Teensy current consumption

Status
Not open for further replies.
I have seen as much as 45 ma running at 90 MHz and as low as 0.3 ma running at 2 MHz using deepSleep from the Snooze library. I have been logging 3 voltages every time they change 25 mV for 11 days now running from a 500 mAh lipo and the Teensy feather adapter and those methods. Cutting the USB power trace allows you connect 3.6 to 6 volts to the Vreg input. I'm not sure if putting 3.3V directly on the 3.3v line is supported.
(Teensy 3.2)
 
Last edited:
I would like to decrease the clock speed of Teensy 3.5 to reduce current. Is this possible without breaking all sorts of timing routines including USB communications? If so, how do I do it?
 
I would like to decrease the clock speed of Teensy 3.5 to reduce current. Is this possible without breaking all sorts of timing routines including USB communications? If so, how do I do it?

Minimum clock speed to use USB is 24 and the speed Menu items Frank B mentioned that don't work are marked (no USB).So far everything I've tried takes into account the CPU speed so delay(1000) is a second at 24 MHz or 90.
 
Ehm.. if you use USB: Why do you want to save power? What's the host?

Of course, RTM applies. I am just in the design stage so had not tried selecting T3.5 from Arduino. Thanks for the help.

I will not be using USB power and will have several T3.5 in the system. I might also use 3.6VIN instead of 5 to cut down the input current on my DC-DC converter. Or maybe I can feed 3.3V directly to the 3.3V pin.

Tom
 
I still do not understand..
I you have USB, it powers your Teensy. Sure, you can disable this, but why, if a USB connection is available (and needed, according to your Question "..Is this possible without breaking all sorts of timing routines including USB communications" ) ? The USB-Host needs power too. Is it a notebook ?

Sorry, but I think, there are important infomation missing :)
 
I still do not understand..
I you have USB, it powers your Teensy. Sure, you can disable this, but why, if a USB connection is available (and needed, according to your Question "..Is this possible without breaking all sorts of timing routines including USB communications" ) ? The USB-Host needs power too. Is it a notebook ?

Sorry, but I think, there are important infomation missing :)

I apologize for the confusion. I will use USB to program T3.5 and initial debug. Once the device is built Teensy will communicate through UART or I2C and will not be connected through USB. I did not know that you could reduce clock speed right in Arduino and I was concerned that reducing it might break USB communications.

I plan to cut the USB power trace so that my board will not back feed the USB host. I will have a blocking diode on my supply so the output voltage will be about 4.3V. Even without cutting the trace, I don't think any current will flow back into the USB host.

Tom
 
Rough current ranges .... with the T3.6 I measured about 80-mA at 180-MHz and 100 or 120 mA [forget which] at 240-MHz, with nothing connected to the I/O pins. I had cut the trace underside of the pcb, and was powering from an external 5V regulator. And as I recall from a year ago, the T3.1 was drawing about 40-mA at 90-MHz. So, T3.5 will be in between these values.
 
Looking at the T3.5 schematic, there is a pair of blocking diodes on the USB power input which are not on 3.2 or LC. If the 3.5 is powered from an external 3.3V supply connected to the 3.3V pins it looks like you would not need to cut the USB power trace. You would need to protect your own supply with a Schottky diode and make sure you have enough voltage at the 3.3V input pins.
 
I apologize for the confusion. I will use USB to program T3.5 and initial debug. Once the device is built Teensy will communicate through UART or I2C and will not be connected through USB. I did not know that you could reduce clock speed right in Arduino and I was concerned that reducing it might break USB communications.

OK :)
USB can be used down to 24MHz, not below.
UART and I2C can work with lower Speeds. For even more power-saving, you can try the great library by DUFF or add a asm volatile ("WFI") which stops the CPU untill the next interrupt occurs (but i'd really try the library).
 
Status
Not open for further replies.
Back
Top