How Much Supply Ripple Can CC430 Tolerate?

Hi Charles,

This thread got today to my attention via a long email chain started between you and Paul. 

Since the post is already pretty old, can you please update us on the status. In the email you mentioned some approaches you plan on taking. Any success?

Also, i assume this thread: http://e2e.ti.com/support/microcontrollers/msp43016-bit_ultra-low_power_mcus/f/166/t/172267.aspx is related to your question here. Have you been able to solve it, or have you figured out that the voltage drop is the root cause?

Regards, Dietmar

As a general note: Details on voltage levels can be found in the datasheet's PMM (Power Management) section. (Starting page 53 on CC430 datasheet, SLAS554 rev F).

Of special interest is Brown Out, but also the SVS (Supply Voltage Supervision). SVS can be user adjusted to fit your needs.

You might also want to check the VCORE voltage to see if the voltage drop propagates to the MSP430 core voltage (Cap).

Charles Gervasi said:

How much ripple can the CC430 radio tolerate?

The short answer is over 1.5V . As long as the supply voltage stays within the specified voltage range and the edge rate is reasonable.

Charles Gervasi said:

150mV of power supply ripple is causing my CC430 to lock up occasionally

Are you sure the ripple is the cause? An addition of simple LC or RC filter should answer this.

If you really need to reduce the ripple you can try Ripple Blocker™ Power Supply Output Noise Attenuators.

http://www.micrel.com/rippleblocker/

( FYI  The Cray supercomputer cards were bench tested in using 1ms power pulse with low duty cycle.

This was needed to prevent the card from overheating. Normally the cards were submerged in coolant.)

Peter

Our approach right now is to take the supply out of power save when operating the radio.  On the next spin I'm putting another 0402 in series with the bead in case I want to try adding resistance.  Filtering out the ripple isn't straightforward.  I suspect the problem is the high edge rate as the voltage increases, not the overall ripple. 

As long as VCC doesn't fall below the minimum operating voltage, a supply ripple shouldn't crash teh MSP.

However, on some MSPs, fast ripple (in V/µs) can cause the brownout circuit to trigger, since due to a racing condition, VCC seems to drop in the comparator when it actually rises.

Also, VCC ripple may have some influence on teh amplitude of the emitted radio waves. This may have impact on modulation as well as range.

For stabilizing VCC, a shunt regulator could be used. Like the TL431. Such a shunt regulator is basically an OpAmp-driven load. It draws as much current as needed to let the voltage break down to a certain level. The breakdown appears on a head resistor. If dimensioned for the peak current of the circuitry, the shunt regulaor will keep the supply current from the power supply constant on the maximum, so before the head resistor, no current ripples appear. Regulation speed is very fast. Only a few Ohm dynamic inner resistance.

However, this approach, as good as it is for a clean VCC (the oputput can even be used as a reference voltage), isn't exactly what one would call low-power. So it is most likely not suitable for battery-powered systems.

p.s.: the TS431BCX (SOT-32, but also available as TO-92 or SOP-8) has an itnernal reference of 2.495V +-0.5% and can regulate an output voltage of up to 37V (by resistor divider) and up to 100mA shunt current. I use this as ADC voltage reference for the MSP, and for the 'voltage good' feedback in our switching power supply.