Sensor Systems for Grid Tied Induction Motor/Generators

[Phil Timmons]

[This started as comment to Eraser on this thread . . .

http://www.fieldlines.com/story/2007/6/25/234126/646

but the comment link does not appear to working on that link, so here is some design discussion on speed sensor systems for induction motors being used for grid-tie generator applications]

Hi Eraser,

I think you are starting with a good question regarding the sensing of motor/generator speeds -

Although I would tend to agree with Finis and Vawtman that using a windmill driven source does not lend itself easily to using an induction motor as a direct grid-tie generator, it is a good method if you are using water or steam turbines, or other engine sources.  Is your application using something other than wind as the primary source?

I am building a system for grid-tied induction motors, as well, and you are correct that tracking the speed of the motor is important, but I guess you follow there is more to this just running the motor overspeed and then slamming a relay to tie it to the grid?

The motor has to be running synchronous with the grid, as well as voltage, current, and frequency need to be tracked - ideally with operating thresholds and alarms - and in most cases automatic disconnects to comply with UL requirements.  I am also thinking that putting some temperature sensors on the motor bearings and windings is cheap and easy failure prevention.  But we can cover that a little later.  For now, your question was methods to determine the actual motor shaft speed, correct?

A quick review gives a few choices - mechanical, magnetic, or optical come to mind quickly.

Mechanical - a lobe or cam bumps a switch - like the old "ignition points" system in internal combustion engines.  Mostly replaced with magnetic now as part of the electronic ignition, but still used in small motors like lawnmowers.  As point systems wear more quickly than the others, I would not recommend this method unless you have some good reason to use it.

Magnetic - magnets either on or in the shaft spin around and pass by a fixed sensor - most often a "hall effect" sensor, that detect the magnetic field pulse pass by as the shaft rotates.  These can be found in the base of most electronic ignition system distributors.  Most junked cars for the last 20 years are a source of these.  Also, some after-market cruise controls use magnets ty-rapped on the car's drive shaft, with sensors to determine road speed.

Optical - These are probably what I would select.  One method is paint the shaft flat black and then put a white stripe lengthwise on the motor shaft.  A light source shines at the shaft, and an optical sensor "sees" the white stripe flash by.  Most handheld tachometers work this way.  

Another optical method is to attach a disk to the shaft.  The disk has slot(s) cut in and one side of the disk is and LED light source, and on the other side an optical photo sensor that "sees" the LED when the slot in the disk rotate by.  These systems are very typical in old floppy drives.  

All of these methods create pulses as the motor shaft rotates.  These pulses are fed into a counter with a time division circuit (e.g. revolutions PER minute), and then can be used to display speed and/or trigger other events - such as the relay you spoke of - in the case of too low or too high of speeds.  

My bias is the slotted disk method described last.  The photo pair LED and receiver are common and cheap as dirt - about 50 cents, new, or free if you want to salvage them from an old floppy drive - and the disk can be stamped metal, or just a washer with some holes drilled in it.

Is that the type of concepts you were looking for?  If so, we can do the details of those circuits, as well.  

[thefinis]

Would love to hear more on the slotted disk method please. I have piles of old computer parts. I kept looking for an induction motor with a centrifugal switch cut in as I know that they make motors with one for a cut out.

I considered going into the problems both turbine wise and grid wise with Eraser but it has been gone over several times here. Sorry about the link did not know that it was bad. I have some links to generator controls but mostly they are for entire control box on a gas powered generator. I think that thermal switches will protect the motor (most have them already) but if used on wind it will cause a run away condition if needed in high winds.

For systems with a near constant rpm like water steam etc. a grid tied induction gen is a cheap and fairly good choice if the utilities don't regulate you to death. I still think that they have their place but many have been lead to believe that it is a cheap and easy sub for pm alts with a battery bank. I was one of those and have not given up entirely on the idea.

Hope you are doing okay in big D with the rain. Out here in west Texas we are wet but not flooding much except for areas with poor city planning that flood with any 2" rain.

Finis

[phil b]

I'm in the same ballpark. Will a 2917 tacho chip work? Otherpower made a module a while back. Hugh also has one on his site. Link that to a relay?

A PLC will work. It registers to voltage coming from analog or digital signals.

I'm up for ideas. Please keep posting.

