I have a project that uses a battery powered ESP8266 to send data. It wakes up, sends a few seconds of data via Bluetooth, and then puts the ESP8266 to sleep for a few minutes. And the cycle repeats.
In my previous version I had one attached device which needed to wake up and sleep with the ESP8266. The current draw from the device was under 10ma so I just powered it via a GPIO pin and had the ESP8266 set the line HIGH when it was awake, which worked just fine. When the ESP8266 went to sleep. the line went LOW and no extra power was consumed by the device.
I now have two devices that also need to wake up and go to sleep with the ESP8266. The current draw is 40-60 ma so I can't power them from a GPIO line(s). I thought about using a transistor: Collector wired to 3.3V VCC, Emitter wired to VIN on the devices, and GPIO controlling the Base via a resistor. However, I'm concerned that the VCE voltage drop of 0.6V will lower VIN down to around 2.7V and not be enough.
So I'm looking to use a FET instead of a BJT. My understanding (limited) is that the Source to Drain resistance (thus voltage drop) can be very low with a FET. Can someone recommend a suitable FET to do this and suggested wiring?
It usually a bad idea to power any active circuit from a logic signal - such signals cannot provide enough current to charge the decoupling capacitors every logic device includes. I'd limit the capacitive load on a logic
signal to a few 100pF worst case (around a thousand times lower than typical decoupling capacitances).
It usually a bad idea to power any active circuit from a logic signal -
Agreed. That's why I'd like to use the logic signal to control a switching device (MOSFET?) which provides the power to the circuits which also run on 3.3V.
What I'm finding seem to have Vgs more than 3.3V. Looking for something where I can attach the Drain to 3.3V, the Source to the load and control it with 0V (off) and 3.3V (on).
Everything I seem to find is for (relatively) high power/voltage requirements.
Are you planning to use P-channel High side switches ?
I believe so. And to make it work, I will probably need a P-channel device. Source connect to 3.3V and Drain connected to 3.3V devices I want to turn on/off. In order for the load to be shut down when the ESP8266 goes to sleep, I believe I'll need to put an N-channel device to control it (maybe, not sure if ESP8266 GPIO lines float when it's sleeping). Something like this:
I'm looking for suggestions for the device(s) to use. I searched Newark, and the logic-level devices with low S-D resistance all seem to be overkill - high amperage and voltage devices, like FQPF27P06. I was looking for something in a TO-92 case or similar, but what I found had a S-D resistance of around 7 ohms, which seems high. I'm only switching less than 100ma at 3.3V.
NOTE: If you can't tell already, I don't know what I'm talking about. I understand BJT transistors, but not FETs.
Also, I tried to insert the diagram from a Dropbox shared folder, but for some reason it isn't showing up. Thus the URL link too.
That diagram (post#5) halves the 3.3volt drive voltage for the fet.
1.65volt is not enough for most logic level mosfets to fully turn on.
Change the pin<>gate resistor to 1k, and the resistor to ground to 100k.
Or use a small-signal NPN transistor there, with just a 10k base resistor.
The p-channel fet with 10k gate pull up is ok, as long as it's one with very low Vgs(th).
Leo..
I used a P-channel MOSFET: TP0604N3-G available from several sources including Mouser. A low signal on the GPIO pin turns power on to the controlled devices. A high signal turns them off. Not designed for a lot of power - just a couple of sensors, etc. - maybe 50 to 100 mA.
