Power meter (cs5463) design problem - Help needed!

[rprodrigues]

Hi all,


I have been facing problems with my power meter design.

I don't know why, but I have already lost 4 units of the power meter ic I use (CS5463 form Cirrus Logic).

I noticed that even with no load applied to the circuit (hall sensors), sometimes, the CS5463 ic is lost. For the others times the ic was lost, there was load applied to the hall sensors (current and voltage).

The IC receives signals from the opamps and communicates with an PIC18F2620 through SPI interface.

I don't know if the I need to implement any protection for the power meter ic. I guess I should, but I don't know how and what one.


Any help is more than appreciated.

Here is the schematic:
**broken link removed**


Thank you,
Rodrigo

 

[eblc1388]

Please check the max allowable analogue input voltage level of CS5463 Vin+ and IIn+ pins. A brief look at the CS5463 datasheet reveals that the analogue voltage should be less than VA+0.3V max.

In your circuit, those pins are connected to the Opamp OPA2277PA (IC1B and IC2B) outputs and these Opamp output pins can briefly go up to Opamp supply voltage level when the circuit is first powered ON or whatever, which is near +15V.

I think this could be a possible cause of the destruction for CS5463.

 

[rprodrigues]

Please check the max allowable analogue input voltage level of CS5463 