Pocketbeagle 5v and 3.3v output issue


pocketbeagle (PB) with serial port 4 (TX and RX) connected to uart on a Curie nano board. PB I2C1 connected to a MPU6050 breakout board. PD 3.5 volt output (P2 PIn 13) supplies power to MPU6050 board.

This configuration works as I can transfer data between the PB and the Curie Nano via the serial port.

I had power off on the Pocketbeagle and was updating a script on the Curie Nano board. The Curie Nano was powered via the USB connection to my laptop.

I noticed that the power LED on the MPU6050 board was on although it was not very bright. I measured the 3.3 volt input to the MPU6050, it showed 1.7 Volts. I also looked at the 5 volt output from the PB, it was 1.8 volts. As the only device that actually had power was the Curie and the only connection between the two boards was the serial port, I unplugged the serial port wires from the header. The voltages on the PB board power output pins went to zero. Putting the serial wires back brought the voltages back. I measured the voltages on the TX and RX wires. The TX was 3.0 volts and the RX was 0.08 volts. I beleve this is as expected as the TX has a Pull up and the RX has a Pull down.

I checked the boards for shorts and did not find any. I in fact did not expect to find any as the boards run fine when both are powered on. When the PB is powered on, the output voltages are at their correct level.


1: Has anyone else seen this sort of behavour on te power output pins?

2: Is it really an issue or is it just how the OSD335x is designed.

3: Should I be concern about this killing my board?

This is pretty common behavior. TTL 232 pins have a limited drive capability and may be sufficient to power or partially power connected loads such as your cutie board. Your 3v measurement on the TX line is evidence that your doing that. You are powering the system up using the ESD diodes inside the chip.

According to the OSD3358 documentation, is not safe to go the other way. You can’t power the OSD in this way safely. I know nothing about the currie board so can’t say anything about it.

Typically it’s safe to do this kind of thing for at least short periods. ESD diodes are designed for high voltage and very short transients. Most of the ESD structures I’ve seen were pretty huge and could tolerate a fair amount of power, but again, you’d want to check with the data sheet of your product.



Thanks for the explanation. I think the solution for me is to power both boards from the same power source so that they are always both on or both off.

One minor problem I have is that when I am downloading a script to the CurieNano, it is powered from the USB connection that is used as the download connection. I think I will fix that by making a USB cable that does not have the +5 volt connection. That will mean if I am downloading a script I will need to use the actual power supply and therefore both boards will be on.

Thanks again.