LED ADAPTATION TO AO SPENCER

[Hobbyst46]

is the LED warm white or cool white? are you going to use the amber filter?

[stjepo]

No, I just showed part of the filters I have at home. I have used B-W 58ES TOP-POL. Is cool white 6000k


Regards

[MichaelG.]

75RR sent this interesting link:

http://www.microscopy-uk.org.uk/mag/art ... ersion.pdf

Regards

Could you please post the full link ... that shortened one doesn't work.

Thanks
MichaelG.

[stjepo]

The web cuts it, copy both parts together http://www.microscopy-uk.org.uk/mag/

artjan17/fp-LED-Conversion.pdf

[MicroBob]

The acceptabe temperature of an LED depends on the the temperature of the bonding wires that connect the semicondutor with the connectors. For an LED of a known maker it is possible to find this temperature in the data sheet. The stated value is often quite high, but it has to be taken into acount that there is a temperature gradient between bonding wire and heat sink. This temperature gradient has to be as little as possible. This can be acheived by using very flat surfaces on the heat sink, a thin layer of very good thermal compound (not the old white zinc paste) and a good amount of pressure via the mountin screws.
A good heat sink has three things: Conductivity, capacity and ways to get rid of the heat. Aluminium and copper have a good conductivity. Massive heat sinks have a good capacity to take up peak loads. The heat can be led away by convection and radiation. For good convection grooves and fins are good or a close thermal coupling to the microscope stand. For good radiation it has to be taken into account that a blank or polishes metal surface radiates nearly no heat!
Painting the heat sink in whatever colour greatly improves this. Black is slightly better than white, but less than expected. This property makes it so difficult to measure the temperature with an infrared thermometer when the surface is of unpainted metal.

A practical design for a good heat sink is an aluminium cylinder, turned for a precise fit in the microscope for good heat transfer, with a grooved, thinly spray painted end protruding from the back of the microscope.

A 20W LED will have more than 15W of heat to be led away without the perfectly coupled heat sink raise to high in temperature. This will make a big heat sink necessary or the time at full power has to be limited. 20W LED power is a lot!

[MichaelG.]

The web cuts it, copy both parts together http://www.microscopy-uk.org.uk/mag/

artjan17/fp-LED-Conversion.pdf

Thank You

http://www.microscopy-uk.org.uk/mag/art ... ersion.pdf

[stjepo]

The acceptabe temperature of an LED depends on the the temperature of the bonding wires that connect the semicondutor with the connectors. For an LED of a known maker it is possible to find this temperature in the data sheet. The stated value is often quite high, but it has to be taken into acount that there is a temperature gradient between bonding wire and heat sink. This temperature gradient has to be as little as possible. This can be acheived by using very flat surfaces on the heat sink, a thin layer of very good thermal compound (not the old white zinc paste) and a good amount of pressure via the mountin screws.
A good heat sink has three things: Conductivity, capacity and ways to get rid of the heat. Aluminium and copper have a good conductivity. Massive heat sinks have a good capacity to take up peak loads. The heat can be led away by convection and radiation. For good convection grooves and fins are good or a close thermal coupling to the microscope stand. For good radiation it has to be taken into account that a blank or polishes metal surface radiates nearly no heat!
Painting the heat sink in whatever colour greatly improves this. Black is slightly better than white, but less than expected. This property makes it so difficult to measure the temperature with an infrared thermometer when the surface is of unpainted metal.

A practical design for a good heat sink is an aluminium cylinder, turned for a precise fit in the microscope for good heat transfer, with a grooved, thinly spray painted end protruding from the back of the microscope.

A 20W LED will have more than 15W of heat to be led away without the perfectly coupled heat sink raise to high in temperature. This will make a big heat sink necessary or the time at full power has to be limited. 20W LED power is a lot!

Thanks for all the advice MicroBob, yes, 20W is too much. The reason I used is because I had it at home. In the future I will get a proper one and installed it.

[stjepo]

In the old Seiwa I used a normal LED sold in the market. Works very well:

[Hobbyst46]

Regarding the circuit for the 20w LED as drawn above, and from the specs of a similar LED chip found on the web, the current through the LED can be 0.7A. The voltage drop across R1+R2, when R2 is set to maximum (R1+R2=27.5 ohm), is ca. 19V. Then the voltage drop across the LED is 34-19=15V. The power due to the LED would be less than 0.7x15 = 10.5W. Most of the heat is generated by the current through R1+R2.

So I suggest to try to keep R1+R2 outside the base, away from the microscope. Then the cylindrical heat sink will need to dissipate less than 10.5W IMO. That can be still a lot, and since the aluminum cylinder is fitted inside the base, there is no option for cooling fins protruding from the base, as would be desirable. So, I thought that a an aluminum riser plate, thermally coupled to the cylindrical heat sink, could improve the heat dissipation. A brass plate would be fairly close in heat dissipation to an aluminum plate, and would stabilize mechanically the scope, by its weight. However, brass is more expensive.

