How powerful of microscope do you need to see the structures in a piece of steel. Is there a inexpensive one that hooks to your computer that would do the job. Not making a career out of it. Just curious about a couple things.
ib2v4u, this would depend on what structures you want to observe. 100X is good for a general overview, or getting started in checking grain size, 400X will allow you to examine most of the products of heat treating that we deal with if the grain size is moderate. If things a very fine 1000X comes in very handy but is more involved because I have found that cheap 1000X objectives are no really worth it and often at these resolutions a less expensive objective could involve oil immersion (and canola oil will not do
just fine 
once again there are oils made specifically for it (I am just so closed minded
).
The scope can be gotten rather inexpensive if you have some experience in working on them, the objectives can range from cheap to more than you would pay for a used scope.
So the basic scope is doable, after all I managed it, however that is only the beginning. One cannot just cut off a piece a steel and drop it under the scope and expect to see anything at all. The steel has to undergo a rather involved preparation process to allow the scope to see anything but a pure white field filled with black scratches.
Samples need to be cut off with zero heat (and remember we work with hardened steel) so a water cooled diamond or abrasive saw is often required. Then I rough grind the end clean and mount he work for polishing (you can go without the mounting but it will make polishing a pain and will make observing the edges of the sample almost impossible due to undercutting, as well as complicating the leveling process on the scope). My polishing begins at 200X, and then 400X and ends on the belt at 600X. Next I lap by hand at 800X, 1200X and 2000X before moving on to the metallography polisher. I built mine myself, but it still required a good DC motor, a variable speed reversible controller, and an appropriate disk attachment. I got diamond lapping disks for the initial work but they were quite pricey, so I found ways to get to the diamond compounds quicker and use leather backed disked made for the job. At this point the sample is polished at 3000X, 8000X and 15,000X before the final polish. Between every step the sample MUST be completely cleaned and rinsed of all previous grit material, I have found that Xylene works well for this, but you must use a very soft wipe. The last steps are at 50,000X and 100,000X.
At this point I etch the sample, you cannot see anything but a solid white field (unless you still have scratches) until it is etched. Not any etch will do, and ferric chloride is absolutely worthless for this. Many microstructures have a specific etch that will bring them out, I use nital the most. Nital is a special solution of nitric acid highly diluted in ethyl alcohol. My nital is a 2% solution. The sample is rinsed and cleaned with alcohol and then immersed in the nital until the "bloom" appears (the surface darkens) then the sample is immediately water rinsed and immersed in alcohol for a secondary rinsing. I remove it from the alcohol and immediately place it at an appropriate angle in a heated air blast. Water or alcohol left to dry on its own will destroy your etched surface. Air is the way to go as touching it with anything, regardless of how soft will be the equivalent of steel wool at this point.
How the etching works is that it makes ledges and ridges on the surfaces for the light transmitted through scope to bounce off at different angles. If that light comes back at you in all the same direction you will only see that solid white, but when the edges of the structures are undercut they begin to bounce light away at different directions and you can then see the stuff.
If working with an inverted scope all you have to do now is drop the mounted sample (another reason I work with standard sized mountings around the sample) onto the holder on the stage and start viewing. If you have an upright scope, you will need to level the work to the stage and objectives or only one side the image will be in focus. I have found the easiest way to do this is to put a small ball of modeling clay on a microscope slide and then set you sample on it. Protect the sample with a non abrasive paper and use a precision machinist vice to squish the whole thing out flat and level, then just put your slide in the scope and start looking.
If you wish to do photography you will need a whole other list of adapters and lenses to mount to a scope, and I had to make many adapters myself since they simply are not out there for my camera. Digital interfaces are the way to go if you are not really skilled in metallography. I tried the whole shoot an entire roll of film hoping to get one good image thing and do not advise it.
So if there is one thing that I would have to say is that metallography is a whole lot more than just buying a scope and going to it. It took me years to get set up to show the images I post occasionally, and in the end it really has little to do with my knifemaking productivity, in fact it is one of the largest time consumers I have gotten tangled up in so in many ways it has hampered my productivity. But it is VERY useful for telling people they are full of &%#! when they are spouting on about some half-assed method being "just fine" or as good as all that fancy industrial stuff- the steel doesn't lie.
from the Sticker Shock
. There is a new place in Ohio that sells supplies at a much more reasonable rate. For most of my simple work I use resins to set the pieces in. I mix it up, lay a strip of 2" masking tape on a tray sticky side up then stick my rough ground samples face down on it. I found a plastic pipe that is the perfect inside diameter for my stage mounting so I sliced dozens of 3/4" rings from it and these I just pres into the tape around the samples. Then I pour the resin and vibrate the snot out of it to remove any bubbles. 