sharp/dull blade drawing The QX3 Microscope small map
Finest abrasives.
Microbevels front and back.
Use a jig.
Copyright (c) 2002-15, Brent Beach


All pictures are taken using an Intel QX3 computer microscope. Images from the microscope are displayed on the monitor and saved as JPG files. To determine actual magnifications, I took a picture of a ruler and determined the width of 1/32" in pixels. This ratio allows me to convert pixels in any other picture into decimal fractions of an inch. A picture taken at 200 times magnification has a pixel to inch ratio of 14,080 in images before February 2005, a pixel to inch ratio of 13,120 since then.

Any microscope that uses visible light to illuminate the object is restricted to magnification of about 200 times. In order to see a feature of the object, the feature must scatter the visible light. In order to scatter visible light, the feature must be at least as wide as the wavelength of the light. This is about what 200 times magnification can produce.

There are lots of resources on the Net that discuss the QX3 microscope. Check out Goggle on QX3.

Topics in this page

Typical Images

The microscope can take pictures at three magnifications: 10X, 60X, and 200X.


Here is the bevel of a scrub iron magnified 10 times.

This bevel is about 1/6" across.

This bevel was produced using a 1" belt sander and a 120 grit belt. The iron is a laminated Stanley iron, with the hard/soft metal junction clearly visible.


This is a 60 times magnification of an Australian made high speed steel plane blade.

This image shows two regions. The lower region is the factory grind. They appear to have used two sharpening techniques - the underlying grind is perpendicular to the edge. After that, they appear to use a somewhat finer (but not very fine) abrasive. Unfortunately, I have no image of the edge before my first microbevel.

The upper region is a microbevel produced using 15 micron abrasive. The scratches in this microbevel are smaller than those in the basic factory grind, perhaps even smaller than the secondary factory abrasive.


This is a typical 200 times magnification of a just sharpened blade designed to highlight the microbevels.

I use three microbevels after the primary bevel. Normally, the scratches on each microbevel are perpendicular to the edge, but I varied my sharpening technique in this case to emphasize the different microbevels.

When honing the bevel on the 15 micron paper, I twisted the jig to the line of travel. This produced skewed scratches on the bevel. In this picture the angle of the 15 micron scratch lines is shown by the blue line. Before the 5 micron microbevel these scratches extended all the way to the edge.

When honing the next microbevel on 5 micron paper, I twisted the jig the other way to the line of travel. This produced a separate set of skewed scratches. In this picture the angle of the 5 micron scratch lines is shown by the green line. Before the 0.5 micron microbevel these scratches extended all the way to the edge.

Finally when honing the last microbevel on 0.5 micron paper, I held the jig square to the line of travel. This produced scratches that would line up with the yellow line, except that the scratches are too fine to show up using a visible light microscope.


Why so Blue?

The microscope has a transparent blue plastic shroud around the sides and back of the lens area. This appears to be intended to protect the lens. Objects in focus are about 1/8" below the shroud.

The shroud works as a protection, but also works to give a blue colour to some pictures.

The microscope comes with a weak light, whose strength can be controlled from the computer. The light is not strong enough to bring out the details in dark metal objects, like plane irons. I supplement this with a small halogen light which I can position anywhere near the plane iron. I can vary its position in all three dimensions around the front side where there is no shroud. If I position it so it shines through the shroud, the picture turns blue.

Normally I avoid the blue lighting, but unfortunately on sunny days light from the window shines through the blue shroud from the side.


Image Acquisition

Software does make a difference.

The QX3 ships with a software package for access to the images. That package is necessary for certain features, like motion pictures, but not for image acquisition.

The image sensor is 352 by 288 pixels wide. The QX3 software processes that image, first stripping off the pixels at the edge, leaving 320 x 240 pixels. It then digitally magnifies the resulting image up to 512 x 384 pixels (achieving the desired magnification levels assuming their standard computer monitor configuration).

If you acquire the image using a graphics program like Irfanview (available free over the internet, google can find it), you capture the full 352 x 288 pixels. If the QX3 is installed and connected and you use the Irfanview File menu, then the Select TWAIN Source... command, the QX3 should appear in the window. Highlight it and click Select. To get an image, in the Irfanview File menu use the Acquire command.