[Tritium]

I believe that optical sensing would be a problem unless the entire assy was sealed against all dust and weather. I have cleaned spider webs out of non protected optical sensor assys that had caused a failure of the system. Dust and mud from dust/rain or dew is also a potential sensor killer as well.

Thurmond

[tdmack]

(Written for Eraser1's post in control section last week.....had broken comment link)

Eraser1

There is nothing complicated or complex about grid connecting an induction motor/generator whether powered by wind, hydro, or steam. A key item is of course the "speed sensor. For an excellent example one could use a PIC chip to control the relays with a hall device situated at the motor shaft to feed pulses into the PIC which are in turn used to detect rpm.

In the case of a wind turbine you'll also need a "powered released" brake, gear reduction, and a stall regulated turbine rotor. A properly chosen motor will include the first two and the latter is a very simple rotor which is designed to have an increase in torque output up to a given wind speed and then plateau.

This is of course a basic explanation and really not meant to sound overly simple.

The real beauty of this set-up is that you can grab some real horse power right off the shelf. We do sell most all these components but you can get them locally as well. As for our PIC controller, one could substitute with a small PLC which are becoming quite affordable.

Added: We are currently recommending the Hamlin 55100 hall-device.  It's a nice small package with a open collector output.  

Tim

[thefinis]

Tim

I hate to argue with someone with real experience on the subject but where I am it is complicated or at least not easy to grid tie. Insurance, disconnects, over/under voltage, same for frequency, plus a few other sensors but it may change as they are rewriting/updating the interconnect agreement now. All work needs to be done or approved by a certified electrician before connecting and UL approved equipment is almost a must.

You have spent how long coming up with a set of blades to match your induction setup? Was it easy or simple? Last time I checked you were still refining the controller. I think that your company has made a great product and that now for you looking back it seems not that hard a task. For us just starting the journey it is rather daunting. Kind of like how do you open a locked door if you have never used a door, door knob or a key before. It is simple if you do it all the time and have the right key in your pocket.

Thanks for the info on the hall sensor and I keep thinking that a plc may be the way to go if and when I ever get that far. I check and see what and where ya'll are on your controls ever so often because I like the idea of just buying one already setup. Thanks again for sharing your knowledge.

Finis

[Phil Timmons]

Finis:

A real simple sort of cartoon diagram of a optical disk arrangement is here:

http://members.tripod.com/simplemotor/oimotor.htm

The application is a little different, but the disk and sensor is about the same.  If that does not make sense, I can do a more detailed sketch.  One feature that is nice about this method, is that a bunch of very thin slots can be put in a disk - typically some hundred and you can get very detailed information about motor speed and shaft rotation.

Btw, if you need electrical license or permit help on a project in Texas, I can help you out with that.  I have a Texas Master and Contractor license, and for RE guys I do that type stuff for free - you will just need to cover the local license or permit fees if there are any.  

--------------------------------

Phil b:

This is what you are talking about, correct?  

LM2917 - Frequency to Voltage Converter

http://www.national.com/pf/LM/LM2917.html

That would produce an output control or signal for any set-points we wanted to cause a control to activate for any particular speed.  A concern I might have is that with it being an analog signal that controls it, there may be some variance with voltage supplies and ambient temperatures.  

But I am still not sure what everyone wants to do information from the speed sensor -  I think we are all clear that we should not be slamming a motor/generator on and off the grid based on a speed point, correct?

I do see three critical points in the operation where speeds need to observed and confirmed:

1. At start-up, while drawing power from the grid, to see the unit has come up to motor speed
   
2. When the turbine starts to spin up, to cross over from motor to generator operation, and
   
3. At full output as a generator to make sure the generator is not over-revving -- to avoid it from burning the windings and bearings.
   
   

But I will cover those in a "Functional Narrative" in my next Diary entry.

------------------------------

Thurman:

That is a very good point about dust, dirt, and creepy crawlers in an outdoor, rugged environment optical system.  

My original bias for optical was because of the high resolution that optical can provide - but thinking it over, that is mostly a waste in this application, as there is not a lot of need to know exactly what the motor/generator speed is at a very frequent sampling.  

--------------------------------

Overall, I am not really married to any particular controller - PLC, PIC, or PC are all fine by me, and my personal taste is usually along the Phillips versions of Intel 8051 family - they have analog and bus controls build right in.

But if we do this generically - first with a Functional Narrative, and then Block Diagram it with options, we can swap any controller in or out, and it will make little difference.  

So I will put up a draft of the Control Narrative, next.