Hi All,
The last 4 times I ran my engine, I have had a hard time getting it to hold a set speed.
When I apply load (say 1.5kw) the engine bogs down to 57hz and stays there (only 1/2 loaded).
So, I increase the speed a bit to get 60hz while it's loaded.
When I release the load, it revs up to 68hz. !!
I add load again, it drops right back to 56...57...58...57 and does not recover.
If I adjust the linkage to get 59hz to 60hz under 1.5kw load, it slowly revs back up to 68 HZ when the load is release (takes about 30 seconds to reach 68hz).
I have previously done the spring modification to reduce any play in the linkage.
Although there seems to be excessive play at the pin joint on the rack.
Very loose Clevice pin fit.
Could this loose pin joint create the symptoms described above or am I chasing two different problems here ?
Has anyone else had to solve this loose plunger/bore fit at the rack connection ? If so, how did you solve it ?
Cheers,
Veggie
Have you tried a spring across that angled linkage and yolk shown at the end of your video? I have a VERY lightweight spring very lightly loaded, connecting between a ring(between yolk and rack) that is around the pin that runs thru the rack, across that 90 degree bellcrank to the pin that connects to the vertical rod that runs down to the governor bellcrank. this takes the slop out of those pins and really cleaned up the lag and overshoot in my governor linkage. It must be a very light spring force, as too much spring tension will cause the yolk connected to the rack to bind.
(http://i270.photobucket.com/albums/jj85/rmarlett/th_linkage.jpg) (http://s270.photobucket.com/albums/jj85/rmarlett/?action=view¤t=linkage.jpg)
I also did the "Utterpower mod" and took that lower hooked adjuster out of the main governor linkage and ran it thru a block bolted to the crankcase where the spring was originally attached. The spring is now connected to where the hooked adjuster was originally with a cotter pin. This also greatly improves governor response.
(http://i270.photobucket.com/albums/jj85/rmarlett/th_187b19b3e367a22abba5c30a50b7877a.jpg) (http://s270.photobucket.com/albums/jj85/rmarlett/?action=view¤t=187b19b3e367a22abba5c30a50b7877a.jpg)
I had the same performance problem, caused by the yoke pin binding. I drilled out the arm for a bronze bushing from ACE glued in with weldbond, then drilled the bushing to fit the pin after I smoothed and polished it. Both the pin and the arm looked like they had been machined with a chain saw.
Check your pin/arm joint for binding when under pressure.
In addition to all the previous suggestions, check out the ease of movement in the injector rack. The symptoms you report are exactly what I saw when the pump was gummed up. The gumming up is apparently due mostly to high temperature. When Thumper was in it's own sound proof box, the injector pumps gummed up regularly but when the whole engine room was sound proofed and the engine enclosure was removed, the problem went away.
BTW, we are talking about 500 hrs before gumming so if you are still running a show engine rather than a work engine, this is unlikely to be an issue.
Point is, the governor has limited power and unless everything moves VERY freely, you will get issues. Some speed droop is normal although your issue seems a bit extreme.
Jens
Quote from: BruceM on March 16, 2013, 08:22:57 PM
I had the same performance problem, caused by the yoke pin binding. I drilled out the arm for a bronze bushing from ACE glued in with weldbond, then drilled the bushing to fit the pin after I smoothed and polished it. Both the pin and the arm looked like they had been machined with a chain saw.
Check your pin/arm joint for binding when under pressure.
The pin and bore were definitely binding.
Terrible finish on the "machined" surfaces.
A week ago I polished the surfaces and lapped them with valve grind past. Then polished the pin with emery.
Glass smooth.
BUT, they still bind because there is so much clearance in the bore that the pin cocks at an angle and grabs the bore.
I too thought of a bushing to get the clearances back to something respectable. Perhaps that is a good upgrade to perform.
With the bushing I was able to support the yoke pin more fully, so that there is almost no exposed pin when the rack is closed. Measure yours with the rack closed, and then add a 1/16 less height when gluing in a bronze bushing. The bush and pin polish job is cheap, at least.
The length of engagement of the pin and sleeve should be increased, and these parts need much better fit and finish. It seems likely to me that something was copied incorrectly or was "value engineered" too much.
Try to give the vertical arm half a turn longer....it worked on my genuine 6/1
BruceM,
The machining is done and the clearances are matched.
The rod slides nicely in the yoke with minimal play.
When lightly oiled it's as smooth as butter :)
Oiler,
Thanks for the idea of adjusting the vertical rod.
I will give that a try, but first I need to know exactly what we are effecting by adjusting that rod.
