PDA

View Full Version : The circle LED mill spindle light -- update



Bill Pace
02-27-2009, 11:20 PM
Problems did crop up with the 'umbrella' circle LED light and though it involved the power going to it - I'm not not up enough on 'lectrics to say just what!

Heres what transpired -- had that 9v wart on it, and the first time I was actually using the mill for a good while leaving the light on, I noticed (after an undetermined time -- 30min/hour? that it was out. On checking it over determined the wall wart was dead and the lite was OK, so tried another wall wart of 5v/10a and after about 5 min the light would start to do do a slow blink -- could not find a resistor in my junk that would change any thing enough, So, based on one of the previous posters suggestion (crrmeyer/Charles) I went to Radio Shack and got a 4 1/2v/1600ma wall wart.
http://www.radioshack.com/product/in...ductId=2049707 (http://www.radioshack.com/product/index.jsp?productId=2049707)
Hooked it up this morning and its been on now some 12-14 hours with nice bright lite, no blinking and no discernable heat off the bulbs, so it would seem this is gonna work.

So you guys that ordered one of the lights ....well, be prepared.

Another glitch arose with using hot glue to stick it -- that didnt work either:( -- that chinese metal was just toooo slick:D The obvious answer to that is, of course, huh Frank Ford?, to use 2-3 little button magnets.

BadDog
02-28-2009, 03:13 AM
I've got a 5V Wall Wart set aside of this project. What would it take to make it work correctly?

Evan
02-28-2009, 05:26 AM
Limit the current to the amount the light normally draws from batteries. The wart also must be able to supply that amount of current comfortably so it will need to be about a 800 ma or better rating based on the reported current in the previous thread. One easy trick to use is to drop the voltage with a diode. Any 1n4000 series diode will work. Radio Shack has them in packs of 20 or so for a couple of bucks. Each diode in series with the supply will drop the voltage by about 1/2 volt. Since the LEDs are all in parallel dropping the voltage has the effect of limiting the current. White LEDs need about 4 volts DC minimum to work properly.

Doc Nickel
02-28-2009, 06:15 AM
I spotted an LED ring-light at the local Shucks auto supply (also known as Checker and a few others, in other places) and gave it a look. This particular one, unfortunately, was too small an ID to go around my Bridgeport-clone spindle, but probably would have been fine for my smaller mill-drill spindle.

Just throwing that out as a possible alternate source, if anyone needs one.

The best light I ever saw was (as was mentioned in an earlier thread) one of those ring fluorescent bulbs. The fellow I saw using it had it on one of those Grizzly mills with the ram-mounted head (Millrite clone) and was doing woodworking with it (as Griz advertises the mill for.)

So he didn't have any real swarf protection for the bulb, and all he did was paint the top half black.

I'm sure somebody could fab up an aluminum-and-plexiglass housing for one for a metalworking application, but on the other hand, it'll start getting even bulkier that way.

Doc.

John Stevenson
02-28-2009, 06:44 AM
One problem I can see with this setup is that often when I'm keywaying shafts unless I use very long cutters I finish up with the extended quill or head bumping a fan or similar on the job.

Anything this size mounted on the head or quill would limit what I need to do as not everything is flat and stands proud.

.

Your Old Dog
02-28-2009, 07:05 AM
I got mine mounted yesterday. I used the magnets. I'm using a small plastic battery holder similar to what radio shack sells on mine. The advertisement says it will run 30 hours on 3 AA batteries, that's good enough for me.


BTW- I made a AA holder for my chi-com digital calipers on my mill and haven't changed a battery in about a year even with our cold winters.

danlb
02-28-2009, 05:00 PM
The claim for 30 hours on a set of batteries, will doubtless be based on "till it's too dim to see". LEDs pull progrssively less current as the battery voltage drops. What is quite bright at first will often fall off to a mediocre glow in 1/2 of an hour when pulling at 800ma from a AA battery pack. At the end of the hour the current will drop to more like 150 ma or so. This is what lets them make such long life claims.

