Power for Siemens LED backlight?

Charles Krebs · Post by **Charles Krebs** » Sat Dec 17, 2011 2:37 pm

This little item came in along with another item I purchased. Very nicely made, and I can see some uses for it, so I don't want to "fry" it when I try to power it up.

In the second image you can see "12VDC/440mA".

Does the "440mA" indicate the power it will consume, or does this indicate that I need to use a current regulated power supply to limit the current to 440mA? (Like the BuckPuck I use for my microscope LED)

As you can see in the last image there do not seem to be any other components inside other than the 64 LEDs, so I guess I will need a power regulated supply. (I can't see behind the array... no screws to remove and it appears to be nicely fixed in with some adhesive, so I would prefer not to try to pull it out).

I've got a couple of unused external hard drive power supplies (12V 2A) laying around. Any way to simply and safely use them?

Ancient1 · Post by **Ancient1** » Sat Dec 17, 2011 7:28 pm

Hi Charles,

The way that the power requirement is specified in the U.S. is:

Operating Voltage/Current Drawn, from source, at the operating voltage

So, you should be able to power the array from a 12VDC power supply that is capable of delivering at least 500ma.

There is something that looks like a ballast resistor in the circuit, so I would just run it from a 12V power supply if it were mine, but it isn't.

To be safe, you might want to power it from a lab supply. Set the voltage to 12V and current limit to minimum. Then, slowly increase the current limit. The current, as indicated on the current meter, should increase up to 440ma and then stay there as the current limit is further increased.

g4lab · Post by **g4lab** » Sat Dec 17, 2011 7:43 pm

I would guess that the 440ma rating is for continuous duty. My observation has been that it is truly amazing how badly LEDs can be overdriven before permanent damage is done. The LEDs are in an 8x8 array. Presuming they are wired in a square (which they may or may not be) that might mean each LED line is drawing 55 ma. A bit high but by no means out of range. Regular indicator LEDs have continuous current ratings from 10 to 30 ma very commonly. I have abused such LEDs running them at 150 ma expecting to cook and they don't. With so many devices to absorb the energy you are not going to cook it on a 12 volt supply. But you might want to make sure its really 12 volts (as opposed to 13.8 v for a car battery eliminator for example)

Similarly if they are wired as a square (ie. series and parallel eight devices each) a twelve volt supply would drop 1.5 volts each which is within the range one would expect for such LEDs.

I don't see a ballast resistor, just a plug connector.

http://www.microscan.com/en-us/products ... ights.aspx

I would say hook it up to twelve volts and see how hot the LEDs get. The illumination LEDs they make are designed to run a lot hotter than the indicator ones. But if it feels like its heating up too fast measure the current and limit it to 440ma.

I doubt you will have a problem. It almost certainly will not go "Pffftt!"

If you look at the catalog pages, strobe versions of these take pulses near ten amps. I think you could strobe this unit quite hard with 64 LEDs there.

SONYNUT · Post by **SONYNUT** » Sat Dec 17, 2011 8:32 pm

any 12 volt supply that can put out 1/2 amp will do...it will not draw more

Wakkies Norman. · Post by **Wakkies Norman.** » Sat Dec 17, 2011 10:02 pm

What does it say on the black component next to the yellow insulated tag?

It looks very much like a limit resistor, it's in the right place.

SONYNUT · Post by **SONYNUT** » Sat Dec 17, 2011 10:25 pm

it more than likely is...led's only need about 3.5 volts...

Charles Krebs · Post by **Charles Krebs** » Sat Dec 17, 2011 11:03 pm

Good eye, Norman!

Looks to be a fuse... "Buss 3/4A 125V"

Unless they are under the array, there are no resistors I can see.

phytoplankton · Post by **phytoplankton** » Sun Dec 18, 2011 2:53 am

I did a little searching, perhaps this will help.

