Improvements to my microscope setup for SMT work

Why do I need a microscope for SMT?

The single biggest challenge to doing “real” SMT work (0805 or smaller components and fine lead pitch ICs) at home is being able to actually see what you are doing.  I know that there are many hobbyists (and maybe even some budget-conscious professionals) who will disagree with me, but I wouldn’t dream of working with surface mount components without using a microscope.  I’ve tried many alternatives, including a 10X handheld triplet loupe, a magnifier ring light, even a nausea-inducing magnifying visor, and none of these even come close.

In case I haven’t made myself clear: I would rather solder SMT’s with a 150W soldering gun than with anything other than a decent stereo microscope.

In January of this year, I scored a stereo zoom microscope on eBay.  While my scope is far from state of the art (it’s a “vintage” American Optical model 569) the optics are fantastic and it quickly became the most prized piece of equipment in my shop.  Here’s a photo of the scope shortly after I added it to my lab, for more photos and information about it, see my original post.

Stereo Zoom Microscope

For the first few months, I used the scope pretty much as it arrived.  One of the first major tasks I used it for was assembling the first batch of AVR HV Rescue Shields, and for this purpose it worked extremely well.  However, as time went on, it became clear that I needed to improve my setup in a couple areas:

  1. The magnification range of 7-30X was great for working on a few tightly grouped 0805 or smaller components, but was too high for general PCB work.  A typical BGA package was larger than the field of view.
  2. The included incandescent projector-style illuminator (shown piggybacked on the scope in the photo above) could only be placed in a limited set of positions and did not have adjustable focus – it made a nice, bright spot in the center of the image that didn’t fully illuminate the field at low zoom levels.  While it is removable from the scope (this provides a workaround for these issues), the included stand took up too much bench space to be practical.

Upgrading the microscope:

The first upgrade I made was to add a secondary objective aka barlow lens to the scope.  A secondary objective serves to increase or decrease the total magnification of a microscope, while simultaneously trading off working distance, the distance between the bottom of the microscope and an object in focus on the bench.  In my case, I added a 0.5x secondary objective, which gave me half the magnification while increasing my working distance by approximately 2x.  While American Optical stopped making accessories for the StereoStar 569 long ago, Reichert, who acquired AO’s microscope line, still sells parts and accesories, including the #575 0.5X secondary objective, shown below.

Supplementary Objective

The secondary objective screws into the existing threads on the bottom of the microscope.  Here it is installed on my scope:

Secondary objective installed

Now with the secondary objective installed, I have a zoom range of 3.5-15X and a working distance of 6-8″.  If I need higher magnification, I can always remove the lens.  Perfect!

The second upgrade I made was to add a fluorescent ring light to the scope.  I picked up the cheapest one I could find on eBay.  This model is sold by Amscope, outputs 8W, and is available for under $30:

Fluorescent ring light

The ring light conveniently attaches to the newly installed secondary objective by tightening three thumbscrews, and provides a decent amount of light that fully illuminates both the object I’m working on as well as the surrounding workbench area, which has been surprisingly helpful.   Best of all, the new light stays out of the way and provides more even illumination than the halogen projector that came with the scope.

Here’s a photo of the microscope setup as it looks today:

New microscope illuminator setup


While the changes I made are significant improvements over my original setup, I have made a few observations that may lead to even more tweaks and upgrades in the future:

  • The increase in working distance due to the 0.5x secondary objective is great, but it puts the scope significantly higher above the bench.  I didn’t appreciate that this could be an issue until I had to buy a taller lab chair to see through the eyepieces!  I’m not sure how to work around this, but it’s good to be aware that more working distance isn’t always a good thing.
  • The color temperature of the fluorescent ring light is very poor (cool) compared to the halogen illuminator it replaced.  This gives everything a slightly depressing blue cast and is far from a true color representation.  Most noticeable are tantalum caps, which go from bright orange in color to a sort of slightly orange-ish dark grey under the scope.  Yuck!
  • Ring lights can create pretty nasty glare.  This might be a side effect of how I have the ring light mounted or the distance to the bench.
  • The 8W fluorescent lamp is ok, but more light would be better.  Fluorescent ring lights are nice and cheap, but better performance can be achieved with a significantly more expensive fiber optic illuminator.  I may look into getting one of these in the future.

Despite these minor issues, I am pretty happy overall with the new setup even after a couple hundred hours of heavy use.

16 thoughts on “Improvements to my microscope setup for SMT work”

  1. I might be talking nonsense here but you could look into the photographic films that are used to correct for tungsten etc. colour casts in the old-school (non digital) world, and place a ring shaped piece over the tube.