As a certified flashaholic (flashlight freak) I've worked with a lot of LEDs on my machines. I have a small LED light on my micro mill and another one on my lathe. Both are driven by leftover wall warts.

I can help with formulas if needed.

Dan

nheng
03-01-2009, 12:09 AM
One way to build an ultra slim light would be to use surface mounted LEDs on a circular PCB and then drop that into a wide face groove on the back of a 1/8" - 3/16" thick polycarbonate ring. The back could be sealed with whatever will withstand oil and coolant.

The surface mounted LEDs will lack the lensing of the T5 LEDs but will cover a large area with diffuse light with the lack of aiming made up by the quantity of LEDs.

Incidentally, for sharpest image, a fluorescent ring or LED ring are not the best solution. Separate light sources from several directions create the "interference" necessary to give the eye and brain a sharper image.

But back to the super thin ring, maybe a group project (buy) is called for if there is interest :)

Den

Evan
03-01-2009, 01:13 AM
I am working (have been for a while) on a lathe work lamp that uses 3 1 watt medium white emitters that cost about 2 dollars each from DealX. It can easily be adapted to work on a mill as it is a triangular configuration with a hole in the middle. The actual light assembly is 1/4 thick with small shields that project another 1/4 on each LED. I'll post a picture later or tomorrow.

There are some cheap and effective solutions to the power issue too using regulated micro power supplies that cost a few dollars. More later.

Peter.
03-01-2009, 03:43 AM
As a cheap & safe power source for lighting LED's would it not be possible to power them somehow by winding the lead around the power cable for the machine? I recall you could power an LED by wrapping wire around a soldering iron.

Evan
03-01-2009, 04:26 AM
Nice thought but a little too simplistic. Wrapping a wire around the outside of the cable won't pick up any power. To do that you need to wrap a wire around a small chunk of iron and also wrap just one of the current carrying conductor around that piece of iron too. That does work and is called a transformer. You are possibly thinking of lighting up a mini neon bulb using stray electric fields near a current carrying conductor. The amount of power available that way is thousands of times less than what is needed.

BadDog
03-01-2009, 04:27 AM
So, I freely admit to being out of my depth here. I do have the background (though not used in MANY years to eventually research and figure out the answer, but it's beyond the point of diminishing returns for me, so I'm hoping someone with fresh and available knowledge can provide a leg up, so to speak. :D

So if I substitute a 4.5V xformer, will that be sufficient? Or is there still too little resistance in the circuit and damage would result. Seems like a 5V circuit limited to drop the voltage would be a better choice. The xformer I have on hand is easily able to handle the amperage, and the voltage could be tweaked any number of ways. Being a static "load", it could be a simple resistor easily calculated, or a varistor for "dimming" and fine control, or perhaps a control diode as Evan describes. Any problem with this?

And I've run into situations like John describes. And for that reason, I was already planning to do a "twist lock" mount for easy removal. You know, holes for screw heads with slots (arcs in this case) like used to mount power strips and such. I thought about magnets, but I do too much steel and have enough problems with swarf intrusion and such as it is, I'm not adding any magnets even shrouded by the main light fixture. If I need to remove the ring, I still have the 300W articulated arm light I'm using now. It's just that the harsh Shadow and heat of that thing causes it's own problems for general every day use, particularly when I have low precision work just cutting to a layout line or prick punch.

Speaking of the fixture itself, I'm pondering whether to use the plastic back, or make an aluminum replacement. But I'm too lazy to drill all those precise holes, so I'll probably use the plastic front as long as it seems to hold up.

Evan
03-01-2009, 04:51 AM
I can't give you an exact answer. The method of wiring all the LEDs in parallel to run them from batteries is a NOT recommended way of powering LEDs. It is commonly used on cheap battery powered LED lamps and flashlights where saving a penny or two is the difference between profit and loss.

The reason I can't give you an answer is that this method of powering an LED or a bunch of them is unpredictable and unreliable. Even with batteries there is the possibility of damage from excess current. Or, they must err on the safe side so much that the LEDs are running at much less than rated current. The problem is that an LED doesn't conduct like a resistor.

