AC Powered White LED Circular Magnifier Work Lamp

,

Use bright LEDs to replace florescent circular light in magnifier work lamp. Let there be light!

A medium difficulty Instructable to fix a circular magnifier work lamp by converting to a very low energy, high reliability alternative light source that avoids using mercury contaminated florescent light bulbs.

Step 1: Background to the LED light bulb idea.

,

Ok, so I cannot turn down a challenge either. I bought this really nice, nearly new, Magnifier Work Lamp for only $4 in a local Thrift Shop. Great bargain, says I, these regularly retail for $50 to $100 and this was a good one. So what if the bulb is blown, says I, as I noticed it would not light when I plugged it in before buying. New bulbs can not cost that much, right? Hahaha. What a bargain, says I!

At home, I tried easy fix #1... replace the bulb... Darn! Still not working!

Followed with easy fix #2... take apart the power box near the base. Yay! Found a wire shorted beneath the cover plate screw. Noted the electrical flash marks by the screw hole. See pix. Repaired insulation and tested the lamp. Darn again!

Next was easy fix #3... take the light head apart (instructions below) Triple Darn! Looked good.

Most likely, the little lamp starter card part (mine was heat-shrink in black) was blown during the short circuit in the lower power box.

Spent the next few days in electrical supply stores learning that the little lamp starter card part that leads power to the lamp is not available in my area. Crap Darn! Just not going to be my week was it? Better get this fixed or I just blew my $4 and possibly my reputation.

Comments

[HoldOnTight]

Great 'Ible! I'd build one when my florescent lamp burns out. I can build a replacement LED-based lamp for less cost, be more reliable, more efficient, more environmentally friendly... Seems like a simple choice. I have several circle-lamps throughout the living room and at the work bench. Any questions?

To clarify some other comments made: LEDs require a narrow forward voltage in order to produce light, usually 0.7 to 1.1 V. Current will begin once the threshold voltage is achieved. The capacitor has an "impedance" (which) depends on the capacitance &frequency (50 or 60 Hz) and the resistor, which when in series, both resist the current to a (hopefully) acceptable level once the diodes have achieved the required forward voltage. So, the supply voltage must be above the voltage threshold required by the number of diodes in series to get enough current for the diodes to emit light and produce a current. So,

supply Voltage > X * (# LEDs) * 0.7 to produce light.

So, the number of LEDs can be significantly increased as the linked circuit schematic shows and it is limited by the source voltage, since the current in a string needs to be about 20 mA, a comfortable operating current for most LEDs.

The current will primarily depend on the impedance of the capacitor and the resistance of the resistor. For safety with the given circuits, the resistor and capacitor should be able to accept the full load should the diodes short out (an avalanche effect for the diodes and resister) as a safety precaution. So the capacitor should be rated to withstand a voltage of 370Vac, and the resistor should be rated to handle the power as well. It would be best to add a fuse in series with the power source to the presented circuits. I would estimate a reasonable fuse would be above the normal operating current, but not so large as to allow current to cause a fire, say 0.5 Amp.

Arcticpenguin, Awesome upgrade! Please add to this 'ible if you do any more upgrades!

[seahorse47]

It must be that you mean ...

supply voltage X > (#LED) x 0.7

If this is not correct, then please advise what X stand for in your formula.

So far I agree with all you said except for the above. Thanks.

[dilate]

Nice work!! Now how can we make that device mobile.
I'm looking for one that can power by battery.. Nothing really good out there on the market that does.. So how can we power this super tool for a portable mobile setup?

[rgbphil]

Great idea, I've just gone through replacing my ring flouro on my own magnifier lamp. Couple of points though....I presume you meant your configuration is 2 sets of 30 LEDs, in parrallel with each set consisting of 2 strings of 15 LEDs in series?? I'm trying to figure out how you could get 60 LEDs in series going on 120VAC, the combined forward voltage for 60LEDs would be around 192V (assuming a forward voltage of 3.2V per LED). Could you add a cct diagram showing the electrical connections? Concerning the 120VAC supply, I'd recommend another warning that when adjusting the beam of the LEDs, that the ring will be at mains voltage.

