Saturday, September 10, 2016

Reconditioning a 50+ year old microscope - part 3 - LED conversion

The microscope came with what I believe it must have been the original illumination system:

  • a 30 Watt / 110 Volts tungsten bulb (in this case a GE branded one):

  • a light dimmer circuit apparently made up of just a Triac, some capacitors and the potentiometer to control at which threshold the waveform is chopped (therefore causing the light to dim proportionally to how much of the wave is chopped):

  • and interestingly, a Hitachi branded mains plug:

However for current use, this system would be impractical several fold:

  • the specific type of tungsten bulb it requires is very difficult to obtain today, and those available (mostly new old stock) are not cheap;
  • as listed in the specs of these light bulbs, the lifespan is stupidly small, in the range of 50 to 60 hours.

  • incandescent illumination is not ideal because of reddish spectrum of light in lower power levels. A blue filter is normally required to adjust to a more even spectrum:

  • given the low efficiency of tungsten bulbs, heat dissipation and IR radiation are high, affecting some types of specimen;
  • in my location the mains voltage is 220 Volts instead of the 110 Volts required. This would imply adding a transformer to convert the voltage;
As such I studied the conversion to an LED light source, which I knew it would have multiple advantages:
  • LEDs are widely available and cheap;
  • lifespan is high, in the order of tens of thousands of hours;
  • light spectrum is consistent across all levels of intensity. Also (depending on the selected LED), a nearly pure white source is possible without having to add any filter;
Nevertheless I had to take into consideration a few aspects:
  • make sure a good thermal dissipation would be allowed. LEDs don't like too much heat;
  • ensure the maximum current delivered to the LED is actively limited, ideally by some regulator circuit;
  • Like in the original system, provide a means for the LED intensity to be varied through a potentiometer installed in the same location of the original one.
As such I put my hands to work. Started by removing the original illumination gear:

Ordered from banggood:
  • a bunch of 3 Watt white LEDs (rated at about 210 lumens each);
  • a power LED tailored heatsink:
  • a step-down (buck) converter rated at a maximum of 3 Amps, with controls for constant voltage and constant current applications:

The work started with the preparation of the heatsink to fit the area originally occuped by the bulb. Some cutting had to be done in order to make it a good fit. Tested the LED mounted on the heatsink for proper operation:

Made some tests with the step-down converter, limiting the current to 500 mA (below the rated 700 mA but more than enough for the lighting requirements), and replacing the on-board multiturn pot by a 500 Ohm panel potentiometer and a 220 Ohm resistor in series (and as such obtaining only the safe voltage range to be used for controlling the light):

  • machined an aluminium disc to provide the necessary width for fitting a DC connector in the hole where originally the thick mains cable would pass through:

Well these specs won't remain true though (in particular the 20 Watts of consumption) :) 

While possibly not making much difference, preserved the original black mask:

Put it all together, inside the microscope base (mounted the LED + heatsink, the step-down converter PCB, the DC connector, and the potentiometer, in the place where the original one would go).

Fired it up and..Bob's your uncle:

See also:

No comments: