Searching Cree LED for tiny dimmers

[Marc]

Hi,

I'm very interested in these tiny dimmers without flickering to try a LED replacement.

One dimmer works with input 6-24V, the other with 3,7-24V


https://aliexpress.com/item/1005004292615287.html


https://aliexpress.com/item/1005002997950962.html

I'm not any expert in electronics and I have some difficulties to find one Cree LED suitable for any of both. And the sellers are not very helpful.

Is there a Cree LED suitable for any of both?

Any help will be appreciated

[jfiresto]

I need more light and did the converse: I found a flickerless dimmer to drive a particular CREE LED. More specifically, I found a MEAN WELL IDLC-25, dimmable, constant current power supply to drive an 18W COB LED.

That way the LED (and illumination) come first and there is much less risk of buying too cheaply. I now, however, have to efficiently dissipate ca. 5W of heat from a 16x16mm square.

[Marc]

I need more light and did the converse: I found a flickerless dimmer

I understand, although your solution is different. Note those tiny dimmers are smaller, and avoid the install of driver and potentiometer .


https://aliexpress.com/item/1005004292615287.html


https://aliexpress.com/item/1005002997950962.html

Somebody knows a Cree LED which can be suitable for these tiny dimmers?

Any help will be appreciated

[jfiresto]

Unfortunately, my oldish browser is not properly rendering the above links so I do not know the technical particulars of those dimmers.

I expect you will have to add a current limiting resistor in series with the LED, sized for its operating current, so that the LED can operate off the constant 12V, 24V or whatever the dimmer module outputs, or that you feed into it. (Again, I don't know the technical data.)

If they really are constant voltage, wouldn't a dimmer have to dim an LED by pulse width modulating it, that is, flickering it: I hope at a high enough frequency that you (or your camera?) don't notice?

[Marc]

apparently, these dimmers are of constant voltage and driver integrated.

Characteristics are these:

- Supply voltage: applicable between DC 3.7V-24V
- Output voltage: equal to the input voltage

- Load power: use within 4A (V*I=P)
- Standby power consumption: 0 power consumption in standby after turning off the knob
- Work type: constant voltage not constant current load must have resistance current limit


there is a video in this page showing the operation (hope you can watch this):
https://aliexpress.com/item/1005004924934358.html

or try this direct video request:
https://video.aliexpress-media.com/play ... 633383.mp4


In many comments the people says there is no flickering, although truly it should be tested with a camera. Although the price is so cheap that I can accept the risks
The more interesting thing is the size which is really small. At least in my case, this could facilitate a lot doing a replacement and its inner placement inside the microscope. Maybe just by moving some components, without necessity to remove them.

However, I suppose it would demands a Cree LED > 3.7v . And I cannot find a Cree Led with higher voltage and around 10 W. As I'm not any expert in electronics, my fear is the install of a 12v LED with excessive power, causing excessive heat or something like that. Or if maybe I choose one Cree LED with characteristics not suitable.

Because these reasons I ask here. I hope that somebody can give some advice, about choosing one Cree LED which could be suitable for these dimmers.
If you can recommend any LED with a suitable electronics, I will be very grateful to know

[imkap]

Mouser has all kinds of Cree leds.
I made my dimmer with a step down converter from Ali express it worked without any flicker with any led connected. Including cree. And cheap maybe 5$
Good thing is that you can limit max voltage and use any Power supply (e.g. 19V 3A old laptop adapter (free) to power and dim a 9V 12w Led)

[imkap]

If you can recommend any LED with a suitable electronics, I will be very grateful to know

This is what I did, look at the last post for the latest version. Dimmer works great. Higher voltage doesn't mean higher temperatures, higher power (watts) mean more temperature...
You can limit the upper voltage, consequentially current (power) and light power up to the limit which your LED (heatsink) can withstand. I limited mine at 17.5V if I remember correctly so it was working at 10W (I measured this, not calculated so it wouldn't overheat on the heatsink provided.

https://www.microbehunter.com/microscop ... 12&t=14944

I bought this kind of LED chip, a few different versions with different power and CRI index:
https://eu.mouser.com/ProductDetail/Cre ... cycode=HRK

This is pretty powerful, but I had some strange halo on my microscope. I think LED is the culprit, but maybe not. I bought another microscope with complete lighting meanwhile so stopped searching.

