Amplify current to UV-LED

[runarh]

I have a function generator (Escort EFG 2210) as a source to a UV-LED (LED255J ). The FG delivers about 7.5 Vrms and 30 mA pulse signal. The LED requires closer to 100 mA for optimal emission. The voltage is good enough. I have tried various amplifier circuits, but end up with less current than before.

Any ideas?

 

[alec_t]

Welcome to ETO!
What frequency are you using?
What is the rated Vf of the LED?
Are you trying to modulate the current in proportion to the generator voltage?

 

[runarh]

The frequency should be from 10Hz to 1 kHz. Vf (forward voltage?) is 8V typical at 100 mA.
Not sure what you mean by the last Q, but I want to end up at about 8V and as close to 100mA as possible.

 

[Colin]

Put a transistor in the circuit.

 

[runarh]

Put a transistor in the circuit.

Already tried a couple of different amplifiers using transistors, opamps and both.

 

[runarh]

It's worth mentioning my knowledge level isn't that high.

 

[audioguru]

On the other website I posted this schematic:

 

[spec]

I have a function generator (Escort EFG 2210) as a source to a UV-LED (LED255J ). The FG delivers about 7.5 Vrms and 30 mA pulse signal. The LED requires closer to 100 mA for optimal emission. The voltage is good enough. I have tried various amplifier circuits, but end up with less current than before.

Any ideas?

Hi run,

Can you provide the following information:

(1) What exactly is the waveform from the signal generator. 7.5V RMS does not mean much.
(2) You will need an additional power supply of say 12V to power an amplifiers to drive the LED with 100ma at 8V. Do you have such a supply available?

 

[runarh]

1. Square, peak is about 10,7V.
2. I have a myDAQ with a 15V output among others, but max current output from it is 2mA.

If necessary I'll have to get another.

 

[spec]

1. Square, peak is about 10,7V.
2. I have a myDAQ with a 15V output among others, but max current output from it is 2mA.

If necessary I'll have to get another.

Thanks, you will need an additional 12V supply, but 15V would do, both with a capacity of say 200mA DC. I will assume 12V.

 

[spec]

OBSOLETE PLEASE SEE POST # 22
2016_02_10

Hi run,

Here is a circuit for you to try.
(1) It needs a 12V supply with a current capability of 110mA
(2) It puts a constant 100mA, into the LED when the pulse input is low. If you want this reversed so that the LED is on when the input pulse is high let me know and I will modify the circuit
(3) Q7 can dissipate around 350 mW so it will need a small heat sink.

ERRATA
(1) place a 100mA fuse in series with the 12V SUPPLY ( fuses normally blow at around 120% or more above their rating).
(2) Place a 9.1 to 10V, 1W Zener diode across the LED (to protect the LED, in the event of a circuit malfunction.
(3) Replace Q1 with a 2N2905A with a 1.1W heat sink

 

[Tony Stewart]

If rated for 7.5V @ 100mADC max, you can possibly drive it 150mA at 50% duty cycle at 8V. The internal ESR will be around 10 to 25 Ω but the forward voltage will drop with rising temperature, so a heat sink is essential and solder time should be limited to 3 seconds ( 5 sec. abs max.)


Any supply voltage above 8V will result in waste heat somewhere but you could use a classic LM317 adj reg from 12V. You can also use any ADJ 3 terminal regulator as a current limiter with the current sense set to 1.25V drop and a minimum 2V drop on the LM317.

For faster turn off of the diode a resistor across the diode of a few hundred Ω is advised as the regulators are emitter followers only with fast pullup only.

The logic level signal input may be any logic level N-Ch MOSFET switched current to ground with RdsOn <=1Ω with drain going to negative or cathode side of LED and source to ground.

Vin DC of LM317 10~15V with suitable heatsink withing constant current when the MOSFET is ON.

This is one of hundreds of possible solutions.

 

[spec]

OBSOLETE PLEASE SEE POST #22
2016_02_10

Hi run,
Here is a version where the LED turns on when the input signal from the pulse generator is high:

​

ERRATA
(1) place a 100mA fuse in series with the 12V SUPPLY ( fuses normally blow at around 120% or more above their rating).
(2) Place a 9.1 to 10V, 1W Zener diode across the LED (to protect the LED, in the event of a circuit malfunction.
(3) Replace Q1 with a 2N2905A with a 1.1W minimum heat sink

 

[runarh]

Thanks for great help. Got another supply with adjustable voltage and current, up to way more than we need. The LED is supposed to light up when the pulse is high, and turn off when low, so the last circuit will probably be good. Does any of the components in that one need a heat sink? The UV-LED is about $500, so I don't want to end up breaking it

 

[spec]

Thanks for great help. Got another supply with adjustable voltage and current, up to way more than we need. The LED is supposed to light up when the pulse is high, and turn off when low, so the last circuit will probably be good. Does any of the components in that one need a heat sink? The UV-LED is about $500, so I don't want to end up breaking it

Hi run,

Wow, what a cost that LED is.

Q1 would need a small heat sink.

Circuits 1 & 2 should be fairly safe, but I would advise putting a 100mA, or there abouts, fuse in the supply rail before the decoupling capacitors. Alternatively, if you have a current limit on your PSU, set it to around 120mA.

You could put a 9.1V to 10V, 1W Zener diode across the LED

Q1 as a BC337, would be fine, but for extra safety you could make it a TO-5 2N2905A metal can NPN BJT.

Incidentally both circuits use current steering so the drain on the supply is fairly constant around 110mA

The LED would be dissipating 800mW @ 8V & 100mA so, if possible, that would benefit from some kind of heat sinking. At the very least make sure there is good ventilation around the LED.

spec

 

[runarh]

Max power dissipation from the LED is 1W, so it should be good without heat sink? It has three pins, one for case.

 

[spec]

Max power dissipation from the LED is 1W, so it should be good without heat sink? It has three pins, one for case.

Hi run,

Yes, that is correct, but there is more to it than that. What counts is the maximum allowable junction temperature of the LED

The cooler, within reason, that you can keep a component the more reliable it will be. In view of the high cost of the LED I would advise taking all reasonable precautions you can to give it a favorable environment. For example, it would help the cooling if the case terminal were soldered to a nice large thick copper area on the PCB. Same for the other two leads.

spec

 

[spec]

Hi run,

I have just had another look at both of my circuits and have found that Q1 will be dissipating around 1.1W when the LED is turned off. For the time being I have suggested changing Q1 to a 2N2905A with a 1.1W heat sink. This will sort the problem, but I will investigate a more elegant solution.

spec

 

[audioguru]

Be careful with that expensive LED. Its case is a T0-39 metal can which was used for transistors (2N2219 for example) many years ago so a heatsink might not be available today.
Its lifetime minimum is only 1000 hours, 25 degrees C and 100mA but they do not say if the temperature is its case or its ambient. It survives longer at 20mA.

They warn that it damages your eyes and may cause skin cancer. I think it might also cause a sunburn on skin.

 

[runarh]

The UV-light should go through an optical fiber cable into a container with fluorescent liquid. At lower currents the UV is not bright enough to cause the liquid to light up. This light should go back through another fiber cable to a photo diode.