With a resistor if you up the voltage the current is directly proportional to the increase. With an LED once it is conducting and making light even a very small increase in voltage can result in a very large increase in current. It's like raising the water behind a dam. Nothing flows until it is over the top but then even a slight rise results in ALL the extra water running over.

LEDs of the same exact type will vary from batch to batch in exactly what voltage it takes for this to happen.

The proper way to run LEDs is with a current limiter for each LED or string of LEDs in series only. The only way to determine a safe value of current for your lamp is to measure how much it draws when hooked to your power source. Then multiply 20 milliamps (0.020 amps) times the number of LEDs to arrive at the maximum safe current. Then reduce that to about 75% to account for variations in the individual LEDs and changes in supply voltage. If the lamps draws that much or less when connected to the wart then you should be OK. If it draws more then try a low value resistor or a diode in series with the power lead. Measure again until you get it below the calculated maximum.

Peter.
03-01-2009, 05:29 AM
What happens tif you over-volt LED's? Do they blow or have reduced lifespan?

Weston Bye
03-01-2009, 08:54 AM
What happens tif you over-volt LED's? Do they blow or have reduced lifespan?
Depending on the degree of overvoltage/overcurrent they may be converted into DEDs (Dark Emitting Diodes) or, in extreme cases, NEDs (Noise Emitting Diodes). The lifespan of a NED is very short, and the NED phase may coincide with a PED (Particle Emitting Diode) mode, followed by a permanent DED condition.

John Stevenson
03-01-2009, 08:58 AM
Depending on the degree of overvoltage/overcurrent they may be converted into DEDs (Dark Emitting Diodes) or, in extreme cases, NEDs (Noise Emitting Diodes). The lifespan of a NED is very short, and the NED phase may coincide with a PED (Particle Emitting Diode) mode, followed by a permanent DED condition.

In which case you need a SHED [ Sh*t happens Emitting Diode ]

.

Evan
03-01-2009, 10:10 AM
What happens tif you over-volt LED's? Do they blow or have reduced lifespan?


You can overdrive a white LED by quite a lot since they are made from a tiny chip of silicon carbide. The life span is inversely proportional to the temperature of the chip. With a nominal drive current of 20 ma a standard 5mm white LED might last 1000 to 10000 hours depending on quality (100,000 hours is total BS for this type of LED).

If you overdrive it slightly, say 20 to 25 ma you will gain a little brightness and cut the life in half. Drive it at 40 ma and it is much brighter but the life may be only 100 hours depending on the cooling of the device. Drive it at 50 ma and it will live for a few minutes. They can withstand very short pulses of current to about 100 ma if the average duty cycle adds up to no more than the nominal 20 ma. This is a way to gain apparent brightness with most LEDs but it doesn't work with white LEDs.

White LEDs are mostly made using an ultraviolet LED that shines on a glob of phosphor which converts the light to visible range, just the same as a fluorescent lamp. The phosphor averages the output so no apparent gain is visible with the pulse method.

Your Old Dog
03-01-2009, 10:23 AM
Depending on the degree of overvoltage/overcurrent they may be converted into DEDs (Dark Emitting Diodes) or, in extreme cases, NEDs (Noise Emitting Diodes). The lifespan of a NED is very short, and the NED phase may coincide with a PED (Particle Emitting Diode) mode, followed by a permanent DED condition.

You got any Wiki's to support your claims? :D

Peter.
03-01-2009, 10:29 AM
Depending on the degree of overvoltage/overcurrent they may be converted into DEDs (Dark Emitting Diodes) or, in extreme cases, NEDs (Noise Emitting Diodes). The lifespan of a NED is very short, and the NED phase may coincide with a PED (Particle Emitting Diode) mode, followed by a permanent DED condition.

