MSP430FE42x2 Voltage Values Problem

Power current measurement is hard, because of the wide dynamic range of the signal (hundreds or thousands to one). Power voltage measurement is easy, because the dynamic range is so narrow (2 or 3 to 1 at the most). If things are set up properly you should get a pretty accurate RMS voltage reading from almost anyone's metering solution. However, you need to make sure things really are set up right. The commonest cause of inaccurate voltage measurement is:

• You are powering the meter from the same supply that you are measuring, and
• You have chokes, or other surge and EMI suprression components, in the wiring, and
• The meter's small power draw is enough to cause some fairly constant voltage drop across those suppression components, and
• Your voltage sensing point is on the meter side of the suppression components, and
• Your voltage reading has a constant offset as big as the drop across the suppression components.

Try measuring the voltage directly across your potential divider resistors, using a multimeter, and see it is significantly different from the true supply voltage. If it is, then hunt down the cause of the difference.

Steve

Hi Steve,

Thanks for your suggestions! I scoped out the signals and found the noise levels to be really high. I believe this was causing the problem.

Can you please explain why would there be a problem if I power the meter from the same supply that I am measuring from? If the offset of the power supply is constant, I can always account for it during calibration, right? Please let me know.

Thanks again,
SF 

There is nothing basically wrong with powering the meter from the same supply that is being measured. In fact, this is how the vast majority of applications of this chip work. It just requires care to ensure that a voltage drop due to the meter's own consumption is not affecting the measured voltage signal.

A constant voltage offset isn't easy to allow for at the calibration stage, in volume production. You really want to keep the calibration of the voltage signal gain to a simple one step operation. An offset, even if precisely constant over all conditions (a big if), would require at least a two point calibration.

Dear SF:

     I'm doing the same job using the same reference. But I don't understand some of the

circuit in the reference guide, so that I have burned three MSP430FE4272 chips. Could you

mailed me your circuit  file? Thank you very much.

YCPU

puyuchi@gmail.com

 

Hello,

I'm sorry I won't be able to email you the circuit because of propreitary issues. I would be happy to help with your problem though. Please post what exactly you are doing. It will help others in the forum as well..

SF

Hi,

   That's ok. My problem is how to connect the AVss ground to the measured load.

Does the jump JP1 connecting to AVss in pp. 27 of http://focus.ti.com/lit/an/slaa203c/slaa203c.pdf

must be short or open? And, does the AVss and DVss connect directly of in another way?

Thank you very much.

ycpu

JP1 is shorted in my design. AVSS and DVSS are connected, but I have used a star ground connection as recommended by TI.

What is your V1-? I believe that might cause damage to the IC if not done properly. And, I am just curious, how do you know that you have burned your MSP?

- SF

 

Hi!

   My circuit is as above. When I powered on  the load, the package of MSP430 was broken, and the AVss lines

on the PCB were also burned. I think I shouldn't connect the AVss to the Live line, because AVss connects to DVss

directly.  I don't understand what's "star ground connection" very well. Thank you very much.

In your circuit, you power teh MSP form a battery. So there is no problem with the electric potential of the digital circuit. This would be one of the usual causes of destruction if the MSP is powered by a cheap switched power supply or (worse) a capacitor power supply. If not properly done, the MSP is gone.

What else might happen? Well, you use a current transformer, so there are no potential problems too (if you're usign a shunt resistor, there's too some way to do it awfully wrong).

What next: well, the shunt resistor on the current transformer. A current transformer tries (limited by it iron core) to produce an output current that corresponds to the input current. Even if this means to produce a lighnting arc. (I worked some time as tester in a current transformer procution, and I've seen quite some arcs due to improperly connected shunts). So if the resistor isn't properly connected or raises its value due to the applied power, voltage may rise easily far above what the MSP can bear.

Another possible problem is the 1M resistor. It acts as a voltage divider with the 1.5k. But that measn that it has to withstand U_NL*1M/(1M+1.5k) V, which (depending on 110 or 230V net)  can be as much as 320V. Usual SMD resistors as well as teh smaller through-hole ones aren't suited for this and will immediately break down, applying the full voltage to the MSP.

yuchi pu said:

I think I shouldn't connect the AVss to the Live line, because AVss connects to DVss

This is not a problem. As long as there is no connection that connetcs AVSS somehow with neutral line too...
With your setup, the MSP will 'ride' on the line voltage, but this only matters in relation earth. If you turn the connector (switching L and N) things are twisted and the MSP will ride on GND, not exposing (usually) any potential to earth. But that only matters if you can touch the circuitry, it has no influence to the circut itself. Voltage is relative. And the relative voltage applied to the MSP is 0 in both cases. You shouldn't touch it, however, if wearing an ESD protection cord and shouldn't put the circuit onto an ESD table too while connected.

yuchi pu said:

I don't understand what's "star ground connection" very well

It means that all points where AVSS or  DVSS are used, have their own wire going to a common point. The main reason for this is to avoid cross-influences. If you have one 'bus' for VSS where everything is connected too, it may happen that soem current flowing from connection A to GND will cause a voltage offset on t he bus, and every AVSS seen by any part of the circuit will se a different AVSS, depending on its position. If every AVSS connection comes form the same point, all points share the same AVSS potential. This is too why you usually don't connect AVSS and DVSS near the processor, but near the power supply. So the supply current for the processor, which flows back to the power supply, will not influence the AVSS level.

Hello.

I really wonder whether you have completed your project?

 

Thank you for your concern.

We have finished the project. 

The circuit posted before is correct, but we made some mistakes on breadboard layout.

Post the final working connection below.

Hello,

I need help understanding how the example code works. I am using a different LCD and I would like to know how the results from the ESP430 are returned? Should I just read directly from the RTx registers? I am completely lost please help