Whats your best guess?

[flywheel]

Looks like there are still guesses coming in so its not quite time to post the answer, sleep well. flywheel

[Henry W]

How about giving the winner.

Henry

[TimSR2]

I vote  for no significant difference in power production.  It's  5 percent or less. The extra revs give 10 percent extra power, and the belt wastes 5 percent of it.  Net difference in power production  is negligible.

However,  the fuel consumption curve of the engine spikes at the higher revs, so from an efficiency standpoint the direct drive would win, using  10 to 15 percent less fuel to do the same work.

TimSR2

[vdubnut62]

Dude! Come on already! We crave answers!
Ron

[flywheel]

OK-OK, Let me dig out my written test results which were done about six years ago when this genny was assembled. I will post the results within a few days.

Hey vdubnut62, what model is that JD tractor in your thumbnail, kinda looks like a 70Dstd from a distance. Is it?
Flywheel

[vdubnut62]

That Sir, is my pride and joy. It was a Father's day gift from my lovely Wife about 10 years ago! Yes really.
1955 70 diesel  372 cubic inches in two cyls rated 50 hp @ 1250 rpm row crop with a wide front end original unrestored showing 1200 hours. It has a pony motor start, John Deere pony motors ARE the Devil.
Ron

[flywheel]

That Sir, is my pride and joy. It was a Father's day gift from my lovely Wife about 10 years ago! Yes really.
1955 70 diesel  372 cubic inches in two cyls rated 50 hp @ 1250 rpm row crop with a wide front end original unrestored showing 1200 hours. It has a pony motor start, John Deere pony motors ARE the Devil.
Ron

To stay on topic as we should lets move any tractor discussion to the general discussion area, see you there.  We have a good forum here and we can keep it that way with very little effort. flywheel

[vdubnut62]

Sorry all. I didn't mean to hijack the forum. But.... you DID ask and I am sorta proud of it
Ron

[flywheel]

Sorry all. I didn't mean to hijack the forum. But.... you DID ask and I am sorta proud of it
Ron
[/quote

You are right - I did ask and you did reply.  I don't think anyone is upset and you did not hijack the forum.  I thought it would be best to move it to the general discussion area.   flywheel

[flywheel]

And now the answer. 
The winner is direct drive.  I looked at my written test results and I see that this testing was done in Sept. of 2003. It was done with a new engine with less than 20 hours on it at the time of testing. The Yanan brand 10kw st head used was manufactured in 2002.

This was not a scientific test and fuel consumption was not recorded.  It was done to see if there is really any difference between belt or direct drive.  I got a good education during this testing. 
A laser tach was used to check rpm and a amprobe clamp-on ammeter used to check amperage.
The were a total of six tests, 3 with belt and 3 with direct drive.  Each set of 3 tests was averaged.

The reason there is a different engine rpm for the belt test was the driven pulley on the head was slightly larger so the rpm had to be increased to produce 120 volts from the head. I could not find the same size pulley that was on the engine at that time.

At 1800 rpm direct drive the head would produce 76.25 amps or 9150 watts, the engine would just begin to produce very light black smoke. 

At 1985 rpm with belt drive the engine would produce 74.5 amps or 8940 watts when the engine would start to produce very light black smoke.

There is a difference of 210 watts, the engine with belt drive is operating at 185 more rpm and produces less so the difference is even more with direct drive. 

I have more to add to this but the answer is clearly more watts with direct drive.   flywheel

[mobile_bob]

your test results confirm what mine have been as it relates to belt losses
in my testing i came out to ~200 watts in losses, which i find quite acceptable
in driving a pair or alternators.

of course that is 200 watts at ~3.5kwatts load on the alternators, so it is a larger
percentage of the total than you have with your setup.

my testing was done in BSFC rather than total loading, because the twin alternator cannot
be driven directly.

interesting how close the numbers are though huh?

bob g

[cognos]

So - in the original case, the difference is about 2% (8940 vs. 9150), In Bob's case, the difference is 6% (I'll guess 3300 vs. 3500) - even though it's still 200 watts, thats a 200% variance between the two cases...

A 6% variance in flywheel's test would be 9150 vs. 8601. 

And of course the variable stating fuel consumed/time is missing in both cases, which would be determinate. I think...

(I'd say 2% can be explained away to 0, just with testing variables. Just to be contentious...       )

[Wizard]

200W is significant.

Cheers, Wizard

[TimSR2]

I think the very small pulleys contributed to the high slip in the belt setup. Larger diameter pulleys would have improved the belt drives results .
Interesting test, thanks for that!

Tim

[mobile_bob]

i went back and double checked my notes and i did not seperate alternator windage from
the drive losses because i have no really good way to do it.

also i found notations that i had made mathematically referenceing my belief at the time that
the belt drive was consuming approx 2% of my power

my methodology was based on BSFC numbers and i would strongly disagree that 2% is within the margins of error

i can clearly measure even a single 60 watt bulb being thrown onto my alternator or ST head during testing
because of the increase in fuel consumtion in grams

and i know the numbers are good because the results repeat with astounding accuracy, to that point that even
my then 15 yo daughter could predict mathematically the results of adding any amount of load (60 watts or above)
in added fuel consumption down to nearly a gram for light loads and a bit more in high load situations.

at first i could not do that because my metering was just not able to measure that accurately and do it consistantly
it was not until i got the GE kw/meter and set it up to test mode where it measures the st power right down to watt/hrs
(+/- 2 watt/hr accuracy) that i was able to get very consistant test results.

before the GE meter i would run 3 tests at each level and average the results, after getting the GE meter the results were
spot on in all three tests at each level, and would repeat on different testing days so long as the shop ambient temp was within
5 degree's of one another,,, if however i would test at 50 degree's ambient and then retest at 90 degree's (summer) the results
while being highly consistant  clearly illustrated the effects of ambeint temperature on air density and resultant power.

too cool

bob g