These pictures were taken one after the other, elapsed time less than 1 minute, in exactly the same conditions. The Irfanview acquired picture includes more of the blade so is larger, but the magnification is the same. It seems a little sharper to me.

Starting in about mid February 2005, all images are acquired using Irfanview.

Using the QX3 software. qx3
Using Irfanview. irfanview

Driver Installation. For Windows after XP

The QX3 drivers were written for Windows XP. If you have an old XP machine, you can continue to use the same driver.

If you have a 32-BIT VERSION of Windows 7 or 8, you can still run the QX3. If you have a 64-bit system, it is time to upgrade to another microscope.

A full set of instructions was supplied to the QX3 group on Yahoo Groups by Mark Mehall.

Razor Blade Images

Razors are, after all, the standard for sharpness. Here are two pictures of blades from BIC disposable razors, one new, one used.

This is an unused razor blade, 200X magnification.

The blade is 0.004" thick. The bevel is 0.020" wide. In that width they manage to fit three microbevels! This geometry suggests an minimum included angle of just under 12 degrees. I cannot tell what angles the three microbevels are, so the final included angle could be much more. [Images of a more recent blade, taken from a three blade razor, suggest that the final include angle is actually over 35 degrees - see below.]

The full bevel is about 0.020" wide, with both sides of the blade appearing to be sharpened the same way.

The secondary bevel is about 0.009" wide, but appears to have been made using about the same sized abrasive as the first bevel. Both sets of scratches appear to be about the size produced by 15 micron abrasives (see next image).

The third microbevel is only 0.0009" wide and shows no scratches. It was probably done using an abrasive with grit under 1 micron. In my testing of honing compounds used on leather strops, I have not found any that produces this fine a finish. This leads me to suspect that BIC is using a very fine abrasive paper to finish their edges, or possibly using a sub-micron diamond honing compound.

bic new
This is not a razor blade - it is a plane blade sharpened to the edge using 15 micron abrasive. The scratches are similar in size to the scratches on the first and second microbevels in the above picture. 15 micron abrasive
This is a used blade. You can just see the changes in the edge - the imperfections that make a blade seem dull. The third microbevel also appears to have a few scratches on it, near the edge.

This is the side of the blade that rubs against the work (your face, the bevel that faces down). The other side showed no such scratches (BICs use only one side of the blade).

Comparing the new and used blades, it looks like the bevel on the used blade is a bit less than 0.0001" narrower than the bevel on the new blade. This is not a before and after test, these are different blades. If approximately correct though, it is interesting how little metal gets worn during use.

bic used
This is also a used blade, but of a more recent vintage three blade razor.

These new razors have the blades floating on springs, one spring at each end. Fixed blade razors pin the blade between plastic struts. Floating blade razors spot weld the blade to a L-shaped piece of metal which rests on the springs. The strut provides the stiffness the blade alone lacks.

In spite of being welded to a support, the new blade is actually thicker than the older blades - 0.006" rather than 0.004".

In this picture the edge is at the top. Below that, the primary bevel, then the body of the blade. The primary is about 0.01" wide and appears to have been honed with an abrasive grit between 5 and 15 micron. The microbevel right at the edge shows no scratches, so was honed with something under 1 micron grit.

Given the thickness and the bevel widths, the included angle at the edge is at least 35 degrees! The single blade razor geometry suggested a minimum 12 degree included angle. It seems to me that it is unlikely there has been a big change in the geometry of the cutting edge, so the actual included angle at the edge, even on the single blade razors, is probably somewhat more than 35 degrees, not the 12 degrees suggested by the earlier blade.

For those interested, there are several interesting web sites on multi-bladed razors, including one that discusses an item in a Mad Magazine in 1975 that touted a 76 blade razor.

three blade used

Knife Edges

I have taken a couple of pictures of new knife blades.

Plane blades cut very differently from knife blades (which tend to be more like saws with very small teeth), so the principle behind knife sharpening are very different.

For more photomicrographs of knife edges, the place to look is on any knife enthusiast forum.

