Grain Drill Calibration Calculator

Why a drill needs calibrating at all

A grain drill does not meter seed by weight. It meters by volume — a fluted roll or a cup wheel turning at a speed geared to the drive wheel, dropping a certain volume of seed per revolution. The rate chart riveted to the lid was produced by the manufacturer with one particular seed, at one particular test weight, through a meter that was new. Everything that has happened to your drill since then has moved the number.

Seed size and test weight move it most. Plump wheat and shriveled wheat fill the same volume at different weights, so the same setting puts down a different poundage. Coated seed is bulkier than raw seed of the same pure-live-seed content — Georgia's B 1510 puts the added weight or volume at "usually 20-50%." Beyond the seed, worn flutes pass more volume than new ones, a slipping drive chain under-turns the meter, and a partly plugged tube passes less than its neighbors.

The manufacturers say so in their own rate books. Great Plains prints that "the rates indicated in the charts are approximate," that they are "based on cleaned, untreated seed of average size and test weight" with "factory-supplied new tires... inflated to factory specification," and that "many factors affect seeding rates including foreign material, seed treatment, seed size, seed weight, field conditions, tire pressure and test weight." The chart is a starting point, described as one by the people who printed it.

None of that shows up while you are drilling. You find out at emergence, and by then the seed is in the ground and the money is spent.

The arithmetic, written out

There is only one equation on this page and it is worth seeing in pieces, because every mistake people make is a mistake in one of the pieces.

• Working width (ft) = opener spacing in inches × number of openers ÷ 12. A 24-opener drill on 7.5-inch spacing is 24 × 7.5 ÷ 12 = 15.0 ft. This is the width the drill seeds, which is not the width of the machine and not the distance between the tires.
• Acres covered = working width × test distance ÷ 43,560. There are 43,560 square feet in an acre; that is the only constant in the job.
• Seeding rate (lb/acre) = pounds caught ÷ acres covered.
• If you sampled some tubes and not all of them, scale the catch first: pounds caught × total openers ÷ openers sampled. The calculator above does this before it divides, and tells you it did.

Penn State Extension's published example is the sanity check on all of it: 2.3 pounds caught from a 10-foot drill over 200 feet covers 0.046 acre, and "2.3 pounds collected represents a seeding rate of 50 pounds per acre (2.3/0.046 = 50)." That example is wired into this site's automated test suite, so the calculator is checked against the extension service's published answer rather than against itself.

Penn State is also honest about where the 200 feet comes from: "this is an arbitrary length, but it is long enough to capture variability in the field yet not too long to be impractical." A longer run averages more of the drill's inconsistency into the answer; a shorter one is quicker to repeat.

The field procedure, step by step

The version to follow standing at the drill, from Penn State's procedure and University of Georgia Extension's B 1510.

1. Do the maintenance first. A catch test on a drill with a plugged tube measures the plug, not the setting. Check the boxes for trash and bridging, that flutes and cups move freely, and the drive chain and sprockets.
2. Load the seed you are actually going to plant. Not last year's, not a different lot. Penn State's instruction is that "seed cups are full and seed metering fluted wheels or sponges are completely covered."
3. Prime the meters. B 1510 says to drive forward 100 to 200 feet before you start catching, so the cups are running full and settled.
4. Set the drill to your intended rate off the chart on the lid. That is what you are testing.
5. Rig the catch. Pull the drop tubes and cable-tie bags to the tube ends. Georgia's supply list: a scale reading to 0.1 gram, flags, a tape or measuring wheel, containers, cable ties. Tare the bags — an empty feed bag weighs enough to change a small-seed answer.
6. Stake the run on ground that looks like the field you will plant, not the yard.
7. Drive it at planting speed, from a rolling start so the meter is already turning at the first flag.
8. Weigh what you caught, subtract the tare, and put the numbers in the calculator above.
9. Repeat it. Penn State: "for increased accuracy, repeat this one or two times and use the average."
10. Adjust and re-run until it lands. AEN-81: "continue to calibrate and adjust the metering cup setting until you are within 5 percent of the target seeding rate."

