+86 400-003-5559 CN


Replacement LED Light Build Uses A Few Tricks | Hackaday

tagsbest capacitor

Microscopes have become an indispensable workbench tool for hackers, enabling them to perform PCB assembly and inspection along with tiny SMD parts. A few years ago, the mad scientist [smellsofbikes] bought a stereo microscope from eBay. However, with its peculiar size, the 12-volt Edison-style spiral base lamp (connected to a 17-volt AC power supply) burned out after a while. He replaced the spare lamp with a burned-out lamp, and the spare lamp blew up after some time. Dumb light. Maybe the original specifications require the use of 24 volt lamps. Because of the strange Edison screw base, they are non-un, but these lamps will emit a rather yellow-orange light. In any case, for a period of time, he used a gooseneck lamp operated by a jury, but often moved the microscope, this lamp became a nuisance. After he had had enough, he decided


Usually, such a build is ordinary vanilla and does not need to be written, but [smellsofbikes] has some tricks worth mentioning. He found several high-power SMD LEDs in the parts box. They are only 1.6 mm wider than the terminals. Therefore, he took a double-sided copper-clad FR4 and mounted the LED edge on one side of the PCB board, twisting it slightly so that the two terminals can be soldered. This is a good heat sink for LEDs while still having a very narrow profile. This is important because the replacement LED board must fit the cylinder of the original lamp.

The LED is driven by a constant current step-down regulator and powered by the original 17-volt transformer. The bridge rectifier and several filter capacitors can produce low-ripple DC power, for which he calculated these values ​​using KiCad's spice function. The LM3414 driver he used was a bit out of routine. It can run LEDs up to 60 watts at a current of 1 ampere and does not require external current sensing resistors. Since he plans to run the LED at only 150 mA, this is too much, which will result in a very powerful, long-lasting solution. He designed the driver PCB in KiCad and milled on it

. The advantage of CNC milling PCB is that you can add custom copper overflow and expand the pad. This technique allows you to weld 0805 or 1206 parts to the same footprint-depending on what you can dig out of the parts bin.

What surprised us was that [smellsofbikes] stated, "The only annoying place is that (LM3414) also needs to connect its enable pin to high level, but the maximum voltage of the enable pin is 5.5V, so I must The rectified transformer voltage of 5.5V is 20 volts."

, There is a VCC output on pin 1, which comes from an internal regulator. From the data sheet, "VCC provides self-bias for the internal gate drive and control circuit." We are not sure why it (enable) (actually the PWM pin) cannot be connected to 5 volts without using an external regulator . Instead, he added a SOT-23 regulator to the design to get 5 volts to make the driver usable. But the voltage regulator he had had a strange inverted pin, and the SOT-23 LDO had to be soldered with dead wires to make it work. In the end, everything is fine, so we guess everything is fine, everything is fine.

In order to install the LED PCB on the side of the original cylinder, he designed a 3D printed collar with tapered fingers that can fix the PCB without using any fasteners. For the main PCB, a simple 3D printed shell is used to install it in the transformer box. Even at a driving current of 150 mA, the LED light will be very bright, and the over-killing driver will ensure that he does not have to worry about burning out the light soon.

Well, good point: I will connect from reg to the enable line. I kicked myself without doing so.

By the way, if someone else decides to study this problem, LM3414 is very suitable for switching 1A, but tps92511 is a similar product with a reduced cost and only handles 500mA. Generally, when using a step-down regulator, you have to work at more than 20% of the maximum rated current, because this is where the efficiency is highest. Therefore, if you are running a lower current, look for a good current match.

Congratulations on the successful construction. I bet it will run for a long time.

First of all, I want to encourage you to use LED substitutes to check the safety of your eyes. You introduced enough lumens, I think it is ok, but I also know enough about using lasers that they are surprising in frightening ways. Are your settings similar to known eye safety settings? From the pictures in the post, it seems that the light does not directly enter the objective lens, but reflects off the workpiece, right?

Also, your solution is small enough, I wonder if you can repackage it and stuff it into the original bulb equivalent? I can't tell from the picture whether you can harvest the old connection in the bulb and epoxy, or use potting compound to ridge it. If you do not need to modify the internal components of the microscope to make it work properly, then you may find an actual market for LED replacement bulbs suitable for old equipment. I can see that one is being built (if I have a microscope and need to replace the bulb, that is the problem I hope to get ;-)

Similarly, NICE BUILD!

