How do I switch to Constant Current instead of Constant Voltage on RD6018?

[sofakng]

I'm trying to test an LED with a max current of 100 mA.

The RD6018 shows [CV] at the bottom of the screen so I'm assuming it's in Constant Voltage mode so how do I switch to Constant Current?

I've already destroyed one LED because of my misunderstanding of the bench power supply.  I had V-SET to 5.00v and I-SET to 0.9A (and no OCP).  I'm confused though because I have current set to 0.9A but I'm guessing that only works in Constant Current mode?

[Fungus]

I doubt it has constant current mode.

For an LED you can set the voltage much too high then use the current limiter. Set it to 100mA.

[wasedadoc]

0.9A is 900mA.  That is 9 times the 100mA max of the LED.

[Martin72]

Hi,

A supply is in it´s constant current mode when you have reached the pre-adjusted current.
Until this it´s in constant voltage mode.
For your LED you should set the current to 100mA before.
BTW, what LED is it?

[sofakng]

Ahhhh... thanks.  I'm an idiot.  0.9A is 900mA as mentioned by wasedadoc.

The LEDs are the SFH 4547.

What is the expected behavior of the following settings?

V-SET: 5.0v
I-SET: 0.1A (100mA)

The LED has a forward voltage of 1.5V so would the PSU reduce the voltage from 5.0V to 1.5V and go into Constant Current mode to limit the current to 100 mA?

Also -- When I measure the LED with my multimeter (Fluke 87V) using the diode mode, it shows 1.1V.  The specifications show 1.5V so I was expecting to see that?

[Martin72]

I-SET: 0.01A (100mA)

0.01A is 10mA....

[wasedadoc]

Ahhhh... thanks.  I'm an idiot. ...

V-SET: 5.0v
I-SET: 0.01A (100mA)

Whatever school you went to did a poor job.  0.01A is 10mA not 100mA.

[Martin72]

The LED has a forward voltage of 1.5V so would the PSU reduce the voltage from 5.0V to 1.5V and go into Constant Current mode to limit the current to 100 mA?

I wouldn´t make such experiments, better set it to 2V or set the power limit to 150mW (1.5V*100mA)
Be aware that the 100mA are the maximum of the diode, you musn´t reach this value.

[TimFox]

The voltage and current settings on a bench supply control the output voltage.
At most one of them can determine the output voltage at a time, whichever one limits the output voltage to a lower value.
Often, the voltage control is more precise than the current control.
One way to check the current limit is to connect the output to a low-resistance ammeter or a voltmeter across a low resistance load, where the voltage at that current is reasonably small.
As you increase the voltage control from zero, the current control function will kick in very quickly and you can adjust the current control to obtain the desired current.
If you increase the voltage control further, the current should not increase.
If you connect a higher-voltage load, such as an LED that needs, say, 2 V, the voltage control must be higher than that value, and the current control can then change the current over its range.

[J-R]

I would not set the voltage to 5V.  The power supply is going to have some amount of capacitance on the output, so you could still damage the LED when you first connect it.

[John B]

That and you have to trust (or measure) that the CC control loop limits the current fast enough to an LED alone, ie you can also add some series resistance for insurance. As above, try reducing the voltage too. No need to risk the output shooting up towards 5V if the Vf is only 1.5V.

[Swainster]

In my experience (which doesn't include RD6018), bench power supplies can be slow to switch from CV to CC. I've blown a few LEDs by just setting the current and wacking the output enable button. On some supplies, you may even have to switch on the output before adjusting the settings, tempting you to connecting the LED while the supply is already at the compliance voltage - been there, done that. It's safer to turn on the output at a low starting voltage then manually ramp it up till it switches into current limited mode. 

BTW, the measured forward voltage depends on the current (and temperature, production batch etc), so your multimeter diode check is unlikely to hit the same value as the datasheet.

[TimFox]

Also, some DMM "diode" modes will not go high enough in voltage for the shorter-wavelength (blue) LEDs that may need > 3 V.

[sofakng]

Thank you so much for the help everybody.  I obviously have a lot to learn (and stop making basic math errors).

I have a follow-up question on the LED but I'm going to post that over in the beginner forum...

Thanks again!

[Fungus]

What is the expected behavior of the following settings?

V-SET: 5.0v
I-SET: 0.1A (100mA)

The LED has a forward voltage of 1.5V so would the PSU reduce the voltage from 5.0V to 1.5V and go into Constant Current mode to limit the current to 100 mA?

Yes.

Except it's not called "constant current mode", it's "current limiting" mode.

A PSU with "constant current mode" doesn't require you to set the voltage to anything special.

[TimFox]

What is the expected behavior of the following settings?

V-SET: 5.0v
I-SET: 0.1A (100mA)

The LED has a forward voltage of 1.5V so would the PSU reduce the voltage from 5.0V to 1.5V and go into Constant Current mode to limit the current to 100 mA?

Yes.

Except it's not called "constant current mode", it's "current limiting" mode.

A PSU with "constant current mode" doesn't require you to set the voltage to anything special.

A true "constant-current" source has a finite compliance voltage, and on a lab supply this can be set to limit the voltage into an open circuit.
For measuring a diode forward voltage, that compliance voltage must be set higher than that diode voltage will be.
For a bench supply with true constant-current mode, the voltage must still be set to a sufficiently high voltage.

[WattsThat]

Do not rely on the power supply to control the current when testing LED’s, use a series resistor.

You have a supply that can deliver 10 amps or more, so excessive current (from too high a voltage applied) will destroy the diode before the supply can ever regulate down to 10 ma which is the maximum allowable continuous current of the diode. Sure, it can handle an amp but that’s only for 100 microseconds. I doubt the supply can even current limit that low, given it’s current output ability. The desired 10ma is incredibly low relative to the output capacity, 0.01/18 is only 0.05% of full scale output, a very big ask of the supply. This may be okay in an older pure analog supply but in a modern device where things are controlled by DAC’s, that low a value may well be impossible.

Since the datasheet says VFD =1.5-1.7 V, set your supply to 5 voltage and calculate the required resistor with R =E/I so 5-1.7 / .01 = 330 ohms.

I don’t know why the diode read 1.1 volts on your Fluke in diode mode, perhaps it’s been damaged by overcurrent? I’d expect it to be shorted if really fried but I don’t know if the Vf could somehow go down with over current, I doubt it. Someone here will know.

[Martin72]

Finally the series resistor is the safest, no question.

[Fungus]

A true "constant-current" source has a finite compliance voltage

Obviously you won't be able to connect up a 1MOhm resistor and ask for an amp. :-)

[TimFox]

A true "constant-current" source has a finite compliance voltage

Obviously you won't be able to connect up a 1MOhm resistor and ask for an amp. :-)

I think that's what I said.  My good bench constant current supplies (old Keithleys) have a front-panel control for compliance.