I have dark-emitting diodes in my security cameras :D

Evan
03-01-2009, 10:34 AM
When I was a kid I would con my dad (easy) to take me on tours of various semiconductor manufacturers in the area (silicon valley). On those tours I would ask the guide (engineer with nothing better to do) if I could have some samples. I sometimes came away with a shoebox load of fallouts, mislabled, malformed epoxy but operational etc parts. Since I had a shoebox of assorted diodes, maybe 50,000 or so I had a few to spare for "life testing".

It turns out they make excellent fuses when connected directly across the mains with a heavy duty foot switch. Safety glasses are recommended as they make a pretty violet flash and a loud report as the bulk of the device vaporizes in a microsecond ( similar to a micron).

Peter.
03-01-2009, 10:39 AM
As a kid I discovered that even blown LED's would explode when connected to mains elecricity :D

Weston Bye
03-01-2009, 11:35 AM
You got any Wiki's to support your claims? :D

No, just empirical and ancedotal evidence - almost as good as a Wiki.

BadDog
03-01-2009, 02:06 PM
That's a great explanation (and analogy) Evan, thanks. And thanks to all the others who have contributed too...

Your Old Dog
03-01-2009, 02:15 PM
I think we should up the anti to play in the "I got Led light and you don't" game :D How about solar powered like this guy did for his 4.5V radio using garden led light kits. I may just try that as I'm battery operation already and I do have a none working garden led light that takes three batteries!

http://www.gotwind.org/diy/DIY_Solar_Powered_Radio.htm

danlb
03-01-2009, 03:21 PM
The degree of damage from over voltage depends on the amount of voltage.

When wired in series, each LED is acting like a variable resistor, with the resistance dropping as the current (or heat) rises. The led Christmas lights are simply 80 or so LEDs in series. Since they are diodes, 1/2 of them have their polarity reversed.

When you supply just a bit too much voltage, the LED will simlpy change color. Often turning a bluish hue. This is because some white LEDs are an ultraviolet light with a white phosphor, but most are a blue LED with a yellow phosphor.


Most 5mm LEDs are supposed to run at only 20 or 30 ma EACH. Typical voltage is 3.2 to 3.5 volts.

If you want to use a 5 volt power supply (wall wart) you need to add a resistor to limit the current to the max for the LEDS or the max for the power supply, whichever is lower.

Voltage / amps = resistance, so...
assume the LED needs 3.5 volts to use 800ma. and your wall wart puts out 5 volts. 800ma is .8 amps. You want to eat up that extra 1.5 volts in a resistor. The calculation is:

.8 * 1.5 = 1.2 ohm resistor.

Dan

Ryobiguy
03-01-2009, 03:29 PM
A few years back a friend wanted to build some LED lighting for camping.
I asked the electronics engineers at work what I should do. They figured out the voltage drops for series strings for each color, ie, something like 5 reds in series and 3 greens in series. That got the voltage just right.

However, they neglected to think about and/or mention the current limiting that is required. Each single or series chain of LEDs needs a resistor to limit the current - they'll suck up as much as the voltage will pull.
In my friend's case, the LEDs lit up REAL bright for a few seconds and then dimmed down as they got hot and were converted into DEDs.

Since (regular) diodes have a forward voltage drop, can you series a bunch together with a resistor, and use that as an over-voltage shunt to maintain a maximum voltage level? Or is that just going to drain current once the voltage is high enough to conduct through the string of diodes?

Or would it just be easier to add over-voltage protection with a voltage regulator widely available from Radio Shack?

-Matt

Evan
03-01-2009, 03:40 PM
The degree of damage from over voltage depends on the amount of voltage.



It depends on the current. The voltage drives the current but an LED acts as a constant voltage device once it conducts. When you increase the voltage it's the current that goes up while the voltage stays nearly the same depending on the power supply "stifness". You can't determine the current by measuring the voltage across an LED. It's a non linear device.

Evan
03-01-2009, 03:54 PM
Why buy from Radio Shack when you can buy direct from their source? Dealxtreme sells every possible LED related item you can want at 1/10 the price of Radio Shack and with free shipping.

http://www.dealextreme.com/products.dx/category.917

To regulate current you need either a resistor selected for the particular load or you can use a cheap little regulator designed for the job and able to handle whatever load you apply.

