Testing a USB digital microscope

 kw: product reviews, microscopes, digital microscopes

From time to time I see ads for digital microscopes, and I decided to get one. I have two microscopes and I had camera equipment for use with them, but since my last film camera broke down, I don't have a convenient way to take photos through them. My digital cameras don't have an adapter to mate them with a microscope's eyepiece tube. 

The ad I clicked on was for a no-brand device, made in China and shipped from China. It came in a box labeled just "Digital Microscope" that has a table of specifications on its bottom:

"Image Resolution" is a key piece of information. 640x480 barely counts as digital imaging these days, but it is OK for a $35 instrument.

"Focus Range" is actually greater than that shown, as we will see below. Also, there is no indication of the zero point, so I had to define one. I suspect my "0" is their "15mm", or close to it.

Though it says it is compatible with Android, you would have to obtain a USB-to-MicroUSB adapter.

The weight, 200g, must include the box. I found that the microscope plus cord weighs 74g, and the stand weighs 46g, for a total 120g.

Red Flag: the stand weighs barely an ounce and a half. For stability it needs to weigh more than twice as much as the microscope. I dug around in a junk box to get three 30g tire weights, and tied them to the top of the base of the stand, for a total stand weight of 164g. That made the setup much more stable.

On the side of the box is another table with check boxes, checked to indicate this is a "X4" model with 1,000X magnification.

Installation did not go according to the little sheet of paper that came in the box. The instructions mention only a website to go to and an installation file to run. That installed the HiView software, but HiView wouldn't connect with the microscope. There is a small CD in the box, and one of its files is a driver loader. After I ran that, the microscope would connect with the HiView software. With that in mind, I don't know how you would get this instrument to connect with an Android phone.

With some fiddling around, I found that with shiny objects at low or medium magnifications the LEDs are too close to the objective lens and make strong reflections that spoil the image. I had made a light stand for use with my inspection microscope, so I tried that out. This is how the setup looks. The card the microscope is looking at is one of my calling cards.

This image is close to life size. The lamps are work lights that were being given away at Harbor Freight a few years ago, and later were sold for about $3, so I bought several. The base is a sheet of steel, curved to shape and painted white. On the stand behind the microscope parts of the tire weights are visible.

The microscope has a thumb wheel for focusing, and at certain distances, it also sets magnification range. Just above the plastic holder from the mounting are two buttons. One turns the lights off and on, with three brightness settings. The other is labeled "Snap", to take an image, but actually using it would jostle the 'scope. Use the software icon in HiView instead. The standoff tip is clear plastic, which is convenient for seeing what you are lining up, and letting light in.

When you run the software and connect to the microscope, you are always given a selection of image resolutions to use. I experimented with this, and found that the microscope indeed takes 640x480 pixel images, but the resolution can be set much higher. However, when you select a different resolution, the software interpolates the pixels, but there is no added information. The sensor is apparently 640x480 pixels. There must be a range of microscopes that HiView can use.

How does the microscope perform? Here are some early images:

This is at a rather low magnification, a midge. The long filament from the head toward the tail is not part of the insect. The microscope's clear standoff tip was about an inch from the specimen. When I put the tip in contact with the paper the insect was on, and shifted the focus to the higher-power area, I got this:

The magnification here is about 12 times greater than the image above. It shows the "hair" on the "knee" of one leg. Both of these images were taken using the LEDs supplied with the microscope.

Next I took a series of images to calibrate the magnification, at least roughly. I used two scales for this, sitting atop my calling card. Firstly, the four calibration images used for low range:

Most photo editing apps let you pick two points, or drag a rectangle, to get the number of pixels of width and height. I used IrfanView to "count" the pixels in one magnified millimeter. In the upper two panels, the scale is a plastic printed ruler with a mm scale. In the lower panels, I used a target that came with the microscope; the smaller lines are spaced at 1/10 mm.

This image shows the highest magnification, with the standoff tip in contact with the card and scale, and the focusing in the high range. The two long lines are 1mm apart.

The standoff distance I used was from the tip of the standoff tip to the object. The four images above were taken at distances of 125mm, 50mm, 10mm, and 0mm (contact).

I made a little table of the results, with effective magnification for two conditions. First, as displayed on my monitors, which are 22", 1680x1050 pixels, and a scale factor of 3.55 px/mm or 90 px/inch. Secondly, for publication, it is recommended that artwork have 300dpi. That means that a 640x480 pixel image will be reproduced no larger than 2.13" x 1.6". 

The microscope can probably be used more than 125mm from its subject, so lower powers than 9x (or 3x) are possible. It appears that the highest magnification possible is about 137x on the monitor, or 40x for publication. That's a far cry from the 1,000x advertised. I intend to test further, and I'll add an update whenever I learn more.

I tried putting a thickish (2mm), small item on the desktop, putting the tip in contact with the desktop, and getting the microscope to focus on it. No dice. It seems that the 0mm setting is at the end of its range, or close to it. The fly leg seems to be at a higher magnification than 137, so I'll experiment further.

I charted the "Monitor" figures, low range only, so I'd have an approximation of how much standoff I need for various magnifications, with these results:

The odd wiggle at the upper left of the curve is an artifact of Excel's low-ball spline used to make a curved line between the points. The curve is good enough for my purposes. A few data points of interest: 

• For 10x, use 100mm
• For 20x, use 40mm
• For 30x, use 18mm
• For 50-60x, use 2mm or less

In the usual case, I'll use it like a zoom lens, putting the subject at a distance such that what I want to see fits in the frame with a little freeboard around it.

I am a little disappointed in a few things, as I mentioned, but this is a workable microscope for (very) general use. One cannot expect high technical quality for $35.