Inconsistent voltage readings from RoboClaw Solo 300

General discussion of using Roboclaw motor controllers
Post Reply
Posts: 3
Joined: Wed Jul 07, 2021 2:30 pm

Inconsistent voltage readings from RoboClaw Solo 300

Post by DosMan » Fri Jul 23, 2021 11:01 am

I have a microcontroller running a RoboClaw Solo 300, and it regularly reads the battery voltage. With the USB plugged in and Motion Studio connected, both my MCU and Motion Studio are getting a reading of 26.5V when 30.00V is applied. This is surprisingly inaccurate, but I found the error is fully correctable over a wide voltage range (I tested 16 - 30V) with a simple linear gain and offset calibration.

However, as soon as the USB is disconnected, the value changes. The MCU reads 27.6V when 30.00 is applied. Different calibration values can again correct it over a wide voltage range (it seems the gain is different, but the offset is about the same). But I would like to get accurate readings whether or not the USB is connected. Any idea what is going here? The reading changes upon connection of the USB cable; Motion Studio does not need to be running.

User avatar
Basicmicro Support
Posts: 1592
Joined: Thu Feb 26, 2015 9:45 pm

Re: Inconsistent voltage readings from RoboClaw Solo 300

Post by Basicmicro Support » Tue Sep 14, 2021 12:38 pm

The USB 5v bus is causing a slight change in the 3.3v power rail. This rail is also used to measure the battery voltage. The reason this is happening is the 5v is clamped through a 100ohm resistor and diode.

This will be corrected in the firmware. Basically we can detect when the USB is connected and can adjust the AD offset appropriately. We are currently working on the next firmware release and I hope to have that included in it.

Post Reply