Experimentations with high powered LEDs for microscope illumination

[gekko]

For what it is worth, and to try to put things in perspective, I've used my old Labophot microscope for phase contrast with a 40x objective at about 80% of lamp voltage with, I think, quite satisfactory results. The illumination from a phase annulus would be only a fraction of that with brightfield. The Labophot uses a 20-W, 6-V tungsten-halogen bulb. The limitations arise when trying to take pictures of fast-moving critters, where you need a fast shutter speed, where more intense illumination has a great advantage, I think.

[75RR]

Update:
Resistors arrived from China in 28 days.
Dimmer presumably on its way, but since it has no tracking number, I can't tell.

Smaller than I imagined.
Caution: Batteries not included!

[billbillt]

These work very well for me... The dimmer will be smaller than you think also..

[Crater Eddie]

Yes, these resistors work fine. Mine does get hot when I crank the brightness up, so I have it mounted on a heat sink.
Still no problems with the PWM controller, I am very satisfied with it.
I did try out a switching power supply to replace the storage battery for my lamp power supply, and ended up with banding in stills and video. It is a 12v 12 amp switching power supply from China. It works fine, but there is just too much noise in the output, so I went back to the battery. This winter I'll work on filtering the supply output, maybe I can clean up the noise and get rid of the banding.
I hope your PWM module arrives soon, when is the estimated delivery date? They usually give a pretty broad range.
CE

[billbillt]

Hi Eddie,
I Bought The following three items from Ebay...
http://www.ebay.com/itm/201204252733?_t ... EBIDX%3AIT

http://www.ebay.com/itm/400786090744?_t ... EBIDX%3AIT

http://www.ebay.com/itm/171764548583?_t ... EBIDX%3AIT

The power supply is coming any day now... The other items are "in transit" from China...
When you figure out how to figure the noise from your power supply, please post an update and I will do it to mine..

[Crater Eddie]

That power supply looks a lot like mine, Bill, just a little smaller. That is an interesting LED unit, will be interested to see how you mount it.
CE

[billbillt]

Hi Eddie,
It is my intention to try to use it as the old style lighting systems that you shine into the mirror... It has up and down and left to right adjustments...

[billbillt]

I also bought one of these big boys a while back... but I didn't want to dedicate it to one application.... That is why I bought a smaller one..

http://www.ebay.com/itm/321564484058?_t ... EBIDX%3AIT

I also bought one of these last year for experimenting...

http://www.ebay.com/itm/30V-5AMP-5A-Dig ... 33a3eccaea

[Crater Eddie]

That first one looks like the one I have Bill, except for the added fan. I thought about getting that one, but decided it was needless overkill for this project.
The last one there looks a lot like one I have at work. Or at least I HAD one until the boss borrowed it.
CE

[billbillt]

Yep, it would be a sure overkill for these small projects... That is the reason I bought the 5A unit...

[Rodney]

Photo of my heat sink fan combination I may use for something.

Rodney

[billbillt]

I have one of those that is almost identical to yours...

[Crater Eddie]

Yep, CPU heatsink / fan assembly. That is the type of heatsink I have my power resistor mounted on. The fan isn't needed since it doesn't get THAT hot... I used the fan to cool the LED as previously described. I have boxes full of this sort of thing, you never know when it will come in handy for something.
CE

[billbillt]

Hi eddie,

That is for sure... I used to have mountains of this kind of stuff... We got flooded in 2001 and I had to throw it all out... I have since attempted to re-stock!...

[75RR]

Well, just goes to show. Despite low expectations, now and again one gets a nice surprise.

The dimmer arrived today!

This is an image of it connected to the Tungsten 12v 60W bulb using DC from the power supply.



Visually no banding and very smooth. Only potential problem (may just be in my head), is that the bulb does not seem that bright at the upper end of its range.
See Video:
Image is of a small piece of color transparency (from way back) taped on to a slide. Use it as a quick testing medium.
Slight fluctuations are not visible by eye, they are probably due to the camera sensor adjusting to change in light intensity.
Tested Dimmer voltage output with my Tester, steady at 11V even when rotating Dimmer.
Is that due to a current change rather than the voltage?



Still have to connect the LED. Is it as simple as adding the resistor to the Red (+) wire?
Am a little leery of losing more LEDs.

[billbillt]

Hi,
The dimmer "dims" by PWM "pulse width modulation".. In essence it just turns the LED on and off quicker than the eye can detect... Longer off cycles mean the LED is less bright... More on cycles the brighter it is.. I don't know what LED you are going to use, but with your supply having a maximum of 12V output, it will be easy to hook up an LED... Most white LEDS use a forward voltage of 3.6V maximum... Without knowing exactly what type and rating yours is I can't give any more info...

