Excessive RPM drop under load ?

[veggie]

Note: New 6/1 engine (1/2 hour run time)

Today I gradually took the engine up to 2/3 load (2kw) and the rpm's dropped from 650 to 575 and did not recover.
While loaded, I adjusted the governor back up to 650 and she held, but was obviously working hard (Thumping with a strong exhaust pop).

As I slowly removed the load, the rpm's climbed to 710 and stayed there  A huge variation between loaded an unloaded.

Any one else have this issue and how did you resolve it?

Thanks in advance,
veggie

[Crofter]

I think that is way too much fluctuation. I would check for proper valve timing first then spill timing and that governor linkage both internal and external has no binds. That slip yoke connecting the bell crank to fuel rack can be the culprit. Picture here of just using an elastic before replacing with a very light metal spring.

[XYZER]

Veggie,
I would bet you governor and linkage needs some tinkering. squirt some oil on it and work it.....

[dubbleUJay]

Veggie,
I would bet you governor and linkage needs some tinkering. squirt some oil on it and work it.....

............and clean off all paint that might be hindering operation!

(That probably goes without saying)

[Ronmar]

That is way too much droop with a load application.  Got a couple of questions for you: 
How are you measuring RPM?
How are you measuring the load?
How are you applying the load to the generator?

Droop is a fact of life with a pure mechanical generator.  Flyweight position is a factor of RPM and spring tension.  Since the flyweights have to move to advance the throttle, the RPM must drop to add throttle.  I don't have a tach.  I use a kill o watt and measure the output frequency of the generator to set RPM.  I set mine for 62HZ unloaded.  It droops 4HZ to 58HZ at full load around 3+ KW.  If my math is correct, that is about 120 generator RPM or about 40 engine RPM.

In a purely resistive load P(watts) = I(current) times E(voltage).  2000 watts divided by 240 volts = 8.33 amps.  Are you actually measuring voltage and current to determine load.

[veggie]

Very god points from everyone.
Thanks for the input.
That's the power of this forum. Sharing experience and information.

Cheers,
Veggie

[Ronmar]

No Jens, I don't think it is to early to worry about this...  Your saying that he needs to wait for 100+ hours before he has a generator that will hold proper useable frequency across it's load range?  In a backup application, that could be years away...  Mine still has only about 83 hours on it...  he is saying he has 75 engine RPM drop with 2KW of load.  That is around 225 generator RPM drop.  If he starts at 1800 generator RPM/60HZ, that is a generator droop to 1575 RPM/52.5HZ...  You should easilly be able to make +/- 10 percent frequency(+/-6HZ) across the load range.  Mine has always done this with lots of room to spare right from it's first loaded runs...  I had perhaps an hour or so of unloaded running before I hooked up the generator and started loading the generator.  I did perform the utterpower governor spring mod before I did any loaded running.  

I think Veggie is either inadvertently overloading it, or something is binding/not setup right in governor, IP misstimed or valvetrain issues...  

[Crofter]

 None of the components of the speed adjusting system are very smart or can do anything to compensate if another part is misfunctioning, or mismatched. governor weights from a thousand rpm engine cannot be made to work optimally at a greatly lower operating range simply by changing spring tension. By the same token a matched governor and spring cannot give good regulation on an engine that for whatever reason is requiring an exceptional rack movement to render a given percent torque increase. Each of the components goes through its inherent range of motions and if the connected parts dont react it doesnt even shrug.

Jens, could the variation in set point you see be connected to atmospheric variations. A change in output at a given rack position would be the result. The mention of the sharp exhaust note suggested to me the possibility of early valve timing. My engine, which was deliberately ordered unrun at the factory, had timing one tooth early. I dont know how well it would have run but would have been somewhat crippled. Because I am running at a lower operating RPM I had to go to a spring with a lower stack rate to get good response.

I have diddled for hours getting governors balance on welding machines where you just cannot live with hunting (or wont hunt at all,) governors. There is no magic to them but if someone happens to have swapped some components, or the engine is not up to snuff, they can make you pull your hair.

If someone with the identical engine could measure their spring tension accurately at its operating length it would be a good help in eliminating one possibility.

[12gauge]

So veggie, have you done the utterpower modification to your governor linkage?  If not that might help.

RH

[dubbleUJay]

Can anyone tell me how to set the governor linkage that connects the bottom to the top assembly? The rod with adjustable end connectors on both sides. Is there a set procedure for that?
I couldn't find any info when I assembled mine and adjusted it so that the 90deg L-lever at the bottom sits horizontal when the engine is not running. I presume this will influence the amount of pressure the spring has to apply to get RPM right

[veggie]

So veggie, have you done the utterpower modification to your governor linkage?  If not that might help.

RH

Hi 12gauge ,

Not yet, I just learned about the modification from this thread.
It looks like a simple thing to do. I can make the necessary part to enable the swapping of the spring and rod positions.
I figure there's nothing to loose by doing the mod. (Along with all the other things that were suggested by the others.)
Might be a while before I get to it. I have to shift my focus to a permanent cooling system and an exhaust system that takes the gasses outside the garage. Winter is near and I won't be able to run with the garage door open much longer. 

Veggie

[dubbleUJay]

Veggie, I can definitely recommend the mod, it made a bigger difference for me than I suspected and I was quit surprised! I've still got issues with drifting, but its much better.
I hope you get the same results, from what I've read, a lot of people did.

[Crofter]

Can anyone tell me how to set the governor linkage that connects the bottom to the top assembly? The rod with adjustable end connectors on both sides. Is there a set procedure for that?
I couldn't find any info when I assembled mine and adjusted it so that the 90deg L-lever at the bottom sits horizontal when the engine is not running. I presume this will influence the amount of pressure the spring has to apply to get RPM right

Generally with a bell crank transfer  you should adjust for 90 degree or equal angles of push rods at the operating position, not the parked position. That keeps the motion transfer most lineal. Mobile Bob had a good article on the linearity of flyball governors and why it was critical to have them correctly weighted for the operating range.

Injector pump rack position and engine output are unlikely to have an exactly linear relationship so it can be a useful tool to tweak the linearity relationships of other parts where possible.

As Jens mentioned try to eliminate any sticky wickets caused by paint or rough machining or lost motion from sloppiness.  Try to determine by manually moving the rack of the engine against a measured load whether the engines capability is normal. Eliminate the most obvious and experiment with the finer variables should get you there. Sometimes a binding condition will go away by wearing in but other times they get more set in their ways and begin to outright jamb rather than merely be cranky. Excessively high rpms after a suddenly removed load is the most potentially dangerous concern. I would be more for finding the why rather than living with a problem to see if it goes away by itself.

[dubbleUJay]

Generally with a bell crank transfer  you should adjust for 90 degree or equal angles of push rods at the operating position, not the parked position. That keeps the motion transfer most lineal. Mobile Bob had a good article on the linearity of flyball governors and why it was critical to have them correctly weighted for the operating range.

Thanks Crofter, I'll try that 2morrow when I have a chance.

The article of Bob, was it on this forum or can you give me a link to it please? That's to say if its not on the LEF which is down. 

[Crofter]

I think Bobs article was on LEF. I will see if I can link below to some pics I have of the 10-1 on photobucket. There is one shot of the linkage showing the rearrangement of the spring, adjuster rod and added pivot anchor but othe guys have much more specific photos. My present arrangement uses two much lighter springs in tandem to give lower stack rate and narrowed the frequency swings a couple of Cycles per second on the Kill O Watt meter; a definite improvement.

http://s655.photobucket.com/albums/uu276/Tijean_photo/