If R1+ R2 are not on/inside the microscope, they need a separate heat sink. R1+R2 which should be high power resistors, for safety margins. The heat sink for the resistors can be a finned common device like for a CPU of computer, etc, or as shown by John B. mrsonchus above.

[stjepo]

Yes R1 + R2 are outside the microscope but do not get hot as the heat-sink does.


Regards

[stjepo]

I have made some measurements (hope they are correct) for you at maximum and at minimum. Hope it helps.




Regards

[Hobbyst46]

Thanks.
I do not understand the 0.02mA figure in the drawings (is it a measured quantity?) and why the circuit is open there below the pair of red arrows (there is a discontinuity). If the voltage measurements were taken on the closed circuit, and the resistance was measured on the open circuit, everything appears OK to me apart from the 0.02mA figure.

I recognize the gross mistake in my estimates of power above - I assumed that 0.7 A flows even when the rheostat is at maximum. This is wrong, since the rheostat limits the current. When R2 is set to maximum (bottom drawing), the current falls down to 0.2 A (0.522/27.7). The LED is probably turned off then.

When the LED is on (in the top drawing) the current through it is 0.634 A (from the measured voltages and resistances).
So indeed the power on R1+R2 would be small - 1.2W in contrast to my erroneous estimate. I think that they still need a heat sink.

But the greater problem now is again, heat dissipation of the LED itself... the heat sink indeed needs to eliminate 20W...

[stjepo]

The measurment of the ohms were taken free of any connection. The the rest in closed circuit. 0.02 mA is the current using my Fluke in the way is shown in the drawing.

I send you a better one.

The R1 and R2 get hot only in the maximum position that is not needed, minimum is more than enough. The AO illuminating system is superb.

Sorry to confuse you sometimes...


Regards

[Hobbyst46]

The reason I was asking about the 0.02mA is because, at least when the LED is on, the current should be several hundred milliamps, for the LED to emit light. Its spec says 600-700mA. Such current would be DC of course, and polarity matters (the LED is a diode). A current of 0.02mA (=20mA) can fire a tiny 5mm LED, not your 20W "lighthouse" IMO.

The important thing is, however, that illumination is superb. Excellent!! you are on the horseback. Just get rid of the heat.

[stjepo]

That is the spirit!!!!!! I keep you posted....

[stjepo]

Hobbyst46 I beg your pardon I have confused you with the reading of the amps.

At last I am sending you the correct circuit with the correct readings.

Sorry,

Regards

[Hobbyst46]

Great! team work is the greatest fun of all! everything now makes sense.

A little off-topic and not relevant to this job:
Some people think that the basic Law of practical electricity is Ohm's Law - others may think it is Kirchoff's Law. They are all wrong.
The most important law is this: At least 95% of all failures of electrical installations are due to either disconnected or erroneous wiring, not faulty
instruments. This Law covers all, from kitchenware to submarines...

Back to business: So now that you have achieved an officially working AO, even if not ideally, please do post more results.

[stjepo]

Great! team work is the greatest fun of all! everything now makes sense.

A little off-topic and not relevant to this job:
Some people think that the basic Law of practical electricity is Ohm's Law - others may think it is Kirchoff's Law. They are all wrong.
The most important law is this: At least 95% of all failures of electrical installations are due to either disconnected or erroneous wiring, not faulty
instruments. This Law covers all, from kitchenware to submarines...

Back to business: So now that you have achieved an officially working AO, even if not ideally, please do post more results.

Yes sir, is a proper comment in this case

Regards

[stjepo]

Meanwhile I find better LED for this microscope I use R2 also as an 'interruptor', using a little electric tape in the area where the heat goes up quickly, to avoid any unwanted damage.



Regards

[Hobbyst46]

The only electric tape that I know is an insulation plastic sticky (one sided) tape. It is made of PVC and is somewhat heat resistant, will not catch fire but its prospected use is under ambient temperatures with occasional heat surges - not excessive. Not for a continuously heated surface.
I think that replacement of the 20W LED with a 5W LED will simplify your project. This is based on the experience of numerous LED conversion projects I read about. With my scope, a 10W LED is too much, even for DF. I use a dimmer and the power is around 5W. At the most.

[stjepo]

The only electric tape that I know is an insulation plastic sticky (one sided) tape. It is made of PVC and is somewhat heat resistant, will not catch fire but its prospected use is under ambient temperatures with occasional heat surges - not excessive. Not for a continuously heated surface.
I think that replacement of the 20W LED with a 5W LED will simplify your project. This is based on the experience of numerous LED conversion projects I read about. With my scope, a 10W LED is too much, even for DF. I use a dimmer and the power is around 5W. At the most.

Agree with you, I was just showing how I was solving problems for the time being...

Regards