For instance, here is a driver board that takes 85 to 265 VAC input and regulates it to drive LEDs at 650 ma and it cost $3.32 with free shipping. Sounds perfect for the ring lights. Cheaper if you buy three or more.

http://www.dealextreme.com/details.dx/sku.13552

danlb
03-01-2009, 04:38 PM
It depends on the current. The voltage drives the current but an LED acts as a constant voltage device once it conducts. When you increase the voltage it's the current that goes up while the voltage stays nearly the same depending on the power supply "stifness". You can't determine the current by measuring the voltage across an LED. It's a non linear device.


I was over simplifying. Probably should have been more precise. Since current follows voltage in this kind of circuit (once it starts conducting) and more people are familiar with the term, it's easier to just say "voltage".

That deal extreme Constant Current Regulated LED Driver sounds like it would work well for 20 to 35 5mm LEDs in series, depending on how high you want to drive them. It could also power two 1 watt leds in series.

I'm always careful when ordering from dealextreme. Of the last 5 orders I placed, they forgot to ship parts of 3 of them. Quite a hassle when you have a 2 week delay due to international shipping. I still order from them anyway. :)

Dan

Evan
03-01-2009, 05:10 PM
I have ordered three times from them for a total of about $500. I have received some items that were exactly as junky as they said they were. No surprises but they did forget to ship one item on the second order worth about $7. I didn't bother claiming it, I have saved much more than that compared to buying the exact same stuff from a North American source. My last order got here in six days and none of them have had any duty or customs fees or inspection. That part I can't figure out.

The regulator at the link I posted is for driving in parallel, not series.

Your Old Dog
03-01-2009, 07:12 PM
Evan, is DealExtreme in the states or Canuckdia? I'm really interested in this.

http://www.dealextreme.com/details.dx/sku.4355

Imagine, peace and quiet during lunch in a restaurant or not having to hear Valentines Day conversations in the waiting room to the doctors office!!

dockrat
03-01-2009, 08:07 PM
Evan, is DealExtreme in the states or Canuckdia? I'm really interested in this.

http://www.dealextreme.com/details.dx/sku.4355

Imagine, peace and quiet during lunch in a restaurant or not having to hear Valentines Day conversations in the waiting room to the doctors office!!

YOD I have looked into those things also. Unfotunately in the USA and Canada, they arn't legal. Stiff fine in USA. Not sure about fine in Canada.

"In the U.S. the FCC (Federal Communications Commission) makes certain frequencies available to broadcasters for public use. When an end-user pays to use that spectrum, jamming the signal is paramount to 'property theft.' The FCC is also concerned about potential "leakage" of jammers interfering with frequencies outside the range of cell phones, like garage door openers or medical equipment; and it's worth noting that over 100,000 emergency calls are made each day from cell phones. Anyone caught manufacturing, selling, owning, or using a jammer in the U.S. is punishable by an $11,000 fine and up to a year in prison for each offense."

dockrat
03-01-2009, 08:07 PM
Deleted....double post

lazlo
03-01-2009, 08:23 PM
Evan, is DealExtreme in the states or Canuckdia? I'm really interested in this.

http://www.dealextreme.com/details.dx/sku.4355


I posted that cell jammer awhile back. DealXtreme is based out of Hong Kong, and everything drops ships to the US.

Despite what they say in the ad text, they ship the cell jammers, and the "illegal" laser diodes to the 'States.

The cell jammer works great, or so I'm told ;)

danlb
03-01-2009, 09:01 PM
Just be aware that using that cell jammer may keep a person from getting a business call they need to stay in business, or might block a 911 call. Doctors use cell phones a lot.

There are more than property rights and obnoxious conversations at risk.