[75RR]

Longer off cycles mean the LED is less bright... More on cycles the brighter it is..

So it is effectively lowering the top of the illumination range?
Thought DC already had that effect vs AC with incandescent bulbs.
How much illumination have I lost and is this going to shorten the life of my incandescent tungsten bulb?

Without knowing exactly what type and rating yours is ...

Crater Eddie and I are using the same Cree XM-L U2 in order to simplify troubleshooting.

[75RR]

Update:
Still fiddling with the Tungsten bulb.

Dimmer does not work bellow 8V.

Dimmer dial has a definite "off click" when turned completely to the left (cuts current off completely), internal off/on switch set towards dimmer dial. Can't remember now if you said that was normal or reversed.
Just checked and polarity is as shown i.e. +- IN and +- OUT reading left to right.

Must have done something strange before because voltage is now changing as the dimmer is turned.
Sending max Power Supply output (Nominal 12V) DC = 13.6V
Dimmer voltage Range 0 to 11.8V smooth as you like.

[Crater Eddie]

Glad it arrived! We'll sure, there is always loss in any circuit. Meaning that you won't get everything at the output that you put into the input. But you don't loose much. I'm on the road right now so I can't check my switch, but you can easily determine the output polarity with your meter and match it to your LED. Yes, you simply connect one output wire to the LED (the correct polarity) and the other output wire to the resistor. The resistor doesn't care about polarity. Then connect the free lead of the resistor to the free lead of the LED. That's it.
I'll check mine this afternoon when I get home.
CE

[billbillt]

The 8V cutoff limit makes sense... The device is rated at 9V - 60DC...

[75RR]

Well, connected it all up and it works. :)

Photo 1: Power Supply, Dimmer and Resistor.
Photo 2: LED heat sinks, both Tungsten Bulb and LED elements are on the same plane to allow köhler.
Photo 3: Lower crenelated edge of focused element on Front Focal Plane (offset so it can be seen) Image reflected on slide.
Photo 4: Zeiss 12V 60W illuminator that I converted. Actually it works with both (LED and Tungsten) so is swappable.

[billbillt]

Yep, Looks like you have a good setup there... All of this stuff is generally inexpensive, so a person can afford to experiment...

[75RR]

The 8V cutoff limit makes sense... The device is rated at 9V - 60DC...

That is for input. Output goes all the way down to 0

[75RR]

Yep, Looks like you have a good setup there...

Thanks. Would like to tidy it up a bit at some point. Maybe put everything in a box.

[Crater Eddie]

Great! I'm glad to see that you have it up and running. I measure 11. 8 volts maximum output on my PWM controller, I consider that perfectly acceptable.
I observed that my controller quit working when my supply battery dropped below 8 volts or so, so that is consistent with your result as well.
I hope the controller and LED work out for you, keep us posted.
CE

Edit:

No, it should not have the effect of shortening the life of your bulb.

[75RR]

Hi Crater Eddie,
I have been looking at some voltage figures, do they make sense to you? Can you run a similar test?

All Tests run from lowest Dimmer setting to highest.

Nominal 6V setting on Power Supply
IN Range 7.7V - 7.31V
OUT Range 1.79V - 3.38V
OUT Range (Incl. Resistor) 1.92V - 2.78V
Current Range 0.01A - 0.29A

Nominal 8V setting on Power Supply
IN Range 10.5V - 9V
OUT Range 2.12V - 9V
OUT Range (Incl. Resistor) 1.93V - 3.12V
Current Range 0.03A - 1.55A

Nominal 10V setting on Power Supply
IN Range 13.4V -11.3V
OUT Range 2.24V - 11V
OUT Range (Incl. Resistor) 1.98V - 3.2V
Current Range 0.05A - 2.12A

Nominal 12V setting on Power Supply
IN Range 16V - 13.2V
OUT Range 2.3V - 12.9V
OUT Range (Incl. Resistor) 2V - 3.3V
Current Range 0.07A - 2.74A

Note: I find the 6V setting is enough. 8V, 10V and 12V settings are too bright.
It is not the highest setting that is the problem but the lowest setting.
When centering or focusing the Field Iris, lowest settings are too bright to look at except for the 6V.

[Crater Eddie]

Yes, these readings make sense to me. This is good data.

