VSD CONTROL MODES 5

[nightfox1925]

Good Day to all! I am not that familiar with Variable Speed Drives (VSD) and I wish to ask for clarification with this subject.

I have encountered two types of VSD control schemes: OPEN LOOP Vector Control and CLOSED LOOP Vector Control. What are the features and difference between these two types of VSD Control? I appreciate any valuable information and links you can share.

GO PLACIDLY, AMIDST THE NOISE AND HASTE-Desiderata

 

[electricuwe]

Open Loop Vector Control:

Without position feedback --> the field-vector necessary for vector control is calculated from voltages and currents only, good dynamic characteristics at medium and high speeds

Closed loop vector control:

With position feedback: the field vector is calculated with thr rotor position information --> Performance like a DC-drive at any speed including standstill

 

[jraef]

Go visit faq237-1062 by ckicking on this link.

Eng-Tips: Help for your job, not for your homework Read faq731-376

 

[DickDV]

Or.....said another way. A Flux Vector (closed loop) Drive gets its shaft speed information from a direct measurement of the shaft, usually thru a tachometer or relative encoder. Neither of these devices provide position information.

A Sensorless Vector (open loop) Drive gets its shaft speed information from a motor model residing in its memory and created with a Motor ID Run when the drive was commissioned.

At shaft speeds above about 50rpm, the sensorless vector performance in most respects is as good as and in a few instances is better than the flux vector control. But, at speeds below 50rpm, the sensorless vector system falls apart whereas the flux vector system maintains good control right down to stall as mentioned above.

Another difference between the two systems is the nature of the speed error. With the open loop system, the error can be cumulative whereas the encoder version of flux vector tends to center its error around zero so it is not cumulative. This can be a critical difference in a few applications like sheet tension control.

 

[nightfox1925]

Thank you very much for kind assistance. Your inputs have been a lot of help for me.

This is the case at hand. The VSD specification prepared by the original designer required a closed loop speed control for the gas compressor. The VSD vendor is deviating by offering an open loop control on the VSD and recommended us to advise the compressor vendor to include the closed loop control on their compressor PLC.

1) It seems that the speed sensor signals will be transmitted to the PLC then PLC to send command signal to VSD to adjust motor speed. I am still in the process to getting the vendor information on the compressor PLC from the compressor vendor.

2) What if the PLC will send speed control commands to the VSD and the speed sensor signals are transmitted directly to the VSD? In this way, the VSD will be able to get the right actual motor speed information and do the corrections/adjustments required.

Is item 2 a better way of a control scheme? I appreciate your second opinions and experience.

GO PLACIDLY, AMIDST THE NOISE AND HASTE-Desiderata

 

[nightfox1925]

jraef, thank you for the valuable FAQ...it sure pretty gave me a better "lay man's" understanding on the subject matter since I am bit on a very squeezed timetable out here. I am in the process of conducting the required research for this subject matter and quick help from people having the experience would be a lot of help.



GO PLACIDLY, AMIDST THE NOISE AND HASTE-Desiderata

 

[itsmoked]

Why would you need a vector VSD for a compressor? Would not a scalar VSD be less expensive and do the job just as well? You don't usually need positional control for compressors but only speed control.

I would imagine your closing the loop, is only needed to match flow to some need. A relatively slow closed loop. Entirely different from a vector/position closed loop.(very high speed)

02101972; Normally a pressure or in some cases a flow sensor would be read by a PLC then the PLC would send an open loop speed setting to the VSD to alter the compressor speed. Alternatively most VSDs can probably handle the this control without a PLC in the loop. The PLC may provide other aspects of control and coordination.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[jraef]

02101972,
I think you may be confusing the terminologies here. When we are referring to "Open Loop" and "Closed Loop", we are speaking of the internal control method by which the VFD creates torque and / or speed accuracy. This is the issue of "Vector" or "Scalar".

It appears to me that your specifier is speaking of an open or closed CONTROL loop, as in a PID control loop that maintains a regulated output of the motor speed based on a feedback loop from a PROCESS sensor. These are wholly separate issues.