There do exist 24V AC LED replacement ES bulbs.

But I doubt whether they have the required light output, they are used for indicators and not for lighting.

How about their performance under 17V voltage is worth discussing ((is it PP or RMS?)

When I do a lot of microscopy, the hottest ticket is fiber optic lighting. But this is (and probably still is) something you can only buy after you have a sufficient equipment budget.

The voltage of the transformer is 12VAC (root mean square is 9V), but the voltage from the transformer is actually closer to 17V.

Yes, I hope to be able to set up the fiber one day.

Regarding the previous question, yes, this is shining on the workpiece rather than optically. I am engaged in LED driver IC design and testing (I work on this specific chip), and we have learned some things about LEDs, which can damage eyesight. I will not look directly at this LED, but once it passes through the optical lens and expands a bit, it is within a safe range. (And I have equipment to verify this.)

cool. Thank you for following up on this information. I thought you might indeed know that they are safe for the eyes, but I want to be a reference for other people studying here just in case. By the way, I helped maintain several IR lasers (95 and 125mJ lasers with 11ns and 9ns pulses respectively-not in a place where human eyes are safe).

Yes, my previous occupation was in the field of high-power lasers, ultraviolet rays and visible light. I have many colleagues who have partial damage to the retina or DIY laser disease. They are all PhDs with years of experience before the accident, so I decided Doing something to reduce the risk of life, so now I play with high-power LED.

(Interesting realization: make sure you can see the visible lasers and know where they are, but they can also harm your retina. On the contrary, you only see the fluorescence of the skin and realize that it is only a second-degree burn I realized that there is a UV laser problem, but those that will never pass through your cornea, you can get a new cornea.)

A bit OOT, but about eye safety: I want to know whether high-power IR LEDs can damage the eyes? Infrared light is invisible to the naked eye, but the combination of a high-power source + a pupil that is fully opened in the dark means that a large amount of infrared radiation can reach the retina.

I originally planned to use a webcam with powerful IR lighting to monitor the nests of wild animals (birds, hedgehogs...), but I am afraid it will cause harm to the eyes of these animals...

It has been a while since I needed to check the safety literature of wind animal laser eyes, but I do remember that there is a lot of literature there. Also remember that many animals and insects cannot see the same spectrum of light as they see things, so things that are safe for us are not always safe for other small animals. I also appreciate your attention to animals, and know that there are also scientific papers studying their safety limits. Please note that along these routes, there are night vision safety devices and wildlife cameras specially made for this purpose, and there may be at least some efforts to protect wildlife. I recommend starting by searching Scholar.google.com, then checking the specifications of the wildlife cam, and then still creating your own cam. The thing I have never seen is setting up multiple infrared illuminators around a space (pointing to the camera) so that none of them are very bright, but in the end they get similar overall brightness.

Also remember that there is a lens in the eye that focuses light on the retina, so you don’t have to have a very strong beam of light focused to a certain degree to burn the spots on the retina. Also don’t think it’s safer if the pulse is very short-the pulse is less than 10ns, and enough power will evaporate the eye fluid in the eyeball, and the expanded bubbles will send shock waves through the eye, causing more damage than a single point burn to the retina. Hmm... it has been more than ten years since I conducted a laser safety training review. Hackaday will be a good place for laser safety stations.

I think unless you manage to invent an infrared flash and trigger it at a frequency of a few hertz, you will not be able to manage the infrared light source bright enough.

However, considering only laser safety, why can a visible red laser be seen even when used as the weakest output of an IR guided laser? ? ? It makes the focus of the eye the closest to the focus of the IR, and if you use light blue or green, the focus of the eye to the IR will be very off.

No, no, don’t panic, I said to myself it’s normal... Beeeecause, if the frequency you use deviates far from the IR, it cannot be used as a guide laser, pass it through a prism, and the blue beam has passed The rain cover is removed, and your infrared beam deviates about 10 degrees from the rain cover. Useless, unable to guide where other optical devices pass. I think when it hits other things that are not safe and useless to the eyes, you may look down in the IR, and the blue is driving past you at a "safe" distance Head.