Eg: What is happening to the system when we make it longer or shorter. ??
And how do we know whether to lengthen or shorten?, based on what symptoms ?
Anyone ?
veggie
*** UPDATE ***
Here is some feedback for anyone else who is "tuning" your governor linkage.
View your clevice / IP rack alignment from the top.
Mine was out of alignment enough to create drag once assembled.
I added a machine washer on the Clevice lever shaft between the engine block and the lever in the appropriate thickness to bring the joint back into alignment.
(approx. 0.025" thickness)
Now the linkage is free and smooth.
I tested the engine and the governor is now very responsive.
62 hz unloaded and 60 hz Loaded.
The circle in the picture below shows the interference.
Good post, Veggie. That's a new one and I'm sure it's not unique to your engine. That much side force on the yoke pin would certainly make it bind up unnecessarily.
Bravo, you're governor response is looking good.
Best Wishes,
Bruce
Still working on that seemingly universal gov problem here. In a no load to 2800W load situation, I have 5Hz difference.
What I've done so far:
Utterpower mod on the spring/hook.
Bushed the top clevis, and extended the bushing a bit. Never had any binding beforehand, just lots of slop.....in fact so much that it was not necessary to drill our the bellcrank in order to fit a bushing!
Bellcrank pivot bolt bearing surface and thread sections were not parallel, bolt wobbled as it was screwed in, moving the bellcrank side to side. Made a new bolt.
Made new, slightly larger diameter pins for all the clevis ends, drilled clevis' and bellcranks to give close tolerance fit.
Found top clevis binding on the rack due to misalignment of the top bellcrank (probably the reason for the original bent pivot bolt.
Carefully (!) bent the bellcrank a few thou to eliminate any binding.
Everything moves freely as individual parts and as fully assembled. The slop in the linkage is gone.
Changed adjustable rod to find optimum length.
Originally, I had about 8Hz difference between no load and 2800W load, all the above has netted me 3Hz. Now have 5hz difference, maybe that's the best I can hope for.
I am in a remote area, so have had a spot of bother in finding different gov springs, but perhaps that's the next step?
Open to suggestions for further refinement.
Thanks for any input.
Regards to all
Hugh
Well you've covered everything except for the weights and ring on the cam shaft. Since you've gone as far as you have, the cover can be removed without pulling the flywheel by removing the studs. The ends of the weights should be rounded where they contact the ring. If that looks good try a softer spring.
+1 Tom's suggestion. Softer spring until it hunts. I had to use two springs, one was close but hunted, the second was a very soft add on. 2 Hz is hard to get, be happy with 3Hz full to no load.
Thanks to Horsepoor on LEF, I think we have a winner in the governor spring hunt.
I posted the below on LEF, and wish to share it here. This spring modification worked for him, and it worked incredibly well for me.
Quote from: Horsepoor on September 26, 2013, 10:06:57 pm
Home Depot stock number is 759 541 and comes in a pack of two different size springs for a total of four. The sizes are: 7/16 in X 2 inch X .062 in with a safe working load of 18 lbs - these are the two you don't want to use. In the same package is: 9/16 in X 3 in X .054 in with a safe working load of 9.55 lbs - this one worked well for me out of the box. So for a couple dollars you may have a repeatable solution. Remember, where the India guy decided to drill the stud hole for the other end of the spring is probably random but I hope it will be close.
Great find, Horsepoor.
I bought a package of these springs on Ebay (not available from Home Depot Canada) Tried one as you suggested.
With all my previous governor mods, the best I could do was a 4Hz range. Included in mods were an assortment of springs. (See my above post)
Just installed this spring and gave it a try.
JKSON 6/1 driving Utterpower PMG
Results...........DRUM ROLL..............
No load 60.4 Hz
3150 Watt load (measured with Kill-a-Watt meters on both legs) 59.4 Hz
Pretty damned good.
Get yourself a package of these springs and give 'er a try. Sure works on my engine too!
Cheers,
Hugh
Hugh,
Thank you for the compliments. I have achieved 2 hz from zero load to full load on both my listeroids. But, the search for prefection has continued. I recently may have found the ultimate solution for at this site:
http://www.governors-america.com/products/governors/Analog/ECC-Series/ECC328-12 (http://www.governors-america.com/products/governors/Analog/ECC-Series/ECC328-12)
I asked for a price, if it is less than $200 I am going to buy one.
Bruce
Horsepoor- better check with the manufacturer about this unit's stability with very low speed engines. It may not be stable for our 650 rpm engines.
