SparkFun Forums 

Where electronics enthusiasts find answers.

Have questions about a SparkFun product or board? This is the place to be.
By angelsix
#158013
Ok I want to make a little board that I would use in so many of my projects, and it's something I always end up needing in various versions and voltages, so I want to make one that will hopefully suit all.

Requirements
Input: To turn on circuit I would like ideally to be as low as about 1.2V, but 1.8V is acceptable but anything higher and its no good.
Output: Again, ideally able to switch up to 12/24V but only need to switch up to 5V
Current: Output current 500mA required, 1A desired

So what I would normally use for this is a BJT such as a 2N3415 which turns on at 0.6 to 1.3V on the base and can switch 25V / 500mA. By the way this is just turning on/off LEDs most of the time, from low voltage low current trigger sources.

However the downside here is mainly the amount of current needed at the base is 3mA for 500mA output, needing a 600 ohm resistor on the gate, which is quite high, and a little volts drop over CE of 0.2V.

I know FETs use practically zero current at the gate as they are voltage controlled, however I would require an inverting circuit of some form to swap the n-channel to activating when the input is high not low, and also the Vgs difference will always be based on the input, so the worse case of say a 20V output from a 1.2V input would be difficult to achieve maybe?

I am presuming there is maybe an IC already out there for switching higher voltage loads from CMOS level inputs with little or no current? If not, what are your suggestions for a circuit to do this? I don't mind making a circuit board for this involving several ICs and passives if needed, I'm just after input firstly in case there already exists such a device that I am just not aware of at the moment?
By waltr
#158019
I would use an NPN BJT, grounded emitter, that switches a P-ch MOSFET, Source to input voltage, Drain is output V.
This way the BJT drives very little current to turn on the MOSFET. The MOSFET should have a low gate voltage (logic level) for full ON with a 5V input.

If looking for an IC solution look at the various 'hot swap' controllers and 'power' controllers.
By rrpilot
#158035
angelsix wrote:I know FETs use practically zero current at the gate as they are voltage controlled, however I would require an inverting circuit of some form to swap the n-channel to activating when the input is high not low, and also the Vgs difference will always be based on the input, so the worse case of say a 20V output from a 1.2V input would be difficult to achieve maybe?
A MOSFET is a transconductance device meaning a voltage controls a current. Not that I'm adding much to the conversation here but I just wanted to attach some schematics of what dlotton already suggested for clarification.

Since the MOSFET he specc'ed has a gate threshold voltage of 1V max it would be well suited for your application. From the spec sheet it also says it can conduct 2A at only 1.8V gate-source voltage. This is specc'ed at room temperature but lucky for you, the gate threshold voltage has a negative temperature coefficient so as the temperature goes up, you need less voltage for the same current.
You do not have the required permissions to view the files attached to this post.
By dlotton
#158047
Here's a simulation of a low-side switch and a high-side switch... essentially the same schematic that rrpilot posted.

The green trace is the control voltage (0-1.8V), the blue trace is the current in the low side switch U1, and the red trace is the current in the high-side switch U3.

Again, for the low-side switch you can use a single N-channel MOSFET. See my previous post for a MOSFET recommendation.

For the high-side switch you could use a part with a complimentary pair of transistors in one package, like the FDC6327C (one package containing one N-channel and one P-channel MOSFET). On this schematic U2 and U3 would be in one package. In the example I used FDMA1032CZ complimentary pair, but that leadless package is not friendly for hand soldering. I'd suggest the FDC6327C, available at DigiKey for ~$0.51 in one piece qty. Note that the FDC6327C is a 20V rated part, not 24V.

Data sheet for FDC6327C: http://www.fairchildsemi.com/ds/FD/FDC6327C.pdf

Image
You do not have the required permissions to view the files attached to this post.
By angelsix
#158052
Thanks guys for all the really useful posts. The SI2336DS looks perfect, very low turn on voltage in a FET and low-side instead of high side, with large current. I'll order that and give it a go. Price is high but sure I can get that down with larger volume.

rrpilot: The 100k resistor to ground, is that just to stop the gate floating high when the uC is powered down as its very low turn on voltage?
By dlotton
#158055
Yes, the pull-down keeps the switch off when the processor is in reset. Also, if/when the processor goes from non-reset to reset state, it will drain the gate capacitance and turn off the switch quicker.

Consider this if you're going to make a board that may get attached/detached to other things a lot; To protect the MOSFET gate you can put a series resistor (~100) and an ESD diode on the gate to lower the chance of blowing the gate with an ESD discharge. This isn't strictly needed, but if you're going to be handling the board a lot, connecting/disconnecting it'll help preserve the gate. Even just the series resistor will help.
You do not have the required permissions to view the files attached to this post.