CVT for rpm and frequency control

[pigseye]

Could someone comment on the possibility of using a constant velocity transmission (CVT) to maintain rpm and frequency control of a gen head with varying loads?

Would it be more accurate and consistent than a governer?

Thanks,
Pigseye

[GuyFawkes]

you'll lose a fair bit of power to it as heat....

sort your governor, if sorted it will do an excellent job, no need for anything else.

[trigzy]

Pretty much impossible.

CVT actually stands for Continously Variable Transmission - ie. it has infinite ratios due to changing pulley sizes.

The unit would try and spin the gen-head as fast as possible for the given input power, acutally causing overvoltage/over frequency problems.

Steve

[listeroidsusa]

If you want to control the engine speed to close tolerances it is pretty easy to add a tach generator to the generator drive, add a cnc servo amplifier, and use a solenoid for the throttle control. The servo amplifier will have a gain control for the speed, and can be simply a 10K ohm rheostat to control a 0-10 volt input to the servo amp, with the servo amp controlling the output to the solenoid, whifh will be acting as a linear actuator.

Mike Montieth
Listeroids/USA

[DirtbikePilot]

Uuh... Mike, I think I speak for most of us when I say that your version of pretty easy and ours are two different things. I get the gist of what you are saying, but I think that very few people actually know enough about that stuff to actually do it. lol Now, if you were to offer a kit to do this, with instructions..... That would be cool! It would be really scary to see what happens if it breaks though. I suppose that could be quite a liability if it broke and flywheels exploded.

[listeroidsusa]

If I am able to get more engines this is the system I plan to use on my generator sets. The original governor is still used, but is set for an idle speed of around 500 rpm. The servo controller has the advantage that there is NO rpm droop when the generator is loaded. A digital system can easily control the rpm to a limit of 1 rpm or less with instantaneous response using the PID loop in the servo controller. The P takes care of proportional loads, the I is the calculus integral that damps the response swings, eliminating hunting, and the D, or derivative, takes care of extreme load change response. I built a sample control using a CA3228 IC chip which is designed for cruise controls but this chip has been discontinued now because new cars use 2 references for speed and the older cars only needed one. I'll probably use the more expensive but available servo controllers such as Advanced Motion Controls. This system is similar to high end woodward electronic governors. (Its an amazing experience to stand beside a 500 hp genset that goes from standby to full load with NO change in rpm!) The setup is similar to the start-o-matic. The SOM used a linear actuator and a gain control  and sensed the generator speed, but is not as sensitive as mine will be.

Mike Montieth
Associated Sciences & Technology
Isothermal Community College

[GuyFawkes]

there is a third way.

hydraulic governing can be as near as dammit instant response (within one crank rpm anyway, which is as good as anyone needs) while also being ultra reliable and not prone to overspeed events....

but, it is more complexity

running up my start-o-matic under load yesterday there was rev drop, I knew there would be, one look at the crudded up governor linkages and springs tells you that, the lister way was KISS, may not be the absolute best, but in reality it was good enough.

anything requiring better quality power should go through a power conditioner anyway

[Doug]

I have a fourth way untested....

Airtronics makes some cheap little current sensors that can be tunes to pick up a set of contacts. Use the contatcs to drive a solonoid to increase or decreas the govener spring tension. This wouldn't be as smooth or stepless as Mike's or Guy's sugestion but a small increase in tension should reduce the frequencey droop at full load.

http://www.airotronics.com/custom/index.htm

Found one of these in a roll up door and I saw the potential for exactly this application right away.

Doug

 

[pigseye]

Wow, there's a lot of engineering going on here.   I appreciate the explanation of CVT, now I understand it better and see that it's really not a viable solution for gensets.

Mike, Your approach is really neat but probably out of my league.  I like the kit idea!  If you build it, we will buy it.  ;-)

Doug, I'm not sure how the current sensors would work or how the systemw ould be closed loop, could you please elaborate?

My main concern is that I want to use my generator to power my home load and a Central AC unit.  I'm concerned that I'll have significant frequency and voltage droop when the compressor of the AC kicks in.  I don't think I need 60HZ +/- 1hz accuracy, just something to keep it closer to nominal.

Thanks,
Pigseye

 

[trigzy]

Doug,
      I rigged up something similar to what you are describing for my little 12V generator.  3.5HP briggs powering a 105A chev alternator.  When load exceeded a preset point, or voltage fell below 14.1, the selenoid pulled in and reved the engine up to full.  Selenoid was released with voltage at 14.4 for and load less than 25 amps or so.  3 minute delay on rev-down to prevent cycling.  I just used a shunt resistor to get a voltage signal from the output current.

Pigseye,
          I wouldn't get too excited worrying about governor systems you dont know if you'll need.  If you try the A/C, and get serious issues, then worry about improving/modifing your governor setup.  No point in adding extra failure points or cost if they are not needed.

Steve

[Doug]

Pigseye:

What I sugested wouldn't be a "Closed Loop".....
Just a dirty little way to add a little more spring tension. Steve also sugested a timer not a bad idea but you would have to muck about with this for a bit and see how the whole system works.

Doug

[pigseye]

Good advice guys.  Thanks