    I had a quick look and came up empty (well excluding filters designed for attaching directly to a camera), but back in 2001 we had a photographer come to the house to take pictures for submission to an architectural prize. He used a film, medium format body (digital was barely invented) and the fluro tubes in the kitchen were clad in, IIRC, orange film (probably an what’s termed an Orange 85 filter in the camera world) to correct the colour cast.

    So it’s definitely available – or was back then.

    1. Oh, I should also add that I too tried a variety of things. I’ve not really progressed to SMD yet but am getting to the point where I could benefit from downsizing to TQFP packages but I picked up a bargain (about 50UKP – which compared to the 10UKP LED loupe that gave me a headache just looking at it seems like a steal) iluminated work lamp on ebay from Waldmann (a to be precise.

      Optically it’s pretty good (but I’ve been putting of an eye-test for a while, so the slight distortion on the edges might be me) and the magnification is upgradable (no idea on price but it ships with 3 dipoters and can be increased to 15 which sounds like lots). It throws out heaps of light and is built like a tank.

      The lens also acts as a partial fume shield if your iron doesn’t extract. This is a long winded way of saying to anyone else reading this, if you find one of these going cheap, it gets a big thumbs up from me.

      If only they did an upgrade that made my soldering look good at these magnifications, I’d be happy 🙂

      1. Wow, Waldmann has a lot of really cool lamps! The Omnivue looks great, nice score! I’m positive that it is far superior to the cheap swivel lamp I have. I’ll have to keep my eyes open for one of these!

    2. I think I’ve seen orange filters before, although maybe I am thinking of something else (cleanroom lights?). Filtering means losing some of the available light, so this is probably not a good option unless I can somehow come up with a brighter fluorescent light – there isn’t enough light as it is!

    1. Ed –

      The nice thing about the 0.5x lens I just added is that it will protect the objective lenses inside the scope from fumes and is much easier to clean. An alcohol soaked Kimwipe usually does the trick, although I doubt that is the “correct” way to do it. It is also possible to get 0x secondary objectives that perform the same protective function without changing the magnification. I have seen these in use in labs I have worked in. Usually they also function as a place to mount a ring light on scopes like mine that don’t come with a good place to clamp a light.

      1. Yes, cleaning with alcohol ought to work. Isn’t there some sort of special cleaner used for resins on PCBs if you’re being really neat? I wonder if that would damage lenses. Still, nice to have a relatively cheap lens in the direct line of fire.

    1. Tim – I actually have used a Mantis before. They are nice (and expensive), although the working distance on the model I used was not nearly as good as my setup.

      Most of my experience with SMT has been with a stereo zoom microscope (usually Nikon or B&L SZ series). I’ve never really had a problem using a traditional style stereo microscope, although admittedly I do not wear glasses while soldering. I can definitely see that glasses would make things more difficult.

      1. We had a couple Mantis, a vision system, and a number of microscopes, most of which were pretty during my days at a contract mfgr. My favorite was the antique AO569. 🙂 I may need to go grab one off ebay.

        I never liked the ring light, nor the LED ring lights we built up. My favorite, albeit a bit hot after a period of time was a dual head halogen lamp from Sunnex.

        1. Ron,

          I have used halogen fiber optic ring lights before and have been pretty happy with them. They put out a truly staggering amount of light and keep the heat away from the work area. Unfortunately, they tend to be very expensive compared to fluorescent or LED ring lights.

  2. Light has two components, (in general usage) those being, color temperature, and color rendering index.

    Color temperature cannot be “poor”, since it’s just the tint of the light. Candles are about 1500K, incandescents are 2500K, general purpose “warm white” florescents are 3000K, and the sun is 5000K. The most efficient color temperature is 5000K, since that matches the frequency response of the eye, but it isn’t quite as big of a deal as the color rendering index.

    Color rendering index measures how well the light renders color, as you may have guessed, and the lower it is, the more efficient. Monochromatic light has a CRI of 0, rendering color in shades of gray, while a really cheap florescent bulb will have a CRI of 6, with more expensive bulbs making a CRI of 7 or 8. An incandescent has a perfect CRI of 10, and a correspondingly lousy luminous efficiency.

    1. That’s the first I have heard of CRI before, thanks for your illuminating (!) comment on the subject! Comparing my incandescent halogen illuminator to the ring light, the difference in color temperature is obvious, but I didn’t realize that the CRI also had big a part in making one light look “better” than the other. Thanks!

  3. I’ve found that white LEDs work well illumination. The first option was a couple MR-16 replacements from WalMart on gooseneck arms borrowed from some desk lamps. Recently, I found a better option – an LED replacement for car dome lights. It’s an array of 6 x 8 3mm white LEDs. Extremely bright. I just mounted it to the bottom of the scope with double-back tape. A nice feature is nothing gets very warm – very nice compared to being in close proximity to a halogen bulb.

Leave a Reply to Tim Cancel reply