How many tubes? Kentucky Extension's guidance is "from 3 to 5 drop tubes across each 10 to 15 foot section of the drill," and AEN-81's own procedure uses five openers over a 200-foot course. Catching every tube is better arithmetic and much more work; a spread across the whole width is the practical compromise, and it is why this calculator has an "openers sampled" box.

The gram shortcut, and why it works

B 1510 contains a trick that removes the arithmetic entirely. Catch from one row unit over the distance in the table below, weigh the catch in grams, and: "the grams of seed collected from traveling the calibration distance equals the pounds of seed being sown on a per-acre basis from that row unit." Twelve grams means twelve pounds per acre.

Distances as published in University of Georgia Extension B 1510, Table 2.

It is worth knowing why it works, because a shortcut you cannot explain is one you will misuse. A pound is 453.6 grams, and an acre ÷ 453.6 = 96 square feet. So if one row unit covers 96 square feet, one gram caught is a 453.6th of a pound over a 453.6th of an acre — exactly one pound per acre. At 7.5-inch spacing a row unit is 0.625 ft wide: 96 ÷ 0.625 = 153.6 ft, the 154 in the table. At 9 inches, 96 ÷ 0.75 = 128. At 15 inches, 96 ÷ 1.25 = 76.8, rounded to 77. Every row is the same division. The derivation is ours; the distances are Georgia's.

Two limits. It holds for a single row unit only — catch two tubes and you must halve the grams. And a scale reading whole grams is too coarse for small seed, which is why B 1510 asks for one reading to a tenth.

Turning the wheel instead of driving the drill

You do not have to move the machine. Penn State describes stationary calibration as "simulating drill operation": jack the drive wheel clear, then "turn the drive wheel the number of revolutions needed to represent 200 feet of travel." Find your drive wheel first — Penn State's step one is to "determine which wheel is responsible for driving the seeding mechanism on the drill," and on many drills only one of them is. There are two published ways to get the count, and they agree.

From the distance. University of Maryland Extension writes it as "100 / (3.14 X diameter of the drive wheel in feet) = Number of revolutions," with the worked case "100 ÷ (3.14 X 3.5' diameter) = 9.09 revolutions." For a 200-foot run, use 200 on top. Rather than measuring a diameter, you can roll the wheel one full turn between two marks and measure that — circumference is what the formula is after, and rolling it measures the wheel you have rather than the one on the spec sheet.

From the area. Mississippi State P3112 works straight from the acre fraction: "wheel turns = (43,560 sq ft per acre x 0.10 acre) divided by (6.6 ft x 8 ft) = 82.5 turns," with circumference from "2.1 ft x 3.1416 = 6.6 ft." That is a tenth of an acre on an 8-foot drill. P3112 lists the common fractions as 1/10, 1/25, 1/50 and 1/100 acre — a hundredth is 436 square feet, a 55-foot run on an 8-foot drill, short enough to do in a shed.

Some manufacturers publish a fixed rotation count for their own machine; use theirs rather than deriving one. Great Plains publishes 45.1 crank rotations per acre on one heavy-duty drill, and on a pair of six-foot no-till drills 913 rotations by crank, 1,628 by tire, and 995 on the neighboring models. Those come out of specific gear ratios. They do not transfer between machines.

A full worked example you can check

A 24-opener drill on 7.5-inch spacing, drilling wheat, chart set for 90 lb/acre. You catch from 6 tubes spread across the width, over a 200-foot run at planting speed.

1. Working width: 7.5 × 24 ÷ 12 = 15.0 ft.
2. Acres covered: 15.0 × 200 ÷ 43,560 = 3,000 ÷ 43,560 = 0.0689 acre.
3. Catch from the 6 bags, tared: 0.30, 0.32, 0.31, 0.22, 0.34 and 0.37 lb — 1.86 lb total.
4. Scale to the whole drill: 1.86 × 24 ÷ 6 = 7.44 lb.
5. Rate: 7.44 ÷ 0.0689 = 108 lb/acre.

The chart said 90 and the drill is doing 108 — 20 percent heavy. On 100 acres that is 1,800 extra pounds of seed wheat, or about 30 bushels, bought and drilled for nothing.

