Hunting noise in mixed systems

At some point most of us need to find and get rid of noise and spurious signals. While for some the source could be the neighbor’s dog barking at 3 a.m., for those of us in electrical engineering it’s more likely (and no less concerning) to be switching power supplies, digital noise from other parts of the system, and external sources.

If you’re designing for battery-powered systems, a switching regulator is typically used, as it has the right balance between efficiency and cost. I recently came across an example involving an embedded wireless system. I’m sharing this example in the hopes that some of my findings may help you as you try to find new ways to deal with noise in a design. Sorry, I can’t help with the barking dog.


For the following discussion, the device under test uses a flexible radio IC already integrated into a module for radio test, the Microchip Technologies MRF89XM8A. This module incorporates the MRF89XA radio IC along with filtering and antenna matching. For demonstration, this module is mounted to a Microchip Explorer 16 board and works with a PC to program the setup of the radio.
We constantly make concessions regarding power, speed, cost and many other factors. So what are the trade-offs of noise versus efficiency while using a switching supply? In this case, the converter switches at about 500kHz, which is common for switching regulators. It can provide the 3.3V output voltage needed by the radio module with an input voltage down to 0.8V. This means the radio can be powered from a single cell, reducing the size of the battery for the product.

Here’s the test setup with the signals that will be measured on a mixed-domain oscilloscope, looking at the analog and RF signals in a time-correlated manner:


First we establish the “golden” reference operation with a laboratory-grade power supply, as shown below. The green trace shows the current ramp up from 0 to about 40mA, which is expected as the radio turns on. The yellow trace, at about the same point, shows a slight dip, indicating a minor amount of ripple when the load is introduced. The orange bits show the frequency-shift keying over time (preamble, data payload, etc.). Near the middle of the screen there is an orange rectangle. This indicates where the RF spectrum (shown in the lower half of the window) is relative to the current and voltage signals. Overall, the system is operating fine.


The screen shot below shows the same RF signal but with a boost-type switching supply powering the radio module. Boost regulators are known for generating noise but are valuable to allow use of a battery with one or two alkaline or NiCad cells and relatively few components, lowering cost. Note the increase in noise at the base of the modulated signal (~5dB higher near the peak). The noise is also readily apparent in the current and voltage waveforms. This noise affects the receiver signal-to-noise ratio and would probably reduce the effective range of the system.


Needless to say, we’re enjoying putting the new mixed-domain oscilloscope through it paces. It’s a powerful tool for debugging embedded systems that include an RF component.