I got advice here to get a high power single LED, not an array. I didn't listen to that, but I think I should have.
So the next thing I'd buy would be that kind of a LED. I think I might do it sometimes, but not yet ...

Maybe one of these?

https://hr.mouser.com/c/optoelectronics ... r%20Rating

[Marc]

This is what I did, look at the last post for the latest version. Dimmer works great. Higher voltage doesn't mean higher temperatures, higher power (watts) mean more temperature...
You can limit the upper voltage, consequentially current (power) and light power up to the limit which your LED (heatsink) can withstand. I limited mine at 17.5V if I remember correctly so it was working at 10W (I measured this, not calculated so it wouldn't overheat on the heatsink provided.

thank you, this is useful info. Because I believed the voltage could condition the heat.. . Glad to know about your experience. Then probably I will try some 9v or 12v Cree Led with some dissipation. Thanks a lot

This is pretty powerful, but I had some strange halo on my microscope. I think LED is the culprit, but maybe not. I bought another microscope with complete lighting meanwhile so stopped searching.

I know in astrophotography these purple halos are a frequent issue, and quite eradicated using UV/IR cut filters. Here there is a useful reading:

https://skyandtelescope.org/astronomy-b ... otography/

it says the lens quality are just another factor. I wonder if some UV-IR cut is already included in the manufacture of expensive lens for microscopes. I don't know really, although maybe there is a similar improvement as happens with the telescope lens. Anyway, I would try to insert one cheap UV-IR filter in front the sensor just to satisfy the curiosity for the results.

I have one halogen lamp so I cannot talk with real knowledge. Although I wonder if maybe also these halos could be reduced with a more natural light, perhaps around 4000k. My microscope doesn't have top quality lenses but I don't notice these halos. And I wonder if a warmer light could mitigate that. Although it could depend of the wished white, of course.

After reading your problem with the halos I think to include in my project different leds, from 4000k until 6500k
I was asking for a cheap component although the reading of this forum can increase the costs by day

thanks again for the useful info

[blekenbleu]

This is pretty powerful, but I had some strange halo on my microscope. I think LED is the culprit, but maybe not.
I bought another microscope with complete lighting meanwhile so stopped searching.

Microscope collector lens for halogen bulb is designed for relatively small square filament;
larger area LEDs provoke stray light that bounces inside microscope
against reflective mostly cylindrical and conical surfaces, resulting in halos.
Better to source LED of desired power and similar dimensions (e.g. 5mm or less).
Single LEDs already mounted to e.g. LED Star MCPCB are easier to DIY.

[Marc]

Microscope collector lens for halogen bulb is designed for relatively small square filament;
larger area LEDs provoke stray light that bounces inside microscope
against reflective mostly cylindrical and conical surfaces, resulting in halos.

that's interesting because I'm reading some things in the subject and the expensive manufacturers seems to put more weight in the light spectrum.
Perhaps they don't comment those other issues. According these explanations about the U-LHLEDC mounted in the Olympus BX43, the led try to mimic the halogen, natural light, in order to reduce the blue traces. Probably with a light temperature between 4500-5000k


page: https://www.olympus-lifescience.com/en/ ... color-led/

they also mount a light diffuser before leaving the light to the condenser:


I have found other few expensive led systems for microscopes which works between 4500k-5000k. Although some expensive microscopes for darkfield and phase contrast mounts 6000k and beyond.

The convex lens for these expensive microscopes seems to include a focal length enough to concentrate the light into the condenser, although I wonder if there are no more special efforts. It seems to be quite standard lens, easy to find. This is for some Zeiss microscope:



I have a diffused convex lens into my microscope, although it seems are not very common. I have seen a newer model of my microscope using led instead halogen, although mounting a transparent convex lens. I don't understand the cause while those expensive led systems are including a diffuser.

[blekenbleu]

expensive manufacturers seems to put more weight in the light spectrum.

