Newby question: Need a 2mA Constant Current Sink

[Windcrest77]

A newby here hoping to see if I am on the right track with this schematic and advice. Firstly everything to the left of the JFET came from the LT3080 datasheet application notes. The LT3080 is unique in that it can regulate down to zero volts instead of the usual 1.25, thats why I want to use it. I need a simple variable regulator board that can output from 0 to 35 volts, the 0 to 3 volt range will be used most often in fact so an LM317 wouldn't work here.

This circuit will be for biasing tube grids experimentally, so there will be no current draw when its used, bias on a tube grid just creates a negative field no current draw. This presents a problem, the LT3080 wont work unless it has a minimum load of 1mA. So I thought I would add a constant current sinking shunt at the output, as simple as I can.

So, I want to ensure that 2mA of current is being drawn at any output voltage setting. I want the simplest possible current sink to do this, I think a JFET with a resistor is simplest correct?

If I am on the right track here what would be a good JFET device number to use? I suspect the resistor will be somewhere in the 100 to 200 ohm range depending on the JFET used correct?

Thanks for any help on how I can sink the output of this supply to a constant 2mA so the LT3080 regulates, even in use it wont be drawing any more than that unless I have a gassy tube.

(I corrected the original schematic in post #6)

 

[Mooly]

The resistor should be in the source rather than the gate. The 2SJ111, 112 and 113 FET's are rated to 35 volts. Something like a 680 ohm. I'm not sure how low a voltage that arrangement would work down to.

 

[Mooly]

Here you go. Supply voltage on the left, current on right. Seems to work down to low voltages with a J113 FET.

 

[grunf]

The depletion BSS139 will have an even steeper curve, Rset around 400ohm.

 

[MarcelvdG]

It will never draw 2 mA at 0 V. If you really need that, I guess you will need something with a negative supply.

Why do you want to bias the grids positively? Or is the cathode at a positive voltage with respect to ground?

 

[Windcrest77]

Thanks! Duh, I see I drew it wrong, so posting it corrected. It looks like the minimum current won't be met when going below 15V, this will stop the regulator according to the datasheet. In the datasheet it says .5 mA minimum then in another place it says 1mA, so I'm going with the higher. I may have to increase the sink target to 10 mA or higher so that voltages from .1 to 2 V can see a minimum current of 1 mA as well. Or figure out how to switch a shunt resistor across the output when working with voltages below 3V. Maybe a high and low switch. I don't need accuracy across all the voltages, just the minimum 1 mA.

 

[Windcrest77]

It will never draw 2 mA at 0 V. If you really need that, I guess you will need something with a negative supply.

Why do you want to bias the grids positively? Or is the cathode at a positive voltage with respect to ground?

Yes I was just going to reference the positive of this supply to the tube circuit ground. Maybe I do need to go with a negative supply that can cover voltages in the -1.25 to 0 gap, there is a negative version of the LT3080 I believe. Back to the drawing board! Any drop out of bias wouldn't be good to the DUT tube, so I wouldn't want to risk using a regulator that quits, and this circuit does have that danger. Another alternative is a shunt regulator circuit too I guess.

 

[grunf]

I would put a potentiometer between the +35V regulator and the -12V regulator, so that you have a range from +35V to -12V on the middle terminal towards gnd. The simplest solution .

 

[Windcrest77]

I would put a potentiometer between the +35V regulator and the -12V regulator, so that you have a range from +35V to -12V on the middle terminal towards gnd. The simplest

I'd like to understand this approach better. So if I made a split rail supply with pos/neg regulators set to let's say +2 to -35 you're saying I could have a pot that smoothly transitions the output from +2 to -35? With no gaps close to the zero volt area? That would be perfect adjustability as it would traverse from positive to negative very smoothly right across those voltages closest to zero, especially with a 10 turn pot. Can you give me some hint as to where the potentiometer hooks up?

 

[chrisng]

Just an idea, not sure if it's actually feasible.