Regulated output generator

[GuyFawkes]

The "discussion" is pointless because you cannot discuss anything with someone who flatly refuses to use their brain (and usually also insists on not being treated like an idiot) which is what we have here, so I will make it real simple for you and explain it in terms that a 5 year old can understand.

There are 720 crankshaft degrees in a 4 stroke cycle, about 100 of which will be useful power stroke.

At 650 rpm those 720 degrees take 0.185 seconds, of which the power stroke embodies 0.025 seconds in "old money" that is about one fifth and one fortieth of a second respectively.

That one fortieth of a second useful power stroke is THE ONLY TIME ANY ENERGY IS BEING PUT INTO THE SYSTEM.

The rest of the fifth of a second the ONLY ENERGY IN THE SYSTEM, available to be converted into electrical power, is the inertia in the rotating flywheels.

If we take energy out of the system we, laws of physics here, are converting RPM into AC, we are doing this all through the 87% of the engine duty cycle that are not a power stroke.

The more energy we take out, the more RPM we sap, broadly speaking if a 1 Kw load sapped 10 RPM then a 2 Kw load will sap 20 RPM.

Now we come to the power stroke, the 13% of the cycle, the fortieth of a second every fifth of a second.

Open the rack enough to accelerate from 640 RPM to 650 RPM, in that fortieth of a second, or open the rack to accelerate from 630 to 650 RPM in a fortieth of a second.

Rate of change of RPM for these two scenarios is 400 RPM a second squared and 800 RPM a second squared respectively.

Both these events take place, near as dammit, in a fortieth of a second (the difference is less than 2%)

But one takes TWICE THE ACCELERATION, which is TWICE THE RPM CHANGE OVER TIME.

=============================

Quite how you are able to imagine that a computer controlled rack vs a mechanical governor is somehow, because it is "more intelligent" (though I will grant it is probably more intelligent than many posters) going to subsume the fundamental laws of physics such that 400 rpm/sec/sq and 800 rpm/sec/sq are now miraculously exactly equal graphs I don't know.

I suspect instead it will be easier to continue to flatly refuse to use your brain and pick some other imaginary "reason" why fundamental physical laws do not apply.

============================

You wonder how the dude who keeps popping up selling 100k hour listeroids manages to keep doing it and keep selling them?

take a look in the mirror, you don't have to look very far to find people who work hard at preferring to believe a load of bollocks because it happens to appeal to them, in preference to actually using the brains god gave them.

And y'all are desperate to label matey with the kid stood next to the flywheel as being stupid.....

Guess its time for me to leave again

[GuyFawkes]

I am not sure what exactly the 13/87 figure tells you but if your controller sets the rack to the correct position the moment the load changes (not based on rpm but based on current draw),
Jens

You see, it is this sort of utterly moronic statement that makes engineers ignore you.

The moment the load, eg AC elecrical load on the gen head, changes.

Not based on RPM

But based on AC current draw (cos voltage is taken care of by the field)

So Jens, by attaching your PID to an indian listeroid + chinese head you get a free energy device, one of the science fiction zero point energy machines, over unity in action.

Must be, to be able to increase current draw without actually converting any rotational energy into electricity, and you can't be doing that because you are going to respond to current draw, not to rpm change, because responding to current draw is, according to you, faster.

[jzeeff]

After a power stroke occurs, there is nothing you can do to increase rpm until the next power stroke.  But it's a long time until that occurs.

So you are going to have short term variations in rpm and voltage - how much depends on how much energy storage the flywheel, etc store.

Yes, over a long period of time (several power strokes), you can control rpm well.

[mobile_bob]

i just can't help myelf,, and i am sure i will regret stepping into this discussion

1. the pid controller can operate at fantastic speeds, lets call it a ferrari/cheetah/rabbit (whatever it is fast)

2. the lister 650 rpm motor is much like a dynosaur, the one with the really long tail

no matter how fast the brain is it takes time to get the signal from the brain to the tip of the tail
a lot of time!

makeing the brain faster doesn't help, you gotta cut the transmit time, and you got to speed up the reaction time

one of the major issues with a slow speed engine, is the eternity it takes to make changes

you have a single event called injection, followed by a delay time before ignition and then 8 more eternities before the
process repeats.

to a pid controller anything that happens in a lister is an eternity.

the only way i can see a pid controller working, and then only with tons for feedback, coding, very specialized design etc.
is with a constant speed (all well and good) and at a set constant load (which is never the case) the load constantly shifts
slightly even if it is a single resistive load such as a heater (drafts will shift the load irratically)

so it basically comes down to incompatable technologies, the pid spinning code like crazy and the lazy old lister plugging along
not giving a crap what the pid wants, and the poor old injector rack being sawed" on like charlie daniels playing "the devil went to
Georgia" on his violin as played  on 78rpm instead of 33rpm.  

just when the pid gets it sorted out, gets the rack to move, waits for the pump to inject either more or less fuel, wait some more
for a feedback of rpm, look up on the table because it missed, recalc and repeat,, well maybe it gets lucky
but then the load shifts slightly and all bets are off.

so there would have to be some hysterisis, a fudge range to operate in
maybe +/- 1 hz would be attainable with a lister with pid controller?

maybe +/- 1 hz would also be attainable with some work on the oem governor
and a 69 cent spring from the hardware store?

and to be perfectly honest i am eaten up with crazy over microcontrollers,, and am working with them
for various other control systems for offgrid use,, control of rpm for a lister is not on that list.

lastly maybe this makes some sense, an analogy

the engine is a battery charger
the flywheel is a battery

the engine charges the flywheel and the generator takes power from the battery/flywheel

it takes time to recharge the battery
and the charger is intermittent by nature, pulsing a charge every other cycle of the engine for a short time.

add more pulses is the only way to improve the problem, more pulses means more cylinders or
much faster rpms so that the pulses come closer together.

i don't think anyone wants a 6krpm listeroid, and i don't even wanna see on on utube!



bob g

[Jim Mc]

...
One should easily be able to stay within a tiny frequency range - maybe 1/4 hz plus or minus. Of course this figure is grabbed totally out of the air but I don't see any reason why this wold not be achievable....