[arcticpenguin]

rgbphil, Thanks for your comments. At the bottom of Step 2 I've included a pdf file that is the schematics. I tried doing up a pix of it but the text is not clear enough. Once you see the pdf you'll see the circuit is rather simple. That with article I quoted will explain the theory better. When adjusting the beam of LEDs the ring is at Mains voltage but no exposed metal is present, it is all taped and insulated behind the whiteboard. Thanks

[coretj]

 On David's site (exact same schematic) he never mentions going over 30 LEDs for the circuit.  Also, once you go over the 30 LEDs you are giving less power to each LED thus limiting it's brightness as well as it's useful life. ..   I would like to know if you used 2 circuits of 30 LEDs.. (2 CAPS , 2 Resistors , and 60 LEDs) or if you made one circuit of 60 LEDs (1 CAP, 1 Resistor , and 60 LEDs.)?  

But from the pictures in step 5 I think it is 1 circuit of 60 LEDs.
 Have you noticed any dimming?
Have you tried doing string of less than 30 LEDs?
Would you be willing to do an experiment to see if you gain brightness from making 2 30 LED circuits vs. 1 60 LED circuit?

[skarfac3]

how did u manage to calculate the lux level....do you have a tool for these? 

[arcticpenguin]

I have a small lux meter that was calibrated a while back.  Not an expensive item so the accuracy may not be too precise. But for a quick measurement works well. 
The light output of the lamp is very dependant on the LEDs being tweeked for minimal dark spots. 

I still use the lamp for magnified viewing of smal things. Like reading part numbers on small components, too small of writing on labels, color codes on small resisiters and surface mount components.

[Entropy512]

In terms of circuit architecture, you might want to take some inspiration from the various Christmas light strings out there.

Most of these use more LEDs in series (often 30 or so) along with a bridge rectifier.  This results in a 120 Hz flicker rate (since the LEDs illuminate on each half cycle), along with more light from the LEDs (since the duty cycle is doubled).  It sounds like you have a 15-series configuration per polarity?  Most don't include filtering of the rectifier output like another commenter suggested, partly because this is difficult for end-to-end stringed LEDs where the bridge rectifier is actually split (half on one end, half on the other).  However, if you can properly calculate the resistance needed with a filter cap involved, you'll be able to get even more light from the LEDs by bringing the duty cycle to 100% and eliminating abuse of the LEDs with high peak currents.

Also, you'll have far fewer fitment problems if you use 5mm or 3mm LEDs, which will also allow you to use more LEDs for more light.  mcd is an AWFUL measure of LED output, since it takes into account the beam pattern of the LEDs.  The only reason the 10mm LEDs have higher mcd numbers is because larger optics -> narrower beam pattern.

[arcticpenguin]

You are very likely correct in this theory.  I'd be interested in hearing the practical application comparison of differences when you finish building one with this circuitry.  A head-to-head comparison is always interesting, so you may wish to also make one as this instructablle is to compare with. As well the cost difference of using more LEDs etc per usable light produced.  So far no dead LEDs anyway. 
I admit mine could be slightly brighter and I could simply add the remaining 70 LEDs in the bag of 100 I bought would do it as well.
There may even be a market out there for someone to manufacture a snap in replacement for the glass lamps.

 

[bilti]

Is it possible to use a bridge rectifier? the blinking at 25 rate is not confortable

[n0ukf]

Where are you finding 25Hz AC? Most countries use either 50 or 60Hz.

[ethersecret]

great instructable...
I got a question regarding the resistor, its driving me crazy. To me this is how a resistor is used (quick googlezia result) http://unclean.org/howto/led_circuit.html
Volts = Icurrent Rresistance

so lets say if I take resistor out and measure voltage drop with a multimeter between cap and leds and if I do the same measurement between cap, leds and resistor v drop will be greater, ohms laaaaww, right?..oh and after resistor current is..wait, zero or pretty close to?

thx

[bilti]

Hello, How many led`s I can use if I have 220V?

[Dr.Bill]

try use a transformer to lower your voltage

[Derin]

double of 110V

[n0ukf]

More specifically, at double the voltage, you use twice as many LEDs per series string. Or you increase the limiting resistor to counter the higher supply voltage by dropping more voltage at the same current. Remember to also increase the power rating of the resistor if necessary based on this voltage drop and the current it's passing trough the LEDs.

[arcticpenguin]

I have no experience with 220V here. Having said that I refer you to the information in step 2 where I quote the article that supplied the schematic and circuit description. My understanding that within reason there is no limit as we are powering by constant current. In paragraph 3 of step 2 I provide a link using the words "DiscoverCircuits". Just click there and you can possibly email Dave Johnson himself for help.

[forced_to_make_an_account]

"Because unless they build one, they will never get one." <-- I love it. Keep on building!