Evan
03-01-2009, 10:12 PM
Not sure about fine in Canada

Intercepting and/or interfering with telecommunications in Canada carries a federal maximum fine for individuals of $5000 per day of interference and up to a year in jail. Also, you may be fined for operating an unlicensed transmitter which is also $5000 and a year in jail. In BC the province has a law that provides for ticketing as a statutory offence with a fine of $288 for jamming cell phones. Not a good plan.

lazlo
03-01-2009, 10:17 PM
Intercepting and/or interfering with telecommunications in Canada carries a federal maximum fine for individuals of $5000 per day of interference and up to a year in jail.

Similar deal in the 'States. But DealExtreme has sold tens of thousands of those jammers around the world (they're quad-band jammers) and no one has been arrested. Yet. :D

darryl
03-02-2009, 01:51 AM
Back to the leds- if you series connect 4 leds, the voltage required to run the string will average out fairly well to what another string of 4 will require. Given this, you can parallel several strings of 4 per string and drive the total of them using a single current regulator. Set the regulator to deliver 20ma per string, so for example 6 strings would require120 ma. The regulator needs to use up some voltage so the power supply would have to deliver about 16 or 17 volts minimum under load. Many 12v warts could do this, and a 15 volt wart would be guaranteed to drive the network of leds through the current regulator. If you needed to be sure a 12v wart would work, then limit the number of leds per string to 3. A 24 led ring light would then have 8 strings of leds and would require 160 ma from the regulator.

This regulator can be as simple as one LM317 chip and one resistor. For 160ma for example, the resistor value comes to 7.8 ohms, so an 8.2 ohm resistor would work fine. The formula is easy- resistor value = 1.25 divided by the current required. After the calculation is made, round the value up to the nearest higher standard resistor value.

The more leds you can wire in series, the more the variations in voltage drops will average out and the more even the light output per string will become. If you have a 24 volt wart, for instance, it will likely have about 27 or 28 volts available under reasonably light loading, and could therefore drive about 7 leds in series per string. In this case, a ring light could be made using say 4 strings of 7 leds, and then would need 80ma of current- thus the resistor value would become 15.6 ohms. You would be safe to use a 15 ohm resistor in this case, which is another standard value. The current per string has jumped to a whopping 21ma per string, close enough to the suggested draw as to be safe. To keep the draw to 20ma or less, the nearest higher resistor value would be 18 ohms, and the current per string would then become just over 17ma per string, close enough.

In any event, pretty much anyone could put together this current regulator circuit, which is just two small parts, and you could get to use pretty much any wall wart you might have on hand. I don't know what those power circuits are like that Evan posted a link to on dealx, but it's possible they present a voltage hazard in the event you accidently or otherwise contact a bare wire- you're pretty much protected from that possibility if you use a wart.

John Stevenson
03-02-2009, 04:57 AM
Got one of those 16 bulb led flashlights here and as an experiment looked at adapting it to the mill to replace the two lo-volt 24v incandescents that have been fitted since new.

However the light they put out is feeble in comparison, now I know they use less power but so do my existing machine lights if i don't switch them on.

The reason I have lights on the machine is to see better what I am doing.
I don't mind putting time and effort into a mod IF it has an advantage but I can't see it with these for what I want i.e. bright light where I need it.

To be honest all my machine lights are crap, they are all old and dated and I need at least 6 to start with. Even 6 import ones are a fair wedge of money, the import heads aren't too bad but at the moment none of the UK importers have any in.

I'm going to investigate some of these small halogen lights the local car park racers are fitting to their cars. I have seen them from 2" up to 6" but the 3" to 3-1/2" size looks very adaptable.

Most of my machines have a 12 / 24v power supply as standard and the arms are still decent it's just the open design of head and hard to source bulbs that don't last that are a problem, the halogen bulb units can be bought for literally pence, 7 for 10 off Ebay.

Anyone done this ?

.

Evan
03-02-2009, 05:21 AM
I doubt you will like it. Halogen lights are HOT. I have one next to my mill but all I use it for is to drape the air hose on as a flexible support. It's much too HOT.

