---
title: Blade vs Burr
url: https://oosioo.com/en/p/%EB%B8%94%EB%A0%88%EC%9D%B4%EB%93%9C-vs-%EB%B2%84
date: 2026-11-10T13:26:28+00:00
author: SYSOP
summary: No peer-reviewed study backs up "buy a burr grinder, not a blade". What exists is one blade grinder tested at different grind times and sieve-based product tests (27–44% in the target band versus over 60% for burrs). A population balance model shows a blade grinder's dial is time, and with no gate at the exit its distribution is wide both ways even at the same median (D10 138, D90 2,941 µm). Shaking trims the coarse tail. With a calculator.
---
# Blade vs Burr

Probably the first advice anyone gets on starting to grind their own coffee is "buy a burr grinder, not a blade grinder". The reason given is that blade grinders grind unevenly and burn the coffee. Yet I could not find a peer-reviewed study backing this up: no paper compares the two side by side under the same conditions.

I worked through what data there is, plus a model. The short version: a blade grinder's grind dial is time. With no gate at the exit, particle size is set by how many times pieces get hit. So even when the median size matches a burr grinder's, the distribution is far wider on both sides. Shaking while grinding trims the coarse tail. How much the taste differs, nobody has properly measured.

## The measurements there are

![An electric blade grinder seen from above. Roasted beans sit in a silver cup inside a white body, with a single silver blade stretching across the middle. A glass French press, a cup of coffee and a spoon are beside it](/uploads/a2bc8b9e1dfcda03.webp =520)
*The blade of a blade grinder* · Photo: OKJaguar, Wikimedia Commons, CC BY-SA 4.0

The only peer-reviewed blade-grinder measurement I found was by Murray and Laredo (2014). They ground 37 g of beans in a single Black & Decker blade grinder for 1.5 to 84 seconds and measured size with eight sieves (9.5 mm to 75 µm). Beans that started at 6–8 mm were mostly 75–300 µm after 84 seconds, with 4% below 75 µm. The longer the grind, the further the distribution moved towards small sizes and the narrower it became, and the authors concluded there is a lower limit to how fine this method can grind. Caffeine in the brewed coffee rose with grind time up to 42 seconds, where it was about 3.5 times the 1.5-second value. No burr grinder was included.

![Horizontal bars. Seven blade grinders in orange (Krups Coffee and Spice 44%, Cuisinart Spice and Nut 41%, Capresso Cool Grind 39%, Secura 32%, Cuisinart Grind Central 29%, Krups GX336D50 27%, KitchenAid 27%) and, at the top in blue, 'Top burr' (Baratza Encore ESP 64%). A blue dotted line at 60% reads 'burr grinders averaged over 60%'. The horizontal axis is mass in the target size band (%), and the title is 'Seven blade grinders, 15 s grind, Kruve sieves (America's Test Kitchen)'](/uploads/9f8e2d2063c52299.webp =760)

The side-by-side data comes from product tests by the US cooking outlet America's Test Kitchen. They ran seven blade grinders for 15 seconds each and sorted the grounds with Kruve sieves into boulders, medium and fines. The share in the target medium range was 27–44%. In the same outlet's burr grinder test, the average was over 60%, with the winner (Baratza Encore ESP) at 64%. The sieve sizes were not given, and these tests are not peer reviewed.

## Blades strike; burrs sieve

In a burr grinder, the gap works as a gate at the exit. Pieces larger than the gap cannot leave and are broken again ("The Path of a Bean Through a Flat Burr"). A blade grinder has no exit and no screen. Beans and pieces stay trapped in the cup until the motor stops, broken at random by a fast-spinning blade. Their size depends on how many times, and how hard, they were hit.

The blade tip is very fast. Home blade grinders don't publish their speed, but around 20,000 rpm is often mentioned. With a 3 cm blade radius, that is a tip speed of 63 m per second. A 64 mm flat burr at 1,400 rpm moves 4.7 m per second at its edge, more than ten times slower. The blade is in the same range as the impact mills (pin mills, 21–87 m/s) seen in "The RPM Mystery". A blade grinder is closer to a hammer that strikes the beans than a millstone that grinds them.

Process engineers model mills like this with a "population balance". The grounds are split into size classes, and for each class you set how often per second it is hit by the blade, the chance it breaks when hit, and which classes the fragments go to, then move mass between classes over time. For the breakage chance I borrowed the Vogel–Peukert-type formula used in "The RPM Mystery". Depending on where along the blade a piece is hit, the impact speed is spread evenly from zero to 60 m/s. Small pieces rarely break even at the same speed, so hardly anything goes much below about 300 µm. Without shaking, fines are assumed to stick to the walls and lid, out of the blade's way, and large pieces to bounce up over the blade and get hit less. Shaking brings both back to the blade. These values are plausible assumptions, not fitted to coffee.

![Two charts. On the left, mass distributions (peak = 100) while shaking at grind times of 2, 5, 10, 15 and 20 s. At 2 and 5 s there is a tall peak at the 6.5 mm whole-bean size; at 10 s a broad hill spans 300 µm to 3 mm; at 15 and 20 s a peak forms around 330 µm and the coarse tail shrinks. On the right, grind time (0–30 s) across and size (log) up, with D50 (solid) and D90 (dashed) stepping down over time for shaking (orange) and still (grey). The shaking D50 meets the blue dotted line 'burr D50 (gap 1.6 mm)' at 12 s and the still D50 around 20 s. The still D90 is still above 2 mm at 30 s](/uploads/a65a66fd114ee0dc.webp =760)

Shaking, D50 fell to 906 µm at 10 seconds and 447 µm at 15. A blade grinder's grind dial is time. As Murray and Laredo saw, the longer you grind, the more the distribution piles up against its lower limit and narrows. Without shaking, it takes much longer to reach the same size, and large pieces dodge the blade and survive.