Low cost and efficient white LEDs indeed have problematic color rendition, with some being more nearly blue + yellow
than red, green and blue. Sourcing more expensive LEDs with better color rendering helps
mitigate chromatic aberration exacerbation by LED's stronger and shorter wavelength blue component.

A microscope designed for LED illumination presumably has illumination optics (collector and field lens)
appropriately designed to control stray light, avoiding halos; my comments above were about LED conversions from halogen.
At least some Nikon and American Optical models (two microscope brands with which I am most familiar)
with halogen illumination optics for Köhler alignment also provide ground glass filters
presumably to improve wide field illumination uniformity at low magnifications.

[rcmjonesmicro]

I converted the illumination on an AO series 20 using a 6W 6500K Cree button style LED and a PWM powered by an external 9v ac/dc adapter. It works fine but at 4x and 10x I have to employ the built-in ND filters as it is too bright without. This is a minor inconvenience but allows plenty of lighting for 100x oil, darkfield and phase contrast. The small PWM went easily above the lamp housing and the LED, mounted on a heatsink, attached to the original lamp holder and is readily adjustable.

[Marc]

A microscope designed for LED illumination presumably has illumination optics (collector and field lens)
appropriately designed to control stray light, avoiding halos; my comments above were about LED conversions from halogen.
At least some Nikon and American Optical models (two microscope brands with which I am most familiar)
with halogen illumination optics for Köhler alignment also provide ground glass filters
presumably to improve wide field illumination uniformity at low magnifications.

thank you for the information

[dedalus]

I have made a dimmer with just a mosfet transistor, a variable resistor, capacitor to prevent self oscillation and a fixed resistor, the mosfet conected in series to a led. You can solder it in the air leg to leg and wrap it in insulating tape while keeping stable. Evidently it is totally flicker free at any frequency and i'm not worried for the colour temperature varying, the regulation has to be very extreme for it to be apreciated. Just be careful with the heat emission of the mosfet as it is totally linear. I think it's worth a try for microphotography as with a digital camera the effect of colour temperature variation will be compensable. I think the light being a higher colour temperature while more dim, as it happens with a led on linear regulation, could be useful in direct observation in some cases.

[cuxlander]

Hello,

I use a similiar mini-dimmer since 5 years, using a Cree XM-L U2 SMD-LED, 280lm, 6000K, CRI 80, (I don't know if still available)
My dimmer is shown below. It has a separate on/off/dimmer-switch, to be safe I have included a 3Ohm resistor.
The whole setup fits neatly inside the base of my little Zeiss KF2. I run it with 3 or 5(USB) Volt. The dimmer is of the pwm-type, you may get stripes on photos.

Altogether a simple effective solution.

Cheers,
Hans

[bkt]

I'm very interested in these tiny dimmers without flickering to try a LED replacement.

The most common way to dim a LED/DC motor/bulb is PWM. Pulse width modulation, a square electric signal with a variable duty cycle. Illumination depending of the duty cycle and the resulting analog level.

For our eyes, it shall not be an issue. I believe such dimmers work at frequencies far above 50/60Hz, maybe at some 100 to 500Hz.



I do not know if this PWN technique could cause flickering with fast cameras (or with the shutter more or less in synch with the PWM frequency). The alternate way to drive a LED could be constant current source. All sort of modules exist also, some working at very high frequencies, producing a constant current, not pulses, so without any flickering risk.

An LM3410 would work at 500kHz or 1.6MHz, an inductor smoothing the current. They still can be driven with PWM (or for some, by an analog level), but the LED current is so constant, smoothed by an inductor.





Keywork for such would be "constant current", not "PWM":

LM3410.JPG (149.04 KiB) Viewed 901 times

One of the most commun and cheap module is made with an LM2596, runs at 150kHz:

[bkt]

[The convex lens for these expensive microscopes seems to include a focal length enough to concentrate the light into the condenser, although I wonder if there are no more special efforts. It seems to be quite standard lens, easy to find. This is for some Zeiss microscope:

You will find a lot of such mini lenses that fit on LED chips, to narrow their beam. Most of LEDs having up to a 120° beam (for general purpose LED illumination), sure not well adapted for microscopes.