I recently bought a new knife made by "J.A.Henckels" which was labelled "German Stainless Steel" and "MADE IN CHINA".

This is a picture of the edge at 60X magnification.

The blade is 0.015" thick just back of the bevel.

The knife face has uniform scratches angled slightly to the edge which give the knife the characteristic matte finish.

They appear to grind the bevels in a single operation - there is no evidence of microbevels. The grinding is quite uniform, with both sides of the knife having the same size bevel (a feature not shared by all the knives I have bought over the years). I suspect that the bevels were done with a machine holding the blade. Or, perhaps it was hand held by a Galoot who also sharpens plane irons and saws without jigs.

knife 1
60 X
This picture is at 200X magnification.

The bevel is ground uniformly, the scratches all pointing toward the handle (which is down in this picture). At this magnification, there appears to be a regular variation in the surface of the bevel - a slightly deeper scratch appears at regular intervals along the bevel. These scratches give the knife its tooth - its ability to cut soft objects. Given the angle of the scratches, this blade is sharpened to cut on the pull stroke.

The bevel is about 0.012" wide. Given the total thickness at the edge of the bevel is 0.015", the included angle at the edge is around 32 degrees from the mid-line, for a total edge included angle of 64 degrees. This is a much larger included angle than I would have expected. Leonard Lee recommends an included angle between 10 and 35 degrees.

This knife does not rely on a small included angle for its cutting ability. Rather its thinness and the [relatively] rough finish on the bevelled edge combine to produce the cutting action.

Depth of Focus

This picture shows the depth of focus of the QX3. Somewhat less than half the bevel is in focus, so the depth of focus is less than one-quarter of the thickness of the blade at the bevel, or less than 0.004".

knife 2
200 X
This knife is also a J.A.Henckels knife, in this case a bread knife, also labelled PROFESSIONAL "S" and ICE HARDENED. Both photos are of the blade as received, both at the 200X magnification setting.

The cutting edge is scalloped and bevelled on the right hand side when in use. The blade is tapered from the back to the cutting edge (0.040" just above the scallops). The spine also tapers from the handle (0.077") to the tip (0.062")

This is the front side of the blade, showing the finish on the scallops at the edge. It appears to me that the scallops were created using a relatively coarse abrasive, then refined using a much finer abrasive at nearly the same honing angle. The darker area at the edge suggests a very fine abrasive.

knife 3
200 X, scalloped side
This is the back side of the blade, which has a matte appearance.

In both cases the scratches are perpendicular to the edge. Had I been able to get an image at the tip or base of a scallop, they would also look perpendicular.

The abrasive used on the front (the inside of the scallop) appears to be a little finer than that used on the back.

The scallop is 0.142" from edge to plane of the blade side. The angle at the edge is then 16.4 degrees. Recall the previous blade had an included angle of around 64 degrees. It would appear that one of the reasons these knives cut bread well is the very small included angle.

knife 4
200 X, flat side

Plane Iron Holding Jig

The platform on the QX3 is fine for very light objects where you want to see the top surface. With heavy objects, the platform just winds back down.

At 200X magnification, the depth of field is very small - around 0.004". If the object is flat, then the whole surface can be in focus. If the surface is sloped, like a plane bevel, then only a very narrow strip is in focus. The combination of these two problems forced me to build this wooden jig to hold the plane irons for viewing.

plane iron jig The slot in the left end holds standard plane irons, the slot in the right end the much thicker Japanese plane irons.

The angle of the slot to the horizontal is the same as the bevel angle, so the jig holds the plane iron with the bevel more or less horizontal (not exactly horizontal since I tilt the jig to raise and lower the bevel for focus).

The little bit of tape helps a little in holding the blade steady.

The slot should be wide enough to handle your thickest blade. This slot is just over 0.13".

The screw provides position adjustment. Advancing the screw raises the jig, lowering the bevel.

Rotating a finely threaded screw a couple of degrees moves the tip of the blade the few thousandths of an inch required to refine the focus.