## At the same median size

To compare with a burr grinder, I matched the median size (D50). The 3D virtual flat burr from "Do Frozen Beans Grind Finer?" (gap 1.6 mm) had a D50 of 660 µm. The blade model reached a D50 of 636 µm after 12 seconds shaking, or 20.5 seconds still.

![A chart with particle size (log) across and mass density (peak = 100) up. The blue line (burr, 3D virtual flat, gap 1.6 mm, D50 660 µm, above 400 µm only) rises in a narrow tall peak just below 1 mm, with almost nothing above 1.2 mm. The orange line (blade, shaking, 12 s, D50 636 µm) is a broad hill peaking around 330 µm and sloping gently down to 6 mm. The grey line (blade, still, 20.5 s, D50 636 µm) has a similar hill plus a second tall peak at the 6.5 mm whole-bean size](/uploads/c65564bdf7c43e53.webp =760)

With the same median, the shapes are completely different. The burr's grounds were packed narrowly just below the gap (D90 1,003 µm). The blade grinder, even shaken, had a D10 of 138 µm and a D90 of 2,941 µm, with 19% of the mass above 2 mm and 6.4% below 100 µm. Without shaking, D90 reached 5,298 µm and the share above 2 mm rose to 26%, meaning beans that were barely ground at all. The same cup holds fines that will over-extract and chunks that will hardly extract. If a retailer's line that "bitter and sour show up in the same cup" is true, this would be why.

Taste has not been measured yet. America's Test Kitchen wrote that most of their experts couldn't tell burr-ground from blade-ground coffee, but didn't say whether the tasting was blind. I also found no temperature data behind the claim that blade grinders "burn" coffee. How much beans actually warm during short bursts is a question a thermal network can answer ("How Hot Does a Grinder Get?").

## Grind it yourself

The calculator below runs the model above. Change the grind time and blade tip speed, and choose whether to shake, and it draws the blade grinder's distribution over the burr's, with D10, D50, D90 and the shares below 100 µm and above 2 mm.

```map /embed/coffee-blade-2026-en.html 16:9
A blade grinder calculator. On the right, sliders set grind time (1–30 s) and blade tip speed (20–90 m/s), and a 'Shaking' button chooses whether to shake while grinding. The chart on the left overlays the blade grinder's mass distribution from a population balance model (orange) and the 3D virtual flat burr's distribution (gap 1.6 mm, above 400 µm only, blue), each scaled to its own peak. The table on the right gives D10, D50, D90, the share below 100 µm and the share above 2 mm.
```

## What this conclusion leaves out

- **The blade model's values are assumptions.** Hit rate, breakage chance, fragment distribution and the effect of shaking were not fitted to coffee measurements. The shape of the distribution (wide on both sides) follows from the lack of an exit, but the numbers depend on the assumptions
- **The lower limit near 300 µm comes from the model's breakage formula.** It points the same way as the "lower limit" Murray and Laredo saw, but its position was not fitted
- **The burr side ignores everything below 400 µm,** because of the 3D model's resolution
- **There are no measurements under the same conditions.** America's Test Kitchen's sieve tests did not publish sieve sizes or test conditions
- **Taste and temperature were not measured**

## Summary

- I found no peer-reviewed study comparing blade and burr grinders under the same conditions. What exists is a study varying only grind time on one blade grinder (Murray and Laredo 2014) and product tests (27–44% in the target size band on sieves, versus over 60% for burrs)
- A blade grinder's dial is time. With no gate at the exit, size is set by how many times pieces are hit. The blade tip moves at around 60 m/s, more than ten times faster than a burr's edge
- Even with the median matched to a burr (D50 636 versus 660 µm), the blade grinder spread widely both ways: D10 138 µm, D90 2,941 µm. Without shaking, D90 passed 5 mm, leaving barely ground beans
- In the model too, shaking trimmed the coarse tail. Differences in taste and the "burning" claim have not been measured

## References

- Murray J, Laredo T (2014). Effect of home grinding on properties of brewed coffee. *Journal of Food Research* 4(1), 77. [doi:10.5539/jfr.v4n1p77](https://doi.org/10.5539/jfr.v4n1p77)
- Bell LN, Wetzel CR, Grand AN (1996). Caffeine content in coffee as influenced by grinding and brewing techniques. *Food Research International* 29(8), 785–789. [doi:10.1016/S0963-9969(97)00002-1](https://doi.org/10.1016/S0963-9969(97)00002-1) — abstract only
- America's Test Kitchen. Blade-style coffee grinders. [link](https://www.americastestkitchen.com/equipment_reviews/1969-blade-style-coffee-grinders); The best burr coffee grinders. [link](https://www.americastestkitchen.com/equipment_reviews/2752-the-best-burr-coffee-grinders)
- Vogel L, Peukert W (2003). Breakage behaviour of different materials — construction of a mastercurve for the breakage probability. *Powder Technology* 129, 101–110. [doi:10.1016/S0032-5910(02)00217-6](https://doi.org/10.1016/S0032-5910(02)00217-6) — abstract only

## Image credits

- Blade grinder: OKJaguar, [Wikimedia Commons](https://commons.wikimedia.org/wiki/File:Coffee_grinder_blades.jpg), CC BY-SA 4.0
- All other charts, the cover and the calculator: made by me