It seems you have a BL-50 series backlight. There is a PDF with some information including a small bit of electrical information, and the pinout of the 9 pin D-sub, and the power supply accessory:

http://www.nerlite.com/pdfs/BL-50-series.pdf

The power supply listed for your model has a PDF:

http://www.nerlite.com/PDFs/DS-000400 POWER SUPPLIES.pdf

^ Edit: this forum software hates spaces in the URL for some reason.

It's a 12V continuous voltage unit with variable current. Considering that the pinout of the D-sub doesn't appear to communicate any information about the connected light, and that they offer flying lead models, I suspect that your light expects 12v constant voltage.

The absence of obvious electronics suggests to me that they are behind the array. For long life the LEDs should dissipate the heat somewhere. That path is probably behind them and through the case. Such a thermal interface would be a likely spot for power regulation components as well.

Though I would still advocate testing it first on a PS that can limit current to a safe value, as Ancient1 suggested. I would hate for you to blow the unit (or just the fuse) based on our educated guesses.

ChrisR · Post by **ChrisR** » Sun Dec 18, 2011 4:27 am

If t'were me I'd have it in bits to see if I could see the copper tracks.
They probably put 3 leds in series across the 12V, or something like that.
(Though 64/3 doesn't go!)
There might be some small resistors built in to the board.

Characteristics of LEDs are something like this

where at the knee, for a white led you'd expect 4 volts ish at 20 - 30 mA.
So if the voltage goes up the current does too but quicker than it would for a straight resistor.
Leds fail by overheating though so there's time factor. If you can just check the current is within range you're fine.

A note about the power supply though - if you use a $5 plug-in DC supply it might not be smoothed. The average voltage may be OK but it'll look like the red line here:

My green lines are meant to show what happens if it IS smoothed, by a big capacitor. It works out that you need darned big capacitors! The humps are at twice mains frequency, so 120Hz for you (100Hz in uk).

(Car batteries don't care, they act just like a huge capacitor)

If there's NO smoothing the LEDs would go OFF 120 times a second whenver the voltage on each one drops below about 3V.
If the Average (rms) voltage is 12v, that's about 70% the way up the sine curve, so you can see that "below 3V" is quite a long interval per cycle.

I have a reading lamp with LEDS. I can't see the flicker, but a camera can ( so can its LCD screen!) It's useless - except for reading by!

The answer is usually a Switched Mode Power Supply, like computers use, as do some table lamps, because they flicker at much too high a rate to worry anything.

Ancient1 · Post by **Ancient1** » Sun Dec 18, 2011 9:03 am

Connecting the LEDs in series to match the supply voltage is not a good idea, as the forward voltage varies from around 1.7V to 3V for different devices . A properly designed circuit requires current limiting, which can be simple resistors or an electronic current source. Some LEDs have current limit resistors built into them, but most do not. Here is a diagram of what is most likely representative of the circuit.

ChrisR · Post by **ChrisR** » Sun Dec 18, 2011 9:52 am

1.7V to 3V

That's true if you cover all the colours including the (original!!) red and green types, but the "white" ones are much higher and don't vary so much. (Though there's more than one way to get "white".)

As mentioned earlier, Siemens may bave embedded resistors into the board ( they may be Surface Mount so hard to spot). It's commonplace though to NOT do so. If the LEDs don't vary much (so the forward voltages and therefore currents are very closely grouped) then it's fine to daisy-chain them. If the 12V varies a bit, so does the brightness, but there's quite a bit of headroom. If the voltage goes too high the fuse blows!

A1 I see you're assuming only 2.5 volts across each diode at 30mA

I have a very cheap small led grid somewhere, and I've been in its knickers - no resistors.
The bigger "task lamps" usually have a very few, often a single "dropper" for a large number of LEDs in parallel.
(eg 21 rows of 3 leds with one resistor for the lot, then another resistor for the odd one that's left, and that's the one LED which stays on if the voltage drops too low)

g4lab · Post by **g4lab** » Sun Dec 18, 2011 11:46 am

Using a car battery as filter will clamp the voltage at 13.8 volts if its a good battery. This probably won't blow the three quarter amp fuse that is in the circuit.

