Reading RPMs by PLCs - Revisited

[AnaDigit]

Dear PLC Gurus,

I don't want to beat this topic to death by I need to read the RPM accurately from 4 pumps using a MicroLogix 14000. I have read many posts about the RPM topic thing and decided to test what results I would get.

I am ready to use a HSC for that purpose reading a 60 PPR encoder, but before commiting to use a PLC, a decided to setup a RPM simulator on the ML1400 using the known frequency of the free-clock-scans to simulate the encoder pulses.

I have mixed reviews of the results I see in real time.

I would appreciate experienced programmers take a look a the attached files:
1) RPM.RSS - RPM Simulateor. Easily adaptable to other AB PLCs
2) RPM.pdf - A starting location for anybody to use and try this simulator.
3) RPM.xlsx - Collected samples of a variety of simulations paged by RPM.

Files from items 2 and 3 were zipped into a RPM.zip

I hope someone can look at this stuff and tell me if what I obtain is correct or if I need to make corrections somewhere, or use a different approach altogether.

Greetings from Bakersfield, CA
AnaDigit

 

[iant]

Welcome to the site mate.
Not certain why you need 60 PPR encoders?
what are the RPMs of the pumps?
Are you using FSD's

 

[AnaDigit]

Hello Ian.

A 60 PPR encoder is what I have on my desk. Actually, the existing equipment has 30 PPR encoders.

I don't know what a FSD is; and, since I never heard of it, I would say no.

What I have is (14) 4-stage centrifugal pumps having synchronous motors rated 80,000 HP, operating at 14.4 KV and drawing 2500 Amps (60 MW). The rated RPM is 600 and we already have tachometers for speed and direction protection.

Given the size of the units, we use a soft-start (generator) to start them. This generator brings the motor up in speed to 570 RPM where the motor breaker (line power) is closed; some seconds later, the generator breaker opens.

I need to investigate a random problem common to only 4 pumps during startup: they fail to reach sync speed.

Synchronous motors start as induction motors, switching to synchronous at 95% of rated speed - 570 RPM in this case. The existing motor tachometer is responsible to apply 14.4 KV line power breaker.

My objective is to use another sensor on the same tach head of every motor and read the RPM from this sensor without disrupting the existing protection. All four sensors will be connected to the PLC. I will also log the startup RPM of every unit only during the initial 5 minutes of every start on a PC for later analysis.

My need is to read fast and accurate RPMs as I monitor a synchroscope during startup. The existing tach meters have a display update latency of up to 2 seconds or more, and that is not useful to me. I hope to overcome that with the PLC.

I read in another post that the AB 1769 platform may read RPMs directly. I need to find a way to obtain fast and accurate RPM data.

Cheers.

 

[Lancie1]

What I have is (14) 4-stage centrifugal pumps having synchronous motors rated 80,000 HP, operating at 14.4 KV and drawing 2500 Amps (60 MW).

Those are some honking big motors, definitely above my experience level and pay grade.

 

[AnaDigit]

Size does not matter - the principle is the same. Yet, given the price tag of the infrastructure I need to be accurate and precise on what I do.

Bottom line is I need to find a way to read fast and accurate RPMs.

Greetings

 

[Ken Roach]

I think that using the clock bits of the controller is introducing another element of uncertainty, because the clock isn't synchronous to the PLC scan. You could have a lot of unproductive head-scratching trying to figure it out.

I would instead use a real signal generator, like the sort that some oscilloscopes include.

There's a pretty good example for calculating Rate using the HSC and the STI features of the MicroLogix 1400 controller in the RA Knowledgebase:

https://rockwellautomation.custhelp.com/app/answers/detail/a_id/19445

The 1769 module you're thinking of is the 1769-HSC, which does have a Rate calculation mode.

https://rockwellautomation.custhelp.com/app/answers/detail/a_id/27622

When you're dealing with slow speeds, the Pulse Interval method is usually more accurate than the Cyclic Rate method. Try them both, but only with a good, physical, hardware-based pulse generator.

 

[AnaDigit]

That was precisely the point. Using the built-in free-clock pulses to simulate the field sensor which will always be asynchronous to the PLC scans.

But, it is worth giving it a try anyways. I may have a dusty signal generator somewhere ...

Cheers

 

[Ken Roach]

Maybe I need to re-state my point: using the clock bits and the ordinary PLC logic scan will be both inaccurate and inconsistent. You've already put a lot of work into illustrating that. Using a physical pulse generator with the ordinary PLC I/O and logic scan will be similarly inaccurate and inconsistent.

The High Speed Counter function of the MicroLogix controllers is an independent hardware subsystem that doesn't rely on the controller's main I/O scan or program scan. The interrupt subsystem runs on 10 microsecond clock pulses, so it can be more accurate than the ordinary I/O scan and PLC logic scan. The inputs that are associated with the HSC are faster than the ordinary general-purpose I/O of the controller.

I haven't used the MicroLogix 1400's Pulse Train Output (PTO) feature to try to generate a steady pulse stream. You could give it a shot... you might even use the acceleration feature to emulate what the motor will be doing.

 

[AnaDigit]

For sure I was planning to using the HSC for that purpose. But, thanks again for remind me of it.

 

[iant]

Sorry for the typo. VSD or VFD. But that is obviously not the case here.
With 4 motors of that size not getting up to sync speed there is a good chance it is not accuracy you may need. At 600 rpm and taking into account the HV nature. You might be able to use a plain sensor of a shaft or similar.
I note that this is a bit more involved than your initial post. Understated slightly
Is the one generator used to start each pump in step?
Am I correct in assuming that the motor is reaching 570 rpm ok?

 

[AnaDigit]

All 14 pumps had been working fine, disregarding random fails to sinc once in a blue moon. This problem started happening after we replaced the old pumps for new ones to improve efficiency. So, what is the common denominator? The new pumps - They are more efficient and lighter.

But, don't be fooled into believing that lighter means less work for the generator during startup - or maybe. For sure the pumps start and must reach sync mode against a closed output valve, where by that time the pumps have develop ~900 psi of output pressure. This is the required pressure to lift the water we pump some 2000 feet to the top of the hill.

There are also fluid dynamic forces that have changed given that inertial forces inside the motor-pump combo also have changed because of the mechanical differences of the new pumps.

Interesting stuff huh? Oh my God! ... Why did I decide to become an engineer?

 

[iant]

I would suggest there is something different with the new motors.
If they are the only ones giving you problems, they may have some other design issues.
Are you using the one Generator to start these motors?

- time to quit engineering and go to the pub (bar)

 

[AnaDigit]

The motors have not changed - The pumps have.

By the way, each pump is rated 355 cfs or 2656 gallons/sec.

Regards.

J R Martinez

 

[AnaDigit]

Thanks for the pointer Ken.

I do not have an active TechConnect account now and in the 2012 KnowledgeBase DVD which I do have, I could only find TechConnect ID 27622 and not ID 19445 ... I'll figure something.

J R Martinez

 

[g.mccormick]

What the heck are these things for? Filling a cities reservoir?