[soktha]

I've just followed the diagram designed by David A. Johnson and adapted to 220 vac ( France ).
the capacitor is not polarized ( 400V capacitor, 0.22uf ) the resistor is a metal oxyd ( 1K 1W ).
also for the second experimentation, i've put 72 Leds ( always followed the provided diagram ) the lamp works fine.

diagram and detail : http://www.discovercircuits.com/H-Corner/AC-Powered.htm

Again, thanks to David A. Johnson.

Soktha

[Derin]

the same cap system is also used inside a hairdryer since there is a 12VDC motor inside

[skrubol]

Unless you're trying to limit non-LED components, a bridge rectifier would make for a better operating circuit (convert to DC.) Run the AC through the bridge, put the cap in parallel with the rectifier output, and the resistor in series (current limiter, necessary for LED's.) You could also put an inductor in series with the input of the rectifier to limit the current instead of the resistor (more efficient,) but the math is more of a pain, and I think it needs to be a very large value inductor (which means the inductor is large and expensive... On second thought, just stick with the resistor.)

[July Dragon]

Great work! But have you try leaving the lighting ON for a long time without switching it OFF and see that all the LEDs are working fine? Thanks.

[lrooff]

It's a neat project and should reduce the use of electricity for the lamp, but as far as keeping mercury out of the environment, it's likely to have the opposite effect for the simple reason that the florescent bulb which would otherwise remain in use in the lamp for several years is now going to wind up in the dump right away.

[arcticpenguin]

I'm not advocating getting rid of a fluorescent lamp to place LED's but only when one has to be replaced anyway. Everyone should be aware by now that fluorescent lamps are mercury safe as long as they are not broken. Once it burns out it must be treated as hazardous waste and needs to be taken to a handling facility. It is unlikely that most people will do this though as they don't do it for fluorescent stick lamps yet.

[Derin]

im sorry but we don't have handling facilities here

[ScottSEA]

An excellent and useful instructable. Well done.

[Stokes]

The schematic shows a polarized capacitor... is that right?

[arcticpenguin]

I see what you mean. Though, in my day this was the symbol for any capacitor and there would be a dot or small plus sign to label polarity.

http://library.thinkquest.org/10784/circuit_symbols.html

This link shows the symbols I'm used to as far as capacitors go anyway.

Because of the written circuit description that I found I went with a non-electrolytic. I believe it is because an electrolytic would not work properly as the current limiting while on AC.

Thanks

[Stokes]

I stand corrected. Knowing that the symbol with one curved end isn't necessarily polarized makes some other schematics I've seen make a lot more sense. I think I have all the parts sitting around to build a small version of one of these; I'll have to take a look!

[offlogic]

Loved your Instructable! I have two such fluorescent magnifiers at home, and a number of them at work. The work lamps are all "O.C. White" brand, which are fine lamps no mistake there, but they use a custom lamp size with a list price of $30 to $40 each. Outrageous! Next time we need a new "bulb" I'll whip out an LED array like the one you speak of....

[GorillazMiko]

Wow, great job! I really want one of these in my rooms, but that would probably be a bit too hard for me to make. Still, great job.

[Spokehedz]

have you checked out some of the constant current LED sources on DealExtreme lately?

They seem like an ideal part for this project. Plus, you could wire the LED's in series, so you could save on a bunch of wiring--and you would be able to use the high-brightness Luxeon LED's. Instead of hundreds of LED's, you just need one.

http://www.dealextreme.com/details.dx/sku.11077

I was thinking about making my own lamp from one of these. Replace a Halogen light with an LED.

[arcticpenguin]

Interesting item for sure. I wonder what is actually inside the $9.15 white box. The circuit I'm using is a simple capacitor to limit current to about 20mA. Change the value of the capacitor will change the current limited. Cost about $2. Then I looked up the Luxeon LEDs you mention and the largest ring they have is only 5.5 inches diameter, much smaller than the 8 inch I required. That lists for $160. Even the cost of the pucks are $17 to $20 each. For the cost of 2 pucks I can buy a used lamp and totally convert it. For the cost of a ring, even though it is too small, I could by 2 new circular Magnifier lamps, more if I go to ebay. While it is a nice idea it's just too darn expensive for my tastes. Maybe worth it for a professional technical application but not for the hobby. An LED is a focused light rather than a halogen, fluorescent and incandescent. You'll find for most applications you'll need the 360 degree dispersion. Thanks

[LinuxH4x0r]

Great Idea! Much better than those florescent tubes. +1

[arcticpenguin]

Maybe not as cheap initially but when you consider decontaminating a mercury spill? Definately more fun though. I figure eventually something like this will be commercially available, but until then. Thanks.