I just finished a prototype of a 30A per channel, four channel current-sink
designed for high-current LED and electronic load applications. Each channel
has an independent current setting and on/off control.
This board plugs into a ZB1, NB1 or NB1A.
My prototype is setup for 1W leds (350mA). Higher currents are possible by
changing the current sense resistors. Pics are at -- Loading...
I have a sketch and library for controlling the current and on/off state for
each channel. I just need to add the PWM functions.
Is this practical? I thought most modern MOSFETs were designed for switching, and dissipated totally impractical amounts of power in their linear regions?
and dissipated totally impractical amounts of power in their linear regions
Well it's a calculatable amount of power but looking at the photos of the FETs just dangling they are not going be able to cope with much power. I think 30A is just the rating of the FET, you will not be able to run that anywhere close to that even with a heat sink with forced cooling.
The web site mentions setting the voltage just a volt or two above the compliance level, although I take that to mean a volt or two above the minimum level for the load as compliance tends to mean compliance with regulations.
Jluciani - have you calculated the total maximum power dissipation and hence the actual current you can control.
The 30A is the limitation of the FET running at a VGS of 3V and a junction temperature of 125degC.
The absolute maximum junction temperature is 150degC for this device.
The temperature rise of the junction is dependent on the power dissipated in the FET
and the heatsinking. Different applications will require different heatsinking.
The FETs hanging in free air are around 70DegC per Watt. With a current of
350mA and a few volts for compliance the power dissipation is around 1W.
The junction temperature is around 100degC.
I will be specifying thermal performance using a couple of extruded heatsinks
under static and forced air cooling. The thermal resistance of the FET is around 1degC/W.
The thermal grease adds apx 0.5degC/W. It should not be difficult to get a
junction to ambient thermal resistance in the 5-10degC per W range
with an inexpensive heatsink.
How much power is dissipated is function of the compliance voltage. For an LED
application the closer you can get to the sum of the forward voltage drops the
less power you dissipate in the current-sink. I am hoping to get within 5V.
If we say 5V maximum across the FET, 10degC/W and a 100deg rise gives a
maximum of current of 2A. The key is to minimize the compliance voltage.