WARNING: The following contains generalizations and simplifications. Detail oriented people might get upset at this. Calm down, we aren't building a nuclear reactor, we are lighting an LED for Pete's sake.

You can see that the resistor is protecting the LED by limiting the current to under 3 amps. If you wanted to, you could experiment with other resistor combinations to get closer to the recommended maximum current, to eek out the last few remaining lumens. But it's always a good idea to leave yourself with a little buffer zone. I would rather loose a few lumens than loose a few LEDs.
You can also see that your power supply is sagging somewhat under load. What I mean is that at low brightness (minimum load) it puts out 16v, and at maximum brightness (maximum load) it has "sagged" to 13.2v. This is to be expected in a small non regulated power supply. It's not a fault, it's just the way it works.
Back to the resistor again.
Lets use the last data set, since that is the range you are going to use. (right?)
At maximum brightness your supply is outputting 13.2 volts at 2.74 amps. That is 36.17 watts. The LED is seeing 3.3 volts at 2.74 amps, which is 9.04 watts. This is a 10 watt LED running at 9 watts. As long as you keep it from overheating it should be happy and run just fine for a very long time.
But 36.17 watts minus 9.04 watts leaves 27.13 watts... where is all that extra power going???
Grab hold of that big power resistor. HOT isn't it? That's where all that extra power is going. The resistor is doing exactly what it is there to do, it's dissipating all that extra unwanted power the only way it can, as heat. That's why it has those fins on it, to help dissipate the heat into the air. That's why I have mine mounted on a big heat sink, to help get rid of that heat faster. You don't have to provide extra heat sinking for the resistor, it's built to take it. But we can, so why not?
But why am I going into all this. It works, right, so why bother?
Because, somewhere down the line someone is going to read this thread and get all upset because we are using such an inefficient circuit. Well that is true, it is very inefficient, I pointed that out a long time ago. We are wasting 27 watts of power as heat to light up a 10 watt LED, that is pretty wasteful. Sure, I get that. But it's relatively inexpensive, and it's easy to hook up and operate. If we were running this off a set of D batteries or something, this would be a terrible idea. But we are running off grid power, so we can afford to be wasteful, as long as we provide cooling for the heat generating components. (Except in my case, I'm running mine using a large storage battery, but that's temporary, and I am aware of the issues.)
Sorry for such a long post.
CE

[75RR]

Did not seem too long. Did find it Very useful. Thanks
Did you see the 6V update on the previous post?
My previous comment on a 8V cutout were based on the Tungsten Bulb, LED will not however work on the Nominal 4V.

Note: I find the 6V setting is enough. 8V, 10V and 12V settings are too bright.
It is not the highest setting that is the problem but the lowest setting.
When centering or focusing the Field Iris, lowest settings are too bright to look at except for the 6V.

[billbillt]

Here is a link to a short and very badly done video of my set up for a 10 Watt LED lamp... For some reason my dimmer does not have an "off" position on the intensity dial.... I din't want to fiddle with the micro switch on it, so I put in a in line switch... I ran this in the vid without the current limiting resistor.. You notice I can't dim the lamp all the way to "off"...

https://vimeo.com/133267543

[Crater Eddie]

75RR:

I didn't see your edit until just now. As long as that works for you it should be fine. The only issue you might have, the dimmer might not be as stable at the lower input, that is it might be struggling. I would run some video tests to check for artifacts. If you don't see any banding creeping in then you are good to go. If you find that you have to run at one of the higher settings you could always place a filter in the light path to tone it down, maybe crossed polarizers or something. Again, not efficient, but a workable solution.
You mentioned putting everything in a box to tidy it up, that is a good idea. If you take the cover off the dimmer you will see that the potentiometer has a fairly long cable on it, this is to allow you to separate it from the dimmer and mount it remotely.
Just remember to allow for good air flow through the enclosure, you don't want things over heating. I would mount the power resistor on the outside to allow for convection cooling. If you really want it hidden inside the enclosure it would be best to add a cooling fan.


Billbilt:

That's odd that yours does not have the "OFF" position on the potentiometer. Slightly different model I guess.
Anyway, mine does not fade all the way down to zero brightness either, as shown in my video earlier in this thread. I don't find it to be a problem, as the lower position isn't too bright for me, except maybe for very low magnification use.
If your power supply isn't capable of providing over 3 amps, then I guess you don't need the limiting resistor. As long as it works well, you're golden.

It's great to have folks experimenting with things like this, as everyone has a different take on things. It is interesting to see different approaches to solving the same problems.

CE