Best Wishes,
Bruce
Good point, but as I think about it, if this is an issue I might be able to use the lever arm on the actuator to reduce the response by placing the linkage closer to the pivot point. I do t know, just thinking about the analog governor in my quest for true listeroid euphoria. Before I buy, I will inquire, I like the ability to regulate engine speed directly from the ST generator head by measuring frequency. If it is less than $200, I will buy one and test it.
The problem with our low speed engines without the heavy SOM flywheels is that there is considerable variation in rotational speed, which shows up as a variation in generator speed (with a 10 Hz period). The speed is slowest during the compression stroke, and highest after the power stroke.
If the governor can be "detuned" enough to ignore this variation, then all is well. But for normal 1800-3600 rpm gensets, this level of insensitivity would allow very poor speed regulation. It would be unusual for an off the shelf governor to handle a 650 rpm engine, unless it has been specifically designed to do so. (Thus my suggestion that the manufacturer be contacted.)
In one video linked to from this forum of a very professional setup, a modern commercial electronic governor was applied to a Listeroid. What resulted was constant 10Hz thrashing of the fuel rack. While it worked, I expect the fuel rack will be worn out rapidly.
Interesting, and good points. To offset this, I am using a ST 7.5 with a very heavy cast iron two speed sheave spinning at 1800 rpm. This 70 lbs of rotating mass eliminated my light flicker and might be enough to smooth out the power pulse onto the governor speed control system. What do you think?
Horsepoor, It seems like your generator's rotating mass at 1800 rpm has done the trick. I hope it will solve the governor stability issue, too.
Bruce
I must admit, with the help of others and this forum, I have solved all my significant issues. Now I just tinker to see how much lipstick I can put on my pig. I've achieved less than 2 Hz drift but I like to tinker, thus my interest in the analog governor if it is cheap enough.
I tried the Home Depot 759 541 spring today on my older Metro 6/1. Alas, it is a bit too soft and hunts badly. This spring, while close, is not a perfect match for all Listeroids.
I added a second even softer spring (one of a selection I got at ACE for my first round of spring fiddling) and now it's better than my old 2 spring setup. No load of 62Hz, and fully loaded 59.7 Hz. No hunting or instability with all the various loads I could try. My old setup had just slightly worse range (loaded 59.5), but also for some loads it would hunt a little.
I suspect I could accomplish the same thing by shortening the Home Depot spring, but I didn't try that.
My rule of thumb for spring fiddling- if it hunts, add another spring, or switch to a stiffer spring. If it varies too much with load, go to a softer spring.
My thanks to Horsepoor and Veggie !
Bruce, see my post on the Lister board about this subject. IIRC I got 1.5 hz with a 7/16 x 2.5 x .043 homodpt spring.
quite a few variables here to proclaim a single spring is the be-all, do-all fix, the biggest of which is flyweight mass. I highly doubt all these engines have the same weight flyweights... After cleaning up the linkage pivots and taking out all the slop of that very sloppy yoke with a VERY soft spring), I am very happilly running the main governor spring it came with. Unless the governor spring rate is matched to the possible flyweight it must counter, it can hunt. Excessive slop or binding of the yoke can also cause hunting as it will take a larger movement force(change in RPM) to cause the governor to move the rack and compensate. Taking the slop out is probably the biggest single improvement you can make. Being unable to take all the slop out of the yoke was impractical, so I used a VERY soft spring from an old tape recorder or CD drive, to gently hold the slop at one end of it's travel without causing binding. This single thing had the biggest effect on stability of any thing I did. It is very simple to experiment with using a rubber band. My ultimate fix would be to use a weed-eater throttle cable in place of all the linkages and add a very soft spring connected to the opposite end of the rack to keep the cable in tension and to close the throttle and shutdown if anything should come apart...
I am happilly getting +/- 2Hz from no load to full load. 62 droops down to 58. I think this is fantastic for an anchient governor design. 2Hz is absolutely phenomenal IMO... Check your utilities advertised specs for frequency. +/- 5Hz is probably pretty typical, although they typically do far better than that...
Yet, nothing was done to remove the slop on the linkage on my machine and with just a spring change now have 1.5 hz response.
Well you must have a fairly tight yoke and linkage then. The time required to take up all the slop in the linkages and IP is the main cause of lag/overshoot. It can also induce a type of random hunting if the yoke is really bad as the rack will move a little on it's own causing uncommanded throttle changes that the governor will then try to sloppilly regulate:)
Well I don't know how tight the linkage is compared to others as I've only had my hands on one other roid. There is a fair amount of slop in it though.