Here's where the trouble starts...

Jens probably meant to say a PID controller could keep the set's average frequency within 1/4 Hz.  And well it could - on average. and over several a several second period.

But he left out the word 'average'.

GF has correctly pointed out that the cyclic variation exceeds 1/4 Hz.  And well it does.

Geez guys, take a breath.  Read the words in front of you...

[mobile_bob]

wow, i thought i succeeded in pissin off another member!

lol

i just assumed you were after getting right on 60 hz +/- 1/4 hz
and not an average over several cycles,

i am sure one could get an average within that region, but am not sure
if this solves the problems most reported by folks that need such precision

but maybe? it would be adequate for those that need an average over several cycles?

bob g

[mobile_bob]

awe crap,, just pitch that listeroid in the scrap iron pile and

buy a changfa!!

(and call it a day)


muwaaahahhahahahha




bob g

[oliver90owner]

awe crap,, just pitch that listeroid in the scrap iron pile and

buy a changfa!!

Or as GF said, buy a 6 pot engine and the Hz change will be reduced by a factor of 6.  Then run it at 2400 RPM and the time between energy input (engine speed correction) events will be reduced by about 3 1/2 - 4 times.  Now you have well over one energy input cycle every generator cycle.  Now it is much easier to reduce the variation.

Again I will say instant increase in engine speed requires no inertia - no flywheel.  Unattainable, but you can see what I mean.  Huge inertia will mean less change of Hz per power event but you may never get the engine started! And the inertial mass will always have a power loss associated with it, whether it is extra bearings, air resistance or other (unless we go to mag lev and operate in a non-leaky vacuum system).

Just give up and live in harmony with your Lister.  KISS.

Regards, RAB

[mobile_bob]

awe darn RAB wheres the fun in that???

KISS is overrated!!

don't you just love overly complicated things in life?



if you really want smooth stable power how bout a 12 or 16v149 detroit
2 stroke and lots of power strokes per cycle. bet flicker is no problem!

course its a bit of overkill for the average 6/1 owner i suppose

bob g

[dieselman]

I think Guy needs to change his medication.

Jet engine?  Hmmm.  How about using the listeroid to store energy.  Lift a large weight on a pulley system like a grandfather clock... Or filling a tank with water. 

I really didn't expect a firestorm from my original question, just some stimulating conversation.  But it has been interesting.

Jim
 

[SteveU.]

Jens, thanks for answering Steves question.

I lay awake tring to craft an easy answer to that. Yours is much better than mine would have been.
I tend to just point people towards the same book/internet sources I've  had to use with only my high school education to learn new bodies of knowledge.
For some reason that seems to piss some people off. (Happens with family and friends.)

KISS is the cake. You can add all the frosting, decorations, and candles you wish as long as the underlaying cake could still provide for life. Example:
Bought a 69 Chev pickup in 1980. It was a stone simple 6 cylinder with point type ignition. I was going to convert it to propane--never did. I did change it over to aftermarket Hall effect electronic  ignition, 50kV epoxy coil and big fat 8mm wires;  and then  even added a piggy back Multiple Spark Discharge modual.
Started much better, ran better, better milage.
It would then eat up a set of spark plugs every 6000 miles-- no platimun plugs then.
Over the course of the next eight years/150,000 miles money was tight and had to be used on a clutch, brakes,shocks, radiator, waterpump, two drivelines, three exhaust systems, ect. So as the electronic add on systems failed one by one; I removed them.
Sold the truck ay 300k+ miles running with the original point type ignition.
In my book: thats KISS.
I learned to like pie much better than cake.

One of the other Steves,
SteveU.

[oliver90owner]

Bob,

how bout a 12 or 16v149 detroit
2 stroke and lots of power strokes per cycle.

The exhaust note mught need a bit of suppression!  Not quite like a 'Johnny Popper', more like a howling banshee?

My experience of the detroit has been limited to a few tractors I have seen.

Regards, RAB

[1958steveflying]

Thanks Jens that did the trick.
  Steve

[compig]

Ultra stable & accurate frequency from a power generator = gas or steam turbine. Continuous combustion is the key. Any IC engine generator always has huge flywheel mass so that kinetic energy does as much as it can to stabilise voltage and frequency within the confines of physics , correctly outlined by GF. For the PID controlled governor curious amongst you , someone on this very site has already done it , but I don't have the time to find the post just now.

[jzeeff]

Did somebody correct the math on this?


> a 600 rpm engine and 60 hz output frequency, there are roughly 10 rotations of the crank in one complete output cycle. That's 2.5 power strokes.

600 rpm = 10 strokes/sec = 5 power strokes/sec = lots of 60Hz cycles between power strokes.