Powering Panda 2 with 5V but still use USB

I have an I2C board which runs (and supplies) 5V.

I want to use a Panda 2 board to control the I2C. Looking at the schematics of the Panda board it should be possible to run the board by supplying 5V from the I2C board to the 5V pin of the Panda 2 header.

  • Can the board be run on (external) 5V supply, by powering that 5V to the 5V pin of the header (and connecting the GND to the GND :slight_smile: )

  • Can the USB still be connected to the Panda 2 for debugging (since this also supplies 5V)?

  • Looking at the schematics / I2C port of the Panda, it looks like it runs on 3V3 (VCC), is it a problem if the I2C runs at 5V?

USBizi have 5v tolerant inputs, but will only bring VCC levels to 3v3 not 5v, so the issue is more a quiestion of whether your device will work if a “1” is only that level (it should :))

If you put 5v as an input to the Panda on the 5v rail, then that is fine. You need to take care of the power regulation. Debug over USB still works.

Well, the I2C board already brings the I2C to 5V…

I have two resistors (10K) which connect the I2C bus to 5V and the Panda2 resistors bring it to VCC (3V3) by using 2.2K resistors

I could remove those 2 resistors from the panda board… then all should run on 5V, right? Or will it work without modification?

Just to be safe, I removed the R15 & R16 from the pandaboard… I’ll fire things up in a few…

yes, you only need one set of pullups, so you should have done the right thing. I personally don’t think it would have caused you a problem to run with both pullups on.

It works… removed the resistors as mentioned before, and I was able to control the I/O extenders (PCF8574) using the examples which were here on the side.

Is it possible to use lower I2C speeds then 100KHz? Or do I need to use the Software I2C then?

@ Brett You are correct, if I have read the thread correctly 2.2k in parallel with 10k then,

1/(1/2.2k + 1/10k) = 1.8k approx and as the pins are seeing 5v then

5/1.8k = 2.8 mA

On an LPC2387 from the data I have read, the pins will current limit at 50 mA but I am not sure if that applies to current sink (which might be 4mA).

However there maybe a gotcha, the 3.3v for the uC needs to be up before the 5v, otherwise the max is 4.6v

@ iwritecode Why would you need a clock slower than 100k, anyway you can clock I2C at any speed, but you need to use software for less than 100k as far as I can see, but I could be wrong.