If nobody can answer, put a measured amount of fertilizer into it and see how long it takes to run out over a given area. That will give you a rough pounds per thousand square feet conversion for a given setting.

You could theoretically test that with something innocuous like sand, but the particle size does impact the flow...

Thanks morpheus!!! I was hoping to avoid that option as this spreader has settings from "A" through to "S".

I also know that each individual spreader does need to be calibrated like this but was hoping for a mapping that I can use as a starting point for each of the 19 or so settings. Thanks Again!!!

Fortunately, there should be a linear regression we can use for this.

And since I've recently been quite open about my neurological issues, I can now be quite open about the meds, which regrettably turn me into a clumsy, affable idiot. Think Gilligan. Alas. My ability to do this in my head is somewhat limited, so bear with me.

You essentially have a function with 19 solvable roots. It's annoying to solve each one individually, and we're hopeful that there's an easier solution. Nolo problemo, this is a common issue.

#1: Call Scott's. They probably have this available and it saves working the actual math.

:-)

#2:

a) Measure the output and pounds per thousand square foot for measurements A (1), E (5), J (10), O (15), and S (19).

b) On a piece of graph p...heck with that, in Excel, in Column A, enter 1, 5, 10, 15, and 19. In Column B, enter the pounds per thousand for settings A, E, J, O, and S respectively.

c) Graph that as an XY scatter and run the line in there.

d) Hope it's linear and that said line is reasonably close to the points. You can chart the equation, which is always very nice and you'll need this later, but we're more interested in the deviation off the curve. We're looking for a pretty low deviation here.

e) If it is, take note of that equation, which will look something like:

aX + b = c, where X is your setting (1 would be A, 3 would be C, etc.), a will be what you multiply by X, b will be a constant that the linear equation will generate for you, and c will be the answer you'll get that will be the pounds per thousand square feet in output for that setting.

f) Accuracy...has to be tested! It's not impossible that (for instance), setting R and setting S are wider open and exceptions to some sort of rule that otherwise governs the other settings. It's unusual, but hardly unheard-of. :-)

I use estimates like this off the top of my head for lots of stuff in an engineering and safety environment. Drives my boss nuts as it works far more often than his "seat of the pants" estimates and can be done in your head with some practice and knowing just a few data points for both linear and parabolic equations.