LEDs lamps can work very well now but it's a matter of using the right type. For a work light on a mill you need at least three watts in the form of 3 x 1 watt emitters, or one of the new very bright multiple emitters like this one. It is rated at 300 lumens which is plenty bright for a work light.

http://www.dealextreme.com/details.dx/sku.5876

If you want to buy one ready made then this one looks well suited. Not cheap but it should last. It's 210 lumens.

http://www.dealextreme.com/details.dx/sku.12958

http://metalshopborealis.ca/pics5/dxled.jpg

Your Old Dog
03-02-2009, 07:39 AM
Similar deal in the 'States. But DealExtreme has sold tens of thousands of those jammers around the world (they're quad-band jammers) and no one has been arrested. Yet. :D

That was my thought when I raised the question! It's okay for some loud mouth to interrupt my meal but if I want to do something about it they are automatically life saving calls, yea right. Could it happen, yea I suppose and to use the
old tired phrase, "if it just saves one ________ it's worth it". Maybe I'll eat outside on the sidewalk next to the traffic, it will be quieter in some cases. :D




.................However the light they put out is feeble in comparison, now I know they use less power but so do my existing machine lights if i don't switch them on.

The reason I have lights on the machine is to see better what I am doing.
I don't mind putting time and effort into a mod IF it has an advantage but I can't see it with these for what I want i.e. bright light where I need it................

It pains me no end to have to say this, I get a headache just thinking about it and it's not like the guy has to be right 100% of the time but I have to agree with Sir John on this one.

I did the mod, got it mounted to the mill now but the light it gives out is a little feelble/soft. Turns out some shadows are a good thing as they help you locate or see/descriminate better like swarf or very fine marks. I think I'll move mine to my drill press and get something else for the mill, maybe two 12V tractor backup lights or something of that order.

Bill Pace
03-02-2009, 09:48 AM
It pains me no end to have to say this, I get a headache just thinking about it and it's not like the guy has to be right 100% of the time but I have to agree with Sir John on this one.

Argggghh! Im afraid I have to agree here Dog, in spite of the concept, application, and yeah, even the "coolness" factor, the light is a disappointment. It just doesnt come through with doing what I had hoped -- helping with those same 'shadows' that you refer too:(

And I agree on the halogens, tried a couple versions of them and after burning fingers a couple times when reaching to adjust, they were discarded.

At present - and is doing the best so far -- are a couple of $8 Walmart spring clip goose necks with 100w fluorescents , these give good light, are much cooler and also much less affected by vibration/short life.

GadgetBuilder
03-02-2009, 10:51 AM
I have two lights on my milldrill, a gooseneck I added mounted onto the switch box and a parallel arm lamp mounted nearby. The gooseneck is on the left, the parallel arm on the right, this to provide illumination from both sides.

The parallel arm lamp has a CF which is protected nicely by the metal lampshade. The light is reasonably bright and well diffused. This works well so I haven't tried other things here.

The gooseneck initially had a halogen compact flood light, 50w and a couple inches diameter. Worked well and provided great light. It was generally positioned within a foot of the work. However, when working on a setup the heat from the intense light was very uncomfortable on my hands if I fumbled for very long. 30W halogen floods are also available and might be a bit more comfortable.

So, a year ago I got one of the DealExtreme 43 LED 5W lamps that look a lot like the compact floods: http://www.dealextreme.com/details.dx/sku.1475

The description seemed a bit over the top in claiming it could be used for general lighting but my thought was it should work well for lighting the relatively small work area on the mill.

The light it provides looks almost purple in comparison to the CF and is dim compared to the original halogen. However, it can be moved very close to the work for brightness and there's no heat on the back of my hands during setup. One of the 43 LEDs has failed in the year I've been using it.

Can't say I'm very happy with this LED light but then I haven't actively pursued another solution either. The more recent offers from DealExtreme have twice the power but cost more than twice as much so I'm waiting for a glowing user report from this group before jumping on that.

Does anyone have a feel for how the 6500K color temperature for the DealExtreme 10W compares to a CF? Or how the 10W light output from this LED lamp would compare to a 13W CF? My impression from the LED lamp I have is that it is inefficient compared to a CF.