Now look at the six bags again rather than the total, because they are telling a second story. The average of the six is 1.86 ÷ 6 = 0.31 lb, so Penn State's 10 percent band runs from 0.279 to 0.341 lb. Four tubes sit inside it. The 0.22 tube is 29 percent light and the 0.37 tube is 19 percent heavy — both fail the check, and both would have been invisible in the 108 lb/acre average. The 0.34 tube, at 9.7 percent over, scrapes inside; that is what a threshold looks like from up close.

What you do next is not one calculation. The ratio 90 ÷ 108 = 0.83 tells you which way to move the setting and roughly how far, but no source we located publishes a linear relationship between a meter setting and its output, and on a stepped or geared adjustment there may not be one. The published instruction is to move the setting, catch again, and keep going until you are inside 5 percent of target.

Check the openers individually

A whole-drill catch tells you the average. It does not tell you that opener four is delivering half rate and opener five is making up for it. Catch several tubes separately, weigh them, and compare each to the average of the group. Penn State's threshold is explicit — "an amount collected from one opener that is 10 percent different from the average on all openers indicates a need for maintenance to the seed-metering mechanism." Usually it is a plugged tube, a damaged flute, a cup out of adjustment, or trash bridged over one opening.

This is not a rare fault. AEN-81 warns of "as much as 15 percent between seed cups across the drill," and Kentucky Extension has reported metering units varying "by more than 10 percent above and below the target-seeding rate, which affects seed costs proportionately." A drill can hit its average perfectly while half its rows are wrong in opposite directions — invisible until the stand comes up in alternating thin and thick rows.

Pounds per acre and seeds per acre are different instructions

A drill setting is a volume. A bag tag is a weight. A plant stand is a count. Those three are related only through the size of the seed in front of you, and seed size moves more than most people expect.

University of Nebraska Extension: "the number of winter wheat seeds in one pound can range from more than 20,000 to less than 10,000, depending on the variety and the year it was produced." In one year's variety evaluation, "seed size ranged from a low of 11,860 seeds per pound to a high of 24,790 seeds per pound" — which Nebraska calls "a 209% difference in seeding rate" if you seed by weight. Kentucky's AEN-81 agrees from the engineering side: "the relationship between the number of seeds discharged and weight varies greatly among varieties and among lots of the same variety," with a trial table running 9,700 to 17,089 seeds per pound.

What that does to the arithmetic: pounds per acre = seeds per acre ÷ seeds per pound. Take a round illustrative 1,200,000 seeds per acre — chosen to show the mechanics, not a recommendation, since rate targets are regional and belong to your extension service:

• At 9,700 seeds/lb (Kentucky's largest-seeded variety): 1,200,000 ÷ 9,700 = 124 lb/acre.
• At 15,200 seeds/lb (Nebraska's suggested minimum, a 30-gram thousand-kernel weight): 79 lb/acre.
• At 24,790 seeds/lb (Nebraska's smallest measured lot): 48 lb/acre.

Same field, same target stand, and the right answer in pounds ranges from 48 to 124. No chart on a drill lid can know which end you are at. Nebraska's summary line is the one to keep: "a simple shift from pounds to seeds per acre can make the difference between underseeding and overspending."

To work in seed counts, AEN-81's method is short: "to estimate seed density, count out 1,000 seeds and weigh them to the nearest 0.1 gram," then convert. It folds germination and purity in at the front, taking a 375 plants per square yard target up to 419 seeds per square yard by dividing by 0.90 germination × 0.995 purity.

Small-seeded species make the point unmissable. UGA B 1251 reports "on average there are more than 750,000 white clover seeds per pound" — some fifty times wheat. That is why a clover rate of 2 to 3 pounds an acre is real and not a typo, and why an error of one pound there is a catastrophe where the same error on wheat is a rounding difference.

The small seed box is a second machine

Most drills that plant forage carry two boxes, and the legume or small seed box meters independently of the main box. B 1510 is explicit: "one should calibrate both the large and small hopper boxes independently to assure both are planting the appropriate seeding rates." A perfect main-box catch tells you nothing about the clover.

