Re: LED illuminator

"Peter" <peterplanapo@xxxxxxxxxxxxxx> wrote in message
On Jun 20, 4:15 pm, "helio" <forbid...@xxxxxxxxx> wrote:
"helio" <forbid...@xxxxxxxxx> wrote in message

Eventually, I found a 100W LED on ebay :

That's incredible.

Best regards

It's a 10x10 array of 100 1-watt LEDs. The light output of 7000 lumens
is spot on at 70 lumens/watt, close to the top of the efficacy range

Aren't all HPLEDs (High Power LEDs) arrays of single diodes wired in
parallel? It seems to me that your Seoul P7 LED makes no exception:

You could make a 100kW array in this way if you wanted. Not
incredible, but huge. However, I'm glad you posted this as I may try
one out.

So, at least let me call it a 'state of the art' LED, as I suppose it is
impossible to increase ad infinitum the number of the arrays on a chip: "A
recurring problem is that efficiency will fall dramatically for increased
current. This effect is known as droop and effectively limits the light
output of a given LED, increasing heating more than light output for
increased current." (wiikipedia).

It can't be used for Koehler illumination because it's far too big
(over an inch square of emitting surface) but it may do a good job for
what Zeiss call "semi-Koehler" wherein they put an array with a
diffuser just under the field diaphragm, thus ensuring that both field
and aperture are evenly illuminated (which was, after all, Koehler's

My education in Optics is near to zero, so I ask: isn't it possible to use a
condenser to reduce the emitting surface? When high-resolution observations
have to be made, isn't the use of a diffuser avoidable?

I reckon this array would put about 30x the light flux into the
condenser compared with my already super-bright Seoul P7 in true
Koehler config. That might well be enough light to be able to make
short enough photo exposures for stop-motion (though I don't think
fast enough for rotifer cilia)

Incidentally, can I suggest to use Peltier cells for cooling ? I found some
interesting modules here:

But, how on earth to make a control circuit? It would need to be a PWM
controller with the facility to produce a single full-intensity flash
with a variable duration of, perhaps, 100 microscecond to 10ms. For
normal viewing the array would be too bright and would need to be
modulated down to perhaps 5%. Any ideas there?

It seems to me that the circuit I suggested in a previous post would be