Not So Simple LED Upgrade for Microscope

[Amen] obtained a microscope whose light source was an incandescent bulb, but the light from it seemed awfully dim even at its brightest setting. Rather than hunt down a replacement, he decided to replace the bulb with a 1W LED mounted on a metal cylinder. The retrofit was successful, but there were numerous constraints on his work that complicated things. The original bulb and the LED replacement differed not just in shape and size, but also in electrical requirements. The bulb was also part of an assembly that used a two-pronged plug off to the side for power. In the end, [Amen] used 3D printing, a bit of metal work, and a bridge rectifier on some stripboard to successfully replace his microscope’s incandescent bulb assembly with an LED. He even used a lathe to make connector pins that mated properly with the microscope’s proprietary power connector, so that the LED unit could be a drop-in module.

Working on existing equipment always puts constraints on one’s work, usually due to space limitations, but sometimes also proprietary signals. For example, a common issue when refitting a projector with an LED is to discover that the projector expects a stock bulb, and refuses to boot up without one. Happily, the microscope didn’t care much about the bulb itself, and with the LED positioned in roughly the same position as the original bulb’s filament [Amen] obtained smooth and even lighting across the field of view with no changes made to the microscope itself.

17 thoughts on “Not So Simple LED Upgrade for Microscope

  1. Actually I would urge caution on this topic, a microscope will concentrate the light onto your retina, onto a spot.
    and a 1 Watt LED is fucking powerfull, and additional to that, it sends the light out directed, the bulb doesn’t, even when it had the same power rating. One should take a look at the resulting power per area.

    1. sorry but that’s just BS a microscope is not a laser. by definition it provides an virtual image not a spot. and considering that the microscopes i know and use every day have 60-100W halogen bulbs i’m really not that worried about the 1W led.lets go through some rough numbers 100w undirected light would be about 30W directed since in my microscope the lamp housing doesn’t have a reflector and the optical path is non reflective. 1W LED is about the same as 5-6W halogen or about 9-10W incandescent. so basically 5-6W is still much less then 30W. as allays YMMV but i really don’t think there is anything to worry about here….

    2. Despite Andy’s confident assertion, a moment of thought would show it’s obvious a microscope will not “concentrate the light onto your retina, onto a spot.”

      The light source must (as the article makes clear) illuminate the field of view evenly, not concentrated into a small spot. How else would you be able to see through the object being viewed if the light were not filling the field of view?

      Though the bulb in the posting looks like an automotive taillight bulb, a proper tungsten microscope lamp has a planar, approximately square filament about 3 mm on a side, with one side facing the microscope’s condenser lens or mirror. Half the light from that lamp goes in that direction (with the back half just going to heat the base of the bulb).

      Again, despite Andy’s misconception, the LED does not send “the light out directed”: it’s more-or-less a lambertian source much like the filament, filling a few steradians with light. See, for example, the spatial distribution graph on page 22 of the Cree LED datsheet:

      The total light power out of a (real) microscope bulb of the class used by the microscope in the article is a few hundred lumens. The 1W LED used produces about one hundred lumens. It’s arguable that the LED lumens can be collected more efficiently, so the total brightness onto the object and up the eyepiece is going to be about the same.

      It’s not going to burn anybody’s eye out. Any user competent enough not to look into the tungsten light at full brightness won’t be blinded by the LED either.

      1. Indeed it does look like a car bulb, 99%.
        In the UK that size/type tend to come in 12V, 6V and 24V versions.
        WIth wattages of 5W, 21W and his pictured is a dual filament of more than likely 5W & 21W.
        Could just be the wrong bulb fitted / it’s using the lower power filament – which is avery common fault made in cars.

        But it’s certainly an elegant hack.

    1. Cree’s datasheet says the maximum reverse voltage is 5V. It would probably survive at the currents you’d get out of the 6V supply, but it would be a bad idea to depend on operating it that way. The linked article says he used a bridge rectifier to make DC from the AC supply.

      1. Smarter people will correct me here but a tungsten filament has thermal properties making the emission spectrum very wide and continuous over much of the visible range. An LED emission has less thermal retention and would switch off and on at these AC frequencies even when rectified. Some filtering with a capacitor is necessary to avoid that effect. Photography might be affected. As to direct connection to AC, some current limiting must be used for both forward and reverse conduction to avoid destruction of the LED.

        1. I guess I was too brief.

          Yes, there will be flicker in an AC-driven LED, just like those intensely-annoying LED Xmas LED light strings. As the original article describes, a capacitor after the rectifier reduces that.

          Of course there needs to be current limiting (in both directions) to avoid LED damage. I meant by my comment that the actual current from a resistor-limited 6V supply (as opposed to a constant-current supply) would be much less in reverse bias than in forward bias. The damaging heating in the LED would therefore be much less than you might otherwise expect.

          In actual, real measurements, a single white LED has a forward voltage of about 2.6-3.4V depending on the LED and current. Rated reverse voltage is typically 5V, but the current in this direction is just a few microamps. Even at 20 volts, reverse current is typically only 100 microamps for a T1-3/4 size, and the LED survives just fine. Still not recommended.

  2. 1w is weak in my opinion. My first microscope was a 3w LED AmScope and it was underwhelming. More recently I converted my 30w halogen scope to a 10w LED. Granted, I only use a camera on it and have not tried looking through the eye-peice. In all honesty, I’m scared to look because the light is blinding without the condensor, and it will completely light a 400sqf room by itself.
    My conversation process was actually simple. I bought the LED and a suitable Ac to DC power supply. If anyone is interested in pictures, just let me know how I can upload them here, and I will.

    1. Electrical power input tells you nothing about the luminous output of LED, a modern reputable vendor VS no-name China brand can easily have several times the flux despite having the same input power, not even mentioning the color rendering…rather look that what the manufacturer claims the luminous flux to be.
      The reputable ones (usually…) don’t lie about it.

  3. Interesting, neat build… great idea and effective design.

    Earlier today I ordered a 60mm angel eyes white LED ring light for a few dollars that I am going to see how works with the Stereo-Graf’s as both stereo microscopes did not come with their lighting modules.

    This got me thinking that I had saw a “desoldering” method on youtube using 12V halogens that have a reflector in the bulb. I didn’t find any bulbs like that when I checked at the local Meijer super store. I haven’t looked since… though was wondering regarding for the two Stereo-Graf’s restoration that something like above though with more of a machined tube with one side that is facing the area to be illuminated milled open and maybe lined with titanium dioxide or something with a higher reflective surface and a glass tube over (cut a pyrex test tube dia that fits). I suppose can even silver… though I’ve not done so with aluminum… only glass. Maybe can try with just glass… thanks for the brainstorming session and inspiration. I just had a duh moment! :-)

  4. Another thing to be careful about is the temperature and spectrum of the emitted light. The tungsten filaments provide a nice wide-spectrum and light geared toward the viewing of specimens. LEDs will output a harsher spectrum with peaks if not careful.

    Just a small example of the difference, but should not be considered the final say:

    1. Agreed on the spectral quality being a problem. That nasty blue spike in LEDs makes for some awful chromatic aberration in my venerable old Olympus scope. Fortunately it’s easy to remove with filters, but I’ll be looking for a higher CRI LED for my next retrofit.

    2. One interesting problem with LEDs is the distinct lack of IR light. With modern cameras having repectable performance in the IR, and common dyes like ICG and even the Alexa Fluors (in absorption mode), lack of IR in the source is a disadvantage. Fortunately, it’s easy enough to swap the halogen back in.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s