I don't think there is much need to worry about how its wired. It obviously can take 12 volts. I think Charles would be well advised to use a lab power supply so he can read the voltage and current. They are likely wired in a fashion that will cause them to all experience the same current and hence the same brightness.

The folks that made this thing Nerlite are in the machine vision business and are well aware of the effects of flicker. Suitable filter caps are very easily available because they are widely used by the car audio types. You can get caps rated a couple of Farads at 12 volts.

Here is a budget switching power supply I recently used on an LED lighting project using some very powerful "Far Red" IR emitting LEDs for a botanical project. It has 5mv of ripple:

http://www.testequipmentdepot.com/bk-pr ... s/1550.htm

You could probably find similar supplies on ebay or from electronic hobby places for even less money. BK Precision is a reasonable well known supplier here in the USA. I have had tons of their gear and though it does not have the joy of ownership of Fluke Tektronix or H/P all the pieces I have had did their jobs perfectly.

ChrisR · Post by **ChrisR** » Sun Dec 18, 2011 12:13 pm

I think that paying Siemens for their super power supply, or involving a car battery, or getting a lab supply which I assume Charlie doesn't have, or spending $149 on something else,
would be rather over the top.

I have a video light similar to this:

http://www.ebay.co.uk/itm/CN-48H-LED-Vi ... 2a1714d15b

About $20 for a 64 led version, runs on four AA cells, bright enough for video or ad-hoc table-top stuff, runs for a couple of hours.

I would be happy to recharge a few AA cells!
If not, try a $2 unit eg
http://www.ebay.co.uk/itm/12V-1-000mA-P ... 4aa54c972d

which might give a flicker to deal with,
or try a standard mains supply like
http://www.ebay.co.uk/itm/12V-5V-AC-Pow ... 20c04c227d

They'll supply 12V and 5V DC for a few bucks, and are switchers so won't flicker.

Charles Krebs · Post by **Charles Krebs** » Sun Dec 18, 2011 12:16 pm

I suppose that the key would be to look behind the array.... but I'm thinking it may be "heat-sinked" to the case so I am not going to mess with it. (The casing is fairly heavy metal, with a warning label "CAUTION HOT")

But I think the excellent internet-searching detective work of phytoplankyon provides the answer. (Thanks! I looked and could not find this).

The power supply PDF:
http://www.nerlite.com/PDFs/DS-000400%2 ... PPLIES.pdf

It shows that the power supply for this unit is as follows:
NER-010500301 CPS-12, 3.5A PS, DB9-F, US

So it looks like a constant 12VDC will do it. The 12V 2A hard drive wall-warts I have in a box somewhere might be OK for this... I'll try that first.

And having the pin-out for the connector even saves a little poking around with the multi-meter. All I need to get is a 9 pin D-sub plug.

Thanks all for the input!

I've got a few other LED projects on my list, so it is probably time to start looking at a small lab power supply.

g4lab · Post by **g4lab** » Sun Dec 18, 2011 12:33 pm

Lab supplies are really handy to have around. You short their inputs and turn the current up until you reach the maximum you want to allow out of them.
Then put your device on and start turning the voltage up. They will stop delivering current where you set them to stop. They also can voltage regulate too.

If you search ebay "DC lab power supply" you get back 244 returns

this would be suitable for the present instance:

http://www.ebay.com/itm/Precision-Varia ... 43abe97988

but I would prefer this one which has enough current capability to drive(and even overdrive a bit) a 12 volt 100 watt quartz halogen lamp should you ever want to do so.

http://www.ebay.com/itm/Lab-Variable-DC ... 230e57da84

It could also drive any 6 volt microscope lamp you are likely to encounter.(usually about 35 watts the biggest. ie. 6 amps)