John

Evan
03-02-2009, 11:14 AM
I have some data on that but am off to town right now. Will answer later. Main point: Lights made up of regular 5mm LEDs are not the answer. High power emitters are.

ninthst
03-02-2009, 01:05 PM
Can I substitute a 6 volt gel cell for the 4 1.5 AA's ?

lwalker
03-02-2009, 02:02 PM
I have to concur with Evan on this one. I've looked at a lot of various lights people have built with the T1-3/4 (5mm) LEDs and they are really weak compared to the LEDs that are built for illumination. A single 1 watt LED is painful to look at (the newer ones come with warnings to not look directly at them) and three of them provide plenty of light for milling. Further, there is a ton of optics: lenses, diffusers, reflectors etc available that are designed explicitly for those LEDs.

The 5mm LEDs are fine if you need a fairly low-level, diffused light, but I wouldn't use them for machine work.


ps, for more "fun with LEDs" check out the guys on the www.candlepowerforums.com Custom & Modified forum. There are some lunatics building 500W portable LED lights :-)


I have some data on that but am off to town right now. Will answer later. Main point: Lights made up of regular 5mm LEDs are not the answer. High power emitters are.

lazlo
03-02-2009, 03:13 PM
And I agree on the halogens, tried a couple versions of them and after burning fingers a couple times when reaching to adjust, they were discarded.

I have the 100W Halogen machinery lights that Enco/MSC have on sale every month. They're Made in USA (I can't remember the name). They have an IP60'ish head, with a thick rubber grommet around the perimeter. You can reach up and move it around without getting burned. Very bright -- highly recommended.

As far as hot, this is Texas. We have no problem with hot :D

Evan
03-02-2009, 03:27 PM
Ok, back for a while.

This is a bulb that I bought from DX to use as a reading light. It fits any standard MR16 base 12vdc lamp including track lights and reading lights. It's a 3 watt single emitter LED flood light with a perfectly even 60 degree beam spread. Absolutely no hot spots or holes. I have had it since last summer and generally leave it on 24/7. It shows no sign of dimming. Construction is top quality with a solid aluminum heat sink body.

http://www.dealextreme.com/details.dx/sku.3264

I took my reading light down to the shop and placed the lamp about where I would want a work light. I took two pictures with my Nikon on manual so it used the same settings for each image. The only adjustment I made to the images was to resize them for web. The pictures were taken with all my shop lights on and I have bright shop lighting.

Without LED

http://ixian.ca/pics6/led1.jpg


With LED

http://ixian.ca/pics6/led2.jpg

This is why I am recommending a minimum of three watts, either 3 one watt emitters or a single 3 watt.


Can I substitute a 6 volt gel cell for the 4 1.5 AA's ?


No, not without some sort of current limiter such as an appropriate resistor.

John Stevenson
03-02-2009, 05:40 PM
Evan,
Colour me confused? although that's not hard to do at times. :rolleyes:

Looked at the DX link but you say
"This is a bulb that I bought from DX to use as a reading light. It fits any standard MR16 base 12vdc lamp including track lights and reading lights."

So two questions, is it a complete lamp or a bulb ?
Reading the feedback about size it looks to be a complete lamp

and the advert says 12 AC or DC so will this run off the machines 12v AC lo-volt rail or require a balanced circuit ?

.

Evan
03-02-2009, 05:47 PM
The pictures above show the lighting from the linked MR16 LED bulb. Not a lamp but can be installed in any lamp that suits that bulb. It is designed for a well regulated 12 vdc so if you can provide that it will work. I wouldn't try running it on straight 12 volt ac as the flicker would be horrendous, especially on 50 hz. Any well regulated 12 volt supply such as a computer power supply will run it just fine. 3 watts is nothing, about the same draw as a night light.

John Stevenson
03-02-2009, 06:39 PM
Ahha

Got it, just spotted the pins, at first glance it looked as if the bulb assembly was the lamp.

.