They are usually adjusted differently too, which is a second reason not to reason from one to the other: on some Great Plains drills the small seeds attachment "is controlled entirely by the seed rate handle" with "no sprocket changes required," while the main box uses sprockets and a handle together. The small box is also where the arithmetic is least forgiving, because the rates are smallest — catching a couple of ounces needs a gram scale and more tubes to get a weighable sample. Calibrate both, and calibrate the small box the more carefully of the two.

Coated seed, treated seed, and what actually changes the rate

These get lumped together and they should not be. B 1510 separates them. Coatings: "seed coatings are likely to influence the rate the seed is metered by a drill and can add a considerable amount of weight or volume to the seed (usually 20-50%)," and "by adjusting the seeding rate for PLS, one will account for the seed coating." Treatments: "seed treatments often include insecticides or fungicides... but usually have no significant effect on seed weight or flow characteristics."

The sources do not agree on that second point, and we are not going to paper over it. Penn State lists "seed coatings or seed treatments" together among the variables affecting seeding rate; P3112 lists "seed coatings or treatments" the same way; Great Plains lists "seed treatment" among its factors. Georgia says treatments usually do not matter much. All four are credible and none publishes a magnitude. The practical resolution is the same either way: if what you are drilling is treated or coated and what you calibrated with was not, calibrate again.

Where this goes wrong

Almost every bad calibration is one of these, and none of them announces itself.

• You measured the machine, not the seeded width. Working width is spacing × openers. A 15-foot drill with two plugged end rows is seeding 13.75 feet, and your rate is wrong by that ratio.
• You forgot the tare. Bags, cups and cable ties all weigh something. On a legume box that error can exceed the sample.
• You caught from a cold meter. Cups not full and running at the first flag deliver light. Prime the drill, use a rolling start.
• You calibrated on the lane and planted in the field. Great Plains lists "field conditions" and "tire pressure" among the factors that move a rate, and its charts assume "factory-supplied new tires... inflated to factory specification." We could not find a land-grant or manufacturer publication that quantifies drive-wheel slip as a percentage of seeding rate, and we are not going to invent one. What the sources support is the instruction: calibrate on ground like the ground you will plant, and confirm in the field.
• You changed speed after calibrating. P3112 lists "variations in ground speed" first among the causes of chart-versus-actual discrepancy. No source we located publishes how much a given speed change moves a given meter, so the honest instruction is to re-check, not to correct.
• You ran the test once. One catch is a data point with no error bar. "Repeat this one or two times and use the average" is the cheapest accuracy on this page.
• You only looked at the total. The average can be right while the rows are wrong.
• You calibrated the main box and drilled a mix. Two boxes, two calibrations.

What a miss actually costs

The worked example above cost 30 bushels of seed wheat on 100 acres. Running light is worse, because it cannot be undone: a thin stand never closes the canopy, gives weeds the room they were waiting for, and is beyond fixing after emergence at any price.

Small-seeded forages are less forgiving still, because the rates are small enough that a modest absolute error is a large proportional one. On a clover seeded at 4 pounds per acre, being off by 2 pounds is a 50 percent error — and at Georgia's figure of more than 750,000 white clover seeds per pound, that is a difference of a million and a half seeds an acre.

When to re-check

Calibrate at the start of every planting season, and again any time you change seed lot, species, variety, or seed treatment or coating. Re-check if the ground conditions are noticeably different from where you calibrated — B 1510's instruction is to "confirm that seeding rates remain on target under actual operating conditions." If you change ground speed substantially from what you calibrated at, check again; P3112 lists ground speed among the causes of chart-versus-actual discrepancy, though no source we found publishes how large the effect is on a given meter.

Keep the numbers. A short note of seed lot, setting, catch weight, and resulting rate, kept year over year, turns into the most useful record on the drill: it tells you when the meter is wearing, because the same setting with the same seed starts drifting.

What this calculator does not know

It does one division. What it does not do is worth stating plainly.

