External power on Teensy 3

Not open for further replies.


Well-known member
Hi, all--

I'd like to use Teensy 3 with a battery pack and want to check my assumptions before doing anything. I have previously modified a Teensy 2.0 per Paul's instructions here using the example with 2 diodes and a battery pack, and I'd like to replicate that here. It appears that the 2 pads on the underside of Teensy 3, next to the USB socket, are where I would cut the trace and put the diode. Is that right? Anything else to do other than putting a second diode inline with the battery?

To clarify: this is going to be powered by 2xAA stepped up to 3.3V with this. The on-board rectifier serves as the diode on the battery lead. This is what I'm using for my Teensy 2.0 project and it works great!

Last edited:
I'm not sure, but wouldn't using 3 batteries, and option 3 be sufficient?

When using a battery, it may be desirable to avoid draining the battery when USB power is available The options above always run the Teensy processor from the external power. Using 2 diodes, you can automatically switch to the higher voltage.

From my little understanding, that device can 'step up' one AA battery to 3.3 volts, draining the battery faster then, say 3 AA's. Why do you need something more complicated? I use 3 AA batteries, and option #3
I may power from 3 AA batteries on some projects, and for me that decision doesn't affect my primary question here: really what I want to know is how Option #3 gets implemented on Teensy 3.0.

The reason I was using the step-up board is because I was making a hand-held device where 1xAA would provide over 24 hours of run time. It needed to be as small and as light as possible, and 3xAA was too much size and weight for that application.
My apologies.

From what I am reading:

First, a the "5V" pads must be cut apart, as in option #1 above. Then a 1N5817 diode is soldered between the pads. The cathod (side with the stripe) must face the center pad.

Looking at the cool picture I got w/ my order: They are on the bottom of the board, left side of the usb mini right next to the ground pin.
Great--thanks! Didn't mean to sound snippy in my previous response...it's been a long morning...
Looking at the size of the solder pads I think an SMT part will be better than the clipped through-hole diode I was able to use on T2. Looks like this part is the SMT equivalent of the 1N5817 diode. Will that work? Which way should the cathode point? I think it goes towards the USB plug and not towards the outer pins.
I didn't get mine up to now (postal delivery time), so I can not try myself. And in addition there is no schematics up to now which might be the question to be obsolete.

I want to feed the 3.0 with 3.3V already there in the project for other procs.
So VIN is not appropriate (3.7 to...).
I also assume that the battery input is not an glorious way to do.
To cut the Vusb from Vin is easy but I assume I have to cut also to output of the Voltage regulator. So of these do not like to be supplied backwards.
Anyone to have an idea on where to do best.

Besides all that congratulations to all of us regarding Obama's second term.
I'm on a similar but trickier path : I want to be able to connect a li-po charger circuit to the T3 to charge the li-po when the µ-USB is connected and run from the li-po when the µ-USB isn't connected. this should be lighter and smaller than 2AA, don't know yet about the battery life.
The li-po and the charger should be connected to Vin, I think I will need to do some clever tricks with diodes to make it work, but not yet really thought about it.
Re: li-po charging--I had asked the same question at Adafruit a few months ago about doing the same thing with Teensy 2.0. I don't know that new diodes were involved since the necessary ones were on their charger breakout. But I don't have the link available at the moment...it might be worth a search on their forum.
Good question--I assume the thought is that the stable 3.3v applied to the 3.3 line won't travel backwards through the regulator and into the USB. Is that how regulators work? I'm not an EE. If so, sounds like a perfect solution to me. I'm using a battery hooked up to that 3.3v step-up breakout applied to the 3.3v pin right now and it's working perfectly.
My goal is to have it be safe for both a battery and USB power to be hooked up simultaneously. When power is applied via USB, the regulator and inductor on my power supply breakout prevents any current from flowing back upstream to the attached battery (as read with a multimeter). If I power 3v3 with the battery-powered regulator and leave USB power disconnected, VUSB reads about 2v. So I guess that much is making its way backwards through the Teensy's 3v3 regulator when no voltage is applied to Vin. I haven't yet tried hooking up power to both VUSB/Vin (still tied together on my board) and 3v3 simultaneously. All I want to ensure is that I'm not pushing current/voltage back up to my computer when using both a battery and USB.

I know that I can bridge the Vin/VUSB pads on the bottom with a diode and that would prevent upstream current from flowing to the computer. Is that necessary for this case?
For a charging, I would use dedicated circuit. Like "Li-po rider" from seeedstudio.com. It should handle charging and switching. And supplies steady voltage on output, so you don't have to worry about anything. The one I mentioned gives 5V, so you can just connect it to Vin.
Hi, lami--to clarify, I'm not interested in using Li-Po myself on this project; instead, I actually will be using 1xAA due to size limitations (and it provides 20 hrs of runtime in my application, which is great).
All I want to ensure is that I'm not pushing current/voltage back up to my computer when using both a battery and USB.

I know that I can bridge the Vin/VUSB pads on the bottom with a diode and that would prevent upstream current from flowing to the computer. Is that necessary for this case?

I also want to know this.
For systems that require 5V power, I use a simple three-pin header with the center pin supplying power to the board. A shunt jumper is then used to either jumper the center pin to either source of power. If you want to add another power option, simply add another pin - up to four sources can be switched this way. But importantly, it is impossible to have two 5V sources on at the same time - they are electrically isolated from one another.

If you all you want is 3.3V power on board that is powered by multiple 5V sources, then an 'automatic' solution could be implemented, using Schottky diodes to isolate the sources from one another. That leaves plenty of 'headroom' for an external 3.3V linear regulator. It may even be enough for the on-board regulator (i.e. supply via Vin) though I'd check the specifications and the voltage drop across the pair of diodes that Paul has between Vin and the Voltage regulator input on the Teensy board. Just remember to cut the jumper between Vin and USB!

With a battery pack, I'd use a linear regulator to ensure that the charging voltages, disconnect spikes, etc. never reach the Teensy 3.0 CPU. Given that a normal voltage range for single LiPo battery packs seems to be 4.2-3.0V depending on the state of charge or whether the cell is being charged, I would consider operating the Teensy permanently at a lower voltage like 2.5V to leave plenty of headroom for the voltage regulator to work with, especially if you want to use a Schottky automatic 'switch over-between-sources' setup described above.

Or, if you have a voltage step-up solution built into your device, use that to apply power to Vin. Just make sure your power supply is regulated well enough to keep the on-board regulator happy. Alternatively, use an external 3.3V regulator. That's my approach, it means less power into the Teensy, less heat to dissipate, and likely a more stable 3.3V power supply than what the Teensy offers. Just remember not to backfeed via Vin or VUSB, i.e. cut the Vin-VUSB connection and attach nothing to the Vin pin.

For my application, I am going with the pin header approach as I have two 5V power sources and multiple devices on board that require 5V power. The nice thing about the pin header approach is that it's fool-proof, easy to implement, and there is virtually no voltage drop. You can always apply some glue to create a 'permanent' solution.
Last edited:
Not open for further replies.