delecting the correct npn

[justwantin]

I have breadboarded a circuit using ap piv and an 2n3904 to switch on 1W LEDs using a power supply that had around a 200mA output a 5v> I'm now ready to run this circuit using the correct current or 350mA of a new power supply but I am not sure about reading the datasheet correctly.

The 2n3904 datasheet says that max IC is 200 mA. Am I correct to think that this npn is undersizd for the task at 350mA?

 

[bychon]

yes.......

 

[Hero999]

Use a higher current transistor such as the BC338 instead.

http://www.datasheetcatalog.org/datasheet/GeneralSemiconductor/mXtwuqt.pdf

 

[justwantin]

Thanks to both you guys,

I stopped at a Jaycar between work and home and picked up 5 MJE340's after a quick scan of the store catalog. They do the job albeit I wont have a power supply with enough amperage to supply 350 mA each to to 6 LEDs until (hopefully) tomorrow.

I've set up the 5 MJE340's with bases to a 12f675 without putting in a base resistor in series and the circuit works as expected. Is there any reason that I should use base resistors if it works OK without them?

 

[Hero999]

A bit overkill on the voltage side (300V!), probably not what I would've chosen but will do the job fine.

See what I mean? The BC338 is much cheaper.

**broken link removed** - $0.42
**broken link removed** - $2.75

It's up to you of course but I'd recommend taking the MJE340s back or saving them for a different project and buying some BC338s. If you do this then I'd recommend getting a pack of 10 or 25 BC338 as they're very good general purpose transistors.

 

[audioguru]

Without seeing your schematic then we don't know if the transistors are emitter-followers that do not need a series base resistor or if they are common-emitter types that need a series base resistor.

EDIT:
Are you making the High Power Mood Light project at Electronics Lab?
The author originally used common-emitter type of LED drivers without series base resistors and people wrote in that the microcontroller got so hot that it stopped or was damaged.

Then the author added series base resistors but selected a value that was 15 times too high which made the LEDs dim when the transistors had less gain than typical.

The current-limiting resistor for the red LED had a value that was too low that caused the LED to burn out.

 

[justwantin]

See what I mean? The BC338 is much cheaper

Yes, makes sense, I'm new to this and grabbed the first npn that looked like it would do the job.

Are you making the High Power Mood Light project at Electronics Lab?

I've attached a hasty schematic. This is my own design, 1 LED on constant and 5 LED's flickering randomly to simulate flame in a paper lantern. I've done this with 15 X 5m LEDs and it worked out well as a very bright night lite but the LEDs ran directly off the 12f675's pins. This time I'm looking for more light for out on the verandah and I'm using Edixeon 1w stars. It's run all night in the shop, no magic smoke and nothing warm to touch, however its only 12C in the shop.

 

[tvtech]

Without seeing your schematic then we don't know if the transistors are emitter-followers that do not need a series base resistor or if they are common-emitter types that need a series base resistor.

EDIT:
Are you making the High Power Mood Light project at Electronics Lab?
The author originally used common-emitter type of LED drivers without series base resistors and people wrote in that the microcontroller got so hot that it stopped or was damaged.

Then the author added series base resistors but selected a value that was 15 times too high which made the LEDs dim when the transistors had less gain than typical.

The current-limiting resistor for the red LED had a value that was too low that caused the LED to burn out.

You guys are sharp.....LOL. Interesting forum.

 

[audioguru]

The base-emitter of your transistor is short circuiting the output of you microcontroller. The base must have a current-limiting resistor because the max allowed output current from a microcontroller output is only 25mA. Without limiting the current then the microcontroller will overheat, smoke then curl up and die. But maybe the flickering reduces the amount of heat.

The 1W LED uses a max current of 285mA. If you limit its current to 220mA then the base current of the transistor can be limited to 22mA.

 

[justwantin]

The 1W LED uses a max current of 285mA

Why would you say that when my specs say 350mA?

My bench top PS doesn't seem to be capable of supplying more than 240mA and I have never seen any one LED (max to its rated current when flashing, however when flashing from all 5 pins 240mA is never achieved according to my multimeter and I am never sure if all 5 might be on at one instance. The readings go up and down within a given range. The same is true for 5mm configurations.

I'll have to read some more on this and figure out what resistor to use. I seem to be having a problem understanding how this works and hence how/what to use for a base resistor

 

[audioguru]

Why would you say that when my specs say 350mA?

You did not post the spec's. A high power LED has a voltage of about 3.5V. 1W/3.5V= 286mA, not 350mA.
If a 3.5V LED draws 350mA then its power is 3.5V x 0.35A= 1.22W, not 1W.
Maybe your LEDs are 1W/350mA= 2.86V?

I'll have to read some more on this and figure out what resistor to use. I seem to be having a problem understanding how this works and hence how/what to use for a base resistor

The spec's for nearly all transistors show a max saturation voltage loss when nthe base current is 1/10th the collector current.

 

[justwantin]

A high power LED has a voltage of about 3.5V. 1W/3.5V= 286mA

Thanks for pointing this out. I can pick up a switchmode plug pack locally that is rated at 5V 2.6A which will provide enough current to run 6 of these so called 1 watt LEDs and 5V is acceptable for my PIC. The Altronics catalog only listed the max If (350mA) and max Vf (5V) when I purchased these LEDs so thats all I had to work with. The LEDs are Edixeons and the bottom of the star is stamped EDSX-1LA5-C1. There's no joy finding that particular LED on any of the Edison spec sheets but the closest 1w Edixeon LEDs all have similar specs such as this:Edison Edixeon ARC 1W Star by LED-TECH.de. This one has a Vmax of 4, not 5

Being that the max Vf was listed as 5V I didn't intend to regulate the voltage to 3.5.

Maybe I'll add an lm317 to the breadboard.

 

[mneary]

The LED maximum voltage is not the maximum voltage you're allowed to apply. It's the maximum voltage drop that you may see at the rated current.

You must take steps (usually current regulator or resistor) to be sure that the maximum current is not exceeded. Feeding a regulated voltage directly to an LED does not limit the current and so is likely to result in under- or over-current in the LED.

 

[justwantin]

Besides regulating the voltage to, say 3.5V, I'd size a resistor using R = (VS - VL) / I.

 

[audioguru]

Besides regulating the voltage to, say 3.5V, I'd size a resistor using R = (VS - VL) / I.

An LED sets its own voltage which might be low or high for two identical LEDs. You must limit the current.

The link above and the link to the datasheet in Google do not work today so I can't see what is the recommended current for the "1W" LED.
The distrubutor's sales sheet lists the 3 typical voltages for the 3 colours of the LEDs but ndoes not show the min and max range of voltages. They say, "LEDs should be run with a constant current in addition to the manufacturers recommendations. To ensure this we recommend to take constant current power supplies because resistors are always a risk factor in high power LED applications."

 

[justwantin]

I have as close as I can find to a spec sheet here:**broken link removed**

It does not list an EDSX-1LA5, but gives specs for EDSX-1xx1 LEDs at Absolute Max Vf 5, Min 2.8 and Max at 4.0. With 5V, I calculated with 3.5V with 300mA and used a 5.6 ohm resistor. With the 5V 2.6A swithmode power plug and 5k resistor on base of a bc338 I don't get magic smoke and my light functions well