Following is a Motic LED option, to replace a bulb. They do not use such a lens. Looks like they found a ref which already has an adapted beam angle... So far, I didn't find a ref of such a LED.

[bkt]

I now found that Osram doc, about LED illumination, including for cameras. "Compared to traditional fluorescent light sources, LEDs instantaneously translate their operating current into light, without much smoothing effects. Simple, cost effective circuit designs can lead to a lower quality of light." https://dammedia.osram.info/im/bin/osra ... %20JM_.pdf

So in case of any flickering or weird effect noticed with a camera (depending of shutter speed vs PWM frequency), PWM could be the cause, the module shall then be swapped out for a constant current / smoothed current module...

OsramCameraApplications.JPG (70.04 KiB) Viewed 893 times

[MichaelG.]

Thanks for posting that Osram link
… it’s a very useful document !

MichaelG.
.
.

P.S. __ I had never seen that Motic LED module before, but I think it features in the second part of this video
https://youtu.be/46PeQ-sMA4M?feature=shared

[Chas]

Looks like they found a ref which already has an adapted beam angle... So far, I didn't find a ref of such a LED.

LED's are quite complex / can of worms
This page and the next one on has some info:
https://luminusdevices.zendesk.com/hc/e ... l-Lighting

Direct projection of a basic led (without a diffuser ) is surprising ; the outside of the image is quite a different colour to the inside, which can sort of mean that with an Abbe condenser you can get rid of the 'LED blues' by lowering the condenser

[bkt]

P.S. __ I had never seen that Motic LED module before, but I think it features in the second part of this video
https://youtu.be/46PeQ-sMA4M?feature=shared

Correct, that LED module is designed for BA310 or BA410, comes in a aluminum housing (heatsink), it plugs in the bulb socket, and as shown in that video, it is to be screewed to tighten it mechanically. No idea of specs of that LED.

Direct projection of a basic led (without a diffuser ) is surprising ; the outside of the image is quite a different colour to the inside...

I didn't investigate it further. For other purposes, I searched for mini lenses to fit top of a 120° LED, for a much narrower beam. But I'm not sure this will produce a nice homogeneous illumination source either. If common LEDs are all that unequal, all solutions should need a minimal diffuser in the path...

This is from commercial descriptions of cheap plastic beam focusing/collimating accessories, to be plugged on LEDs, for different beam angles:

[MichaelG.]

Several variants of the Motic LED module shown here:
https://moticeurope.com/en/accessories/lamps.html
… but not much detail.

MichaelG.

[bkt]

Looks like they found a ref which already has an adapted beam angle... So far, I didn't find a ref of such a LED.

LED's are quite complex / can of worms
This page and the next one on has some info:
https://luminusdevices.zendesk.com/hc/e ... l-Lighting

Direct projection of a basic led (without a diffuser ) is surprising ; the outside of the image is quite a different colour to the inside, which can sort of mean that with an Abbe condenser you can get rid of the 'LED blues' by lowering the condenser

But white LEDs are now cheap. For any basic setup, I would order some and test... Criteria being high CRI (acceptable spectrum), some 3W and a more or less 20° to 60° narrow beam...

If part of the LED produces more blue, it can be filtered before entering the abbe?

[Chas]

If part of the LED produces more blue, it can be filtered before entering the abbe?

Yes possibly, eg if the projected image is larger than the bottom lens of the condenser the outer part could be 'dumped' .. I noticed the effect when projecting the image of the LED in a Jansjo desk light (so probably a pretty basic design of LED) into the condenser, I guess it is the 'flip chip' design talked about on the second page of the Luminusdevices website

But the little filter 'from the past' works pretty well
These images were taken using a large diffused LED source (Ulanzi VL49) that claims a CRI of 95+ using a bottom of the shop 10x chinese plan objective (and an NDPL2 adapter)

without the filter :

Plan x10 with LED.jpg (18.68 KiB) Viewed 572 times

with the filter:

With Kodak F1.jpg (18.6 KiB) Viewed 572 times

Both with the same focus, exposure and white balance setting