On this issue, my first thought is that your lighting will hardly be as bright as the sun in the IR range. So I really want to say that the possibility of danger is very small.

But then I considered that many of these animals will not be extinguished in the sun. If they do, their pupils will be closed, which may not happen under infrared illumination at night.

I think as a starting point, I will compare the lighting level of your infrared light source with the lighting level of the sun. If it is 10%, then there may be some worry. 0.1% may not be.

I was thinking, if you try to take a direct look at a powerful standard LED light (visible light), it can be said that it is even only 15W, and for a long time, I think it will not be very good for your eyes. And you may not bear it.

However, if you replace the visible light with the invisible IR with the same power, you will not feel any; if it is in the dark, your pupils will be fully opened, allowing more "light" to enter.

In addition, what if you put the light in a confined space (nest) very close to the animal, and let the animal turn on the light while resting for hours or even days? Some animals do not even close their eyes.

But the same question may also apply to humans: what if you put it directly close to the crib?

What if you use a powerful IR LED illuminator (50W or higher)?

I found a reference to "IEC/EN 62471 (risk group 2)" in some Axis luminaire data sheets (T90D30). The standard seems to consider the following infrared lamps: "Eye-infrared radiation hazard exposure limit". But you have to pay for it...

But according to:

"GR2-Risk Group 2 (medium risk): maximum exposure time is 100 s"

Therefore, staring at the illuminator for more than 100 seconds does not seem to be a good idea. Unless it only works with visible light?

"The advantage of using CNC milled PCBs is that you can add custom copper overflows and expand the pads."

You can use any PCB for processing, whether it is milling, etching at home or manufacturing through PCB services.

Indeed it is. You can use it for any form of homemade method, not only for CNC, but even for PCB services. Although when ordering from PCB fab service, I hope to dial in the right footprint.

At work, we only need to use approved footprints to send our things. le

What I have done with gEDA PCB is that there are few or no traces, and I used the Voronoi processor to make the largest pads for all the components on the board. In this way, it is completely automatic, and I don't have to make special oversized pads. I haven't learned how to do this in KiCAD.

This is also mixed. The absence of a solder mask means that you don’t have to remember to put in a solder mask cutout (and there are many potential short circuits).

Kikkad to Voronoi

Great build! My oscilloscope needs such a light-I will have to try PCB tricks to connect to the LED and dissipate heat.

I rebuilt an old B&L understage illuminator for the student oscilloscope, which used a spherical bulb in the Edison base. I printed an Edison base to fit the slot and installed the LED, and it worked great!

A few years ago, I made a strobe illuminator for bio-oscilloscopes, which used PIC uC to blink LEDs at an adjustable rate. It's nice to see the moving cilia of the bioists.

3D printing the bulb base is a good idea. I must remember this.

[smellsofbikes]> "The only annoying place is that (LM3414) also needs to connect its enable pin to high level, but the enable pin has a

>The maximum voltage is 5.5V, so I have to derive 5.5V from the 20 ish rectifier transformer voltage. "

I have seen components that also have an enable pin and a regulated output, but unless the enable is declared first, the regulator will not work. A bit of a chicken/egg dilemma.

I have used a white or blue LED powered by a series resistor from the main power supply as a 3 to 4 volt shunt regulator, which is only used to drive the enable pin. It works, but I am not satisfied with the required power resistor. (This is something I can quickly organize.)

Why not have more LEDs in series and a resistor for current limiting?

Part of the reason is that I work in LED driver design, part of the reason is that I like efficiency, and part of the reason is that there is too much light and too much heat to handle in the optical tube. It’s also much more difficult to use this LED mounting system for serial strings: each LED needs its own PCB, or I need to stagger the isolation cuts in the copper, which is cool, but also causes too much heat to make the available surface area dissipate.

Eye-safe lasers are usually marked in milliwatts rather than watts. 15W to 50W can be a dangerous problem. In other words, I think you may mean LED-type bulbs (with diffusers, etc.). This should help, but I'm not sure to what extent the microscope light source is related to the original article.

Please be kind and respectful to help make the comment section great. (


The site uses Akismet to reduce spam.

By using our website and services, you expressly agree to our placement of performance, functionality and advertising cookies.