Plane irons vary in length. Rather than build a new jig for each length, I put things under the jig or under the microscope to accommodate most of the variation (I use pocket books since I have them in so many thicknesses). Another good way to provide a whole series of different support heights is to use slips of paper - adding a slip adds between 0.002 an 0.004".

full setup
If you have square drive screws, you can leave the screw driver in the screw during adjustment. This is actually quite important, since the weight of the screw driver is enough to affect the focus. If the screw driver will not stay in the screw during the adjustment steps you will have a much tougher time getting a really good focus.

When the edge is almost in focus you can further fine tune by putting small weights in various positions along the top of the jig. I use a small knife, a small metal ruler, coins, any small object on my desk. Often a slight position change of these fine tune weights along the top surface of the jig will get the best focus.

Here is a picture of the jig in use. My jig is too light on its own to support some heavier irons, so I often rest some other plane iron on it, with the screw head going through the blade slot.

Notice the small Halogen lamp aimed at the bevel to provide the extra light necessary for a clear, bright picture.

QX3 illumination

You can replace the upper halogen light with bright white LED in 5 minutes.

The QX3 has a halogen light in underside of the microscope column and and another in the base. For viewing opaque objects like I do, only the upper light is important. The existing lower light is fine for providing a little contrast at the edge.

Changing the light will make a small difference, not a big difference. To get a really good image of a reflective surface you have to move an external light around until light from the external source reflects up into the microscope. You cannot do this with the fixed light on the bottom of the microscope column. So, even with the replacement bright white LED, you will still need a very adjustable external light source.

Bright white LEDs can be bought at any local electronics parts store - check your yellow pages under Electronic equipment & Supplies and ask if they sell single bright white LEDs. A 3mm or a 5mm LED will do - size is as important as the electrical characteristics. It seems just about any bright white LED will work fine.

You will need some wire snips and a small Philips screw driver.

QX3 Limits

The QX3 can only display features large enough to scatter visible light.

The wavelength of visible light is between 0.00004 cm or 0.000016" (purple) and 0.00007 cm or 0.000028" (red).

For abrasives graded in microns, we can directly compare the grit size with the wavelength of light. To do this we have to make an assumption about how big a scratch an abrasive crystal with a given diameter will make. If we had just one such crystal and we pushed really hard, we could make a scratch the width of the crystal. If we had thousands of such crystals, the force per crystal is very small. For example, if 1000 crystals are in contact with the bevel and we are pushing down with 5 pounds of force, the force per crystal is only 0.005 pounds. With this very light force, the crystal does not dig deeply into the blade. As a result, the scratch width is a fraction of the crystal width.

Now, I don't know what fraction. If you do, send me an email and let me know. I believe the scratch width is less than a quarter of the grit particle width. So, the Chrome Oxide abrasive, with grit size 0.5 microns, would produce scratch widths a bit more than 0.1 micron. A micron is one millionth of a meter, or 0.00003937". So, the CrO scratches around around one-tenth of this, or 0.000003937". This is about one-quarter of the shortest wavelength of light.

These scratches are not only not visible using the QX3, they are not visible using any microscope that uses visible light. As a result, the surface scratched up by the CrO abrasive looks scratch free.

In the images produced by the QX3, there are about 13,120 pixels per inch. That is, each pixel represents a distance of 0.00007622 inches. This is 2.5 times the longest wavelengths of visible light.

Other cheap microscopes

Not everyone is going to make a hobby of looking through a microscope at a plane iron.

Many people might learn a lot from a look through a cheap scope at their irons, especially if they are having problems.

People in the UK can get a very cheap 100x magnifier at Maplin - model number L11BK. The are a number of good hand magnifiers, usually called loupes, sold on ebay almost all the time. As usual with ebay, make sure you buy from someone with a good reputation.

When trying to see what is happening with a particular grit, it helps a lot to skew the tool slightly to the direction of the honing action. You will be able to distinguish scratches left by the current grit from those left by an earlier grit if you change the skew for each grit.

Once you figure out what is happening you probably won't need the microscope any more. However, if you have a tool that does not seem to sharpen, it might be helpful.

Site Navigation

Sharpening Home Page

Back to the Sharpening Overview


Try looking around the site map. You can also reach the site map from the little map at the top of each page.

Questions? Comments?

You can email me here.