• It cannot tell you your target rate. Rates are regional, species-specific and date-specific; they belong to your extension service and your seed tag. This page tells you what the drill is doing, not what it ought to be doing.
• It does not know your drive wheel slipped. It takes the distance you type as the distance the meter was driven for. If the wheel spun, the two differ and nothing in the arithmetic can see it.
• It assumes every opener you did not sample behaves like the ones you did. That is what scaling by openers-sampled means, and it is exactly what hides a plugged row — hence the per-opener check.
• It works in bulk pounds, not pure live seed and not seed counts. Convert before you calibrate, using the tag.
• It cannot convert one meter setting to another. We found no published linear relationship between a setting and its output, so the answer to "20 percent heavy" is to move the setting and catch again, not to multiply.

Once the drill is putting down what you intend, the companion question is what to put down on the manure side — see spreader calibration for the tarp-test version of the same idea.

Frequently asked questions

How do you calibrate a grain drill? Catch what the drill delivers over a measured run, weigh it, divide by the acres covered. Width = spacing × openers ÷ 12; acres = width × distance ÷ 43,560; rate = pounds ÷ acres. Penn State's example: 2.3 lb from a 10-ft drill over 200 ft = 0.046 acre = 50 lb/acre.

How many revolutions of the drive wheel equal the test distance? Distance ÷ circumference. Maryland writes it "100 / (3.14 X diameter of the drive wheel in feet)" — 9.09 revolutions for a 3.5-ft wheel over 100 ft. P3112 works from area and gets 82.5 turns for a tenth of an acre on an 8-ft drill. Some manufacturers publish a fixed count for their own machine; use theirs, and do not carry it to another drill.

How far off is the seeding chart on the drill? Nobody publishes one number. Penn State says charts "may not be very accurate." P3112 says discrepancies "can range from 25 to 50 percent." Kentucky Extension reported tables varying "by 10 percent or more from measured values." Great Plains calls its own charts "approximate." The spread is the answer: measure your own.

Why does the same setting put out a different rate with a different seed lot? A drill meters volume, not weight. Nebraska reports winter wheat from 11,860 to 24,790 seeds per pound in one year's evaluation. AEN-81: "the relationship between the number of seeds discharged and weight varies greatly among varieties and among lots of the same variety."

Do I have to calibrate the small seed box separately? Yes. B 1510: "one should calibrate both the large and small hopper boxes independently to assure both are planting the appropriate seeding rates."

How much should openers vary from each other? Penn State's threshold is 10 percent from the group average. AEN-81 notes "as much as 15 percent between seed cups across the drill" is possible.

What do I do when the catch disagrees with the chart? Move the setting and catch again. AEN-81 gives it as an iteration: "continue to calibrate and adjust the metering cup setting until you are within 5 percent of the target seeding rate."

Sources

• Penn State Extension, "Calibration of Grain/Seed Drills" (Hoover, Duiker) — 200-ft method, the 2.3 lb / 50 lb-per-acre example, the 10 percent per-opener threshold.
• Mississippi State Extension P3112, "Reduce Seed Cost by Properly Calibrating Your Drill" (Lemus) — 25-to-50-percent chart discrepancy, acre fractions, wheel-turns-from-area.
• Univ. of Georgia Extension B 1510, "Preparing and Calibrating a No-Till or Conventional Drill..." (Baxter, Maddy) — gram shortcut and distance table, both-boxes rule, PLS, coatings and treatments.
• Univ. of Kentucky Extension AEN-81, "Grain Drill Calibration Procedures for Winter Wheat" (Shearer, McNeill, Watkins, Higgins) — 5 percent target, 15 percent cup-to-cup variation, thousand-seed weighing.
• Univ. of Maryland Extension, Maryland Agronomy News, 3 May 2019 (Morris) — revolutions-from-diameter formula.
• Univ. of Nebraska–Lincoln Extension / CropWatch (Klein) — wheat seeds-per-pound ranges, thousand-kernel weight.
• Univ. of Georgia Extension B 1251, "White Clover Establishment and Management Guide" — white clover seeds per pound.
• Great Plains Mfg. seed rate manuals 151-122B and 196-522B — chart accuracy, tire condition, rotation counts, small seeds attachment.