pulsetrain (may require larger rectifying capacitor).
50 steps resolution is sufficient for many applications, such as HVAC, non-intelligent conveyor, etc.

... Who needs Sallen-Key or 7th order Butterworth filters when you can just use a single resistor and cap. Hahaha...