Using LM334 for higher current

[carbonzit]

I'm stumped.

I'm trying to use a LM334 in a higher-current configuration, as per the datasheet:

**broken link removed**

I tried my hand at calculating Rset, but I'm getting the wrong result. Let me show my work (I"m the student here):

My goal was to get 30mA out of the regulator.

From the datasheet:

The current flowing through RSET is determined by VR, which is approximately 214μV/°K

So taking the temperature at 68°F, I get 293°K, so Vr should be ~ 62mV.

Since n is typically 18 for 2μA £ ISET £ 1mA, the equation can be further simplified to
Iset = Vr/Rset × 1.059

(n is the ratio of the two currents in the circuit, Iset and Ibias)

Rearranging the equation to

Rset = Vr × 1.059/Iset

I get Rset = 2.1Ω

So I built the circuit with 2Ω for Rset. For the "stability" cap and resistor, I used 0.01µF and 15KΩ. I used a load resistor of 330Ω. I used a BC328 transistor.

Applying 10 volts to this, I measured a current of only a few mA. So I changed Rset to only 1Ω; the current rose to 7mA. The current didn't change when I decreased the load resistor, which tells me that at least the current regulator is doing its job.

The datasheet gives no explanation of how to compute Rset for the "higher current" circuit, so I ASS-U-med that the calculation would be the same as without the transistor. So what am I doing wrong here? Do I need a fractional Ω Rset to make this work properly?

Also, how do you select that R and C "for optimum stability"? I'm not smart enough to figure this out yet ...

 

[crutschow]

The current calculation should be the same.

Perhaps it is oscillating and that is causing the problem. Do you have an oscilloscope? I would try a smaller stability resistor value.

You determine "optimum stability" by looking at the response with an oscilloscope. Optionally, if you can find a Spice model for it, you could simulate the circuit to get an idea of the stability.

 

[Sceadwian]

Where is your load connected to?

 

[carbonzit]

Where is your load connected to?

On the bottom (-Vin) side, +Vin is connected to the positive power supply output. I don't really think it matters, does it? whether load is on "top" or "bottom" here. (As I said, load is just a resistor for now.)

My goal was to build a current-limited source for testing LEDs.

So, am I correct about calculating Vset & Rset here? If so, why doesn't this work the way it should?