You probably do want Closed Loop Control, as in PID process control, in your gas compressor system, i.e. you want to maintain a specific pressure and want the VFD to modulate the compressor speed to maintain that consistently amidst variable conditions. This would have NOTHING to do with whether or not the VFD uses Open Loop or Closed Loop Vector control in determining its torque or speed accuracy. In fact, it is unlikely that the VFD needs to have Vector control at all, you will likely be operating in a speed range that makes it somewhat irrelevant as itsmoked pointed out.

PID control was traditionally done in specialized controllers that one would wire to the sensors and the VFD speed input. Over the years, these have been integrated into the VFD control architecture because the microprocessor power is available.

What your VFD vendor is saying is that he recommends NOT doing the PID control loop inside of the VFD, do it in the external controller (it appears you have a PLC anyway?), and there is a very valid reason for his recommending this. If anything goes wrong with the VFD and you must swap it out to quickly get back on line, your critical PID loop programming and tuning remains behind in the external controller, saving you the time and trouble of redoing it in the replacement VFD. Conversely, if the PID loop control portion of the VFD goes bad, your entire VFD is down while PC boards are being replaced. If an external loop controller goes down, you still have the option of manual control with the VFD.

Eng-Tips: Help for your job, not for your homework Read faq731-376

 

[DickDV]

That's a really good point about where to locate external control loops, jraef.

 

[waross]

I agree with you completely jraef.
respectfully

 

[nightfox1925]

Thank you jraef for your comments. The VSD SPECIFICATION (not the compressor PLC)done by the original specifier/engineer indicated Closed-Loop Speed Control which was also been part of the tender document wherein the design contractor is trying to comply at present. I may do round up everybody (instrumentation and process people) to discuss this officially since the required item may not be economical and a normal scalar type of VSD might serve the purpose.

GO PLACIDLY, AMIDST THE NOISE AND HASTE-Desiderata

 

[nightfox1925]

I have checked the process PID and observed that the speed signals (4-20mA) from the motor is interfaced directly to the VSD controller and speed feedback goes to the compressor PLC controller (DCS). Then a remote speed control signal is interfaced from the compressor PLC back to the VSD controller. The PLC draws pressure indication/control (PIC) signals from high discharge pressure and tank pressures. Hence, it appears that the PLC is remotely controlling motor speed from pressure control set-points by sending raise/lower command signals to the VSD controller. At the same time, the VSD internal control do its job of ensuring that the speed command given by the PLC is attained by the motor and apply necessary speed corrections where applicable (via internal closed loop control). Is this observation valid and acceptable as far as the VSD operation is concerned?


GO PLACIDLY, AMIDST THE NOISE AND HASTE-Desiderata

 

[itsmoked]

Hi 02101972; Yes! This is what we are expecting.

The only other point I would like to make is that I believe a run-of-the-mill scalar VSD does not even feed back the speed, (even internally), and I see no reason why your application would want or need that feedback. Instead the VSD is fed the 4-20mA signal which essentially commands a specific output frequency by the drive. The motor just strives to reach that speed minus its normal slip. If the speed with the slip is too slow it matters not a bit because the PLC will be up-ing the 4-20mA signal until the flow is at the correct level.

Another way to say this is, you don't care at all what the motor speed actually is! You only care what the flow or pressure(process variable) is. Hence no need for any motor/speed feedback.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[nightfox1925]

Hi Itsmoked. I will bear that in mind once I deal with the concerned process and instrumentation people dealing with the control system of the compressor.

Thank you very much for all the assistance.

GO PLACIDLY, AMIDST THE NOISE AND HASTE-Desiderata

 

[itsmoked]

You're most welcome. Thanks for the clear descriptions and we wish you success.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[jraef]

Yes, excellent posts 02101972, I am giving you a start for not leaving us wondering what happened!

It appears as though your PID control loop is in fact being done outside of the VFD, which is the best way as I said. I am also of the opinion that whoever wrote that specification requiring Closed Loop speed control either was under the same mistaken impression I stated earlier, or has a penchant for wasting money and needlessly adding complexity. Closed loop feedback will require a shaft encoder on that compressor motor, which is, IMHO, a level of mechanical failure risk that is unwarranted in this application. Itsmoked hit the nail on the head in his description of how it will work; the VFD speed will modulate to the PID loop regardless of the speed control method, the loop controller will not look at speed, only output of the machine. Speed accuracy is therefore irrelevent.

Eng-Tips: Help for your job, not for your homework Read faq731-376