This is the application circuit from the datasheet and not necessarily the actual circuit or values in the power supply module that I am using. The green line (switching node) is where I attach the wire needed to the power supply module.
The negative power supply rail uses a charge pump that leeches from the switching node in the module. Alternatively, a switched cap inverter could be used to provide a negative voltage. This is the switching node waveform: 14.6V square wave.
Here is the AC noise at the output of the charge pump left of the ferrite bead L2. The filter cap C6 is doing a pretty good job.
Here it is on the right hand side of L2. There are still some narrow spikes.
Here is at the output of the regulator. The regulator remove most of the ripples, but the high frequency spikes still comes through. Note: Scope is on a more sensitive voltage scale.
Here is an improvement on the layout. By shorting the 2 ground pour areas with the red trace, I have reduced the return path (in purple) for the charge pump circuit. This stops the RF current going up towards the regulator. As a result, it eliminated the voltage spikes on the negative rail.
Here is the noise at the negative regulator output with the ground mod. This is not too bad for something that starts out as 14.6V square waves!
Here is the positive rail:
'll throw in a couple of RC filter on the supply rails at analog module.
No comments:
Post a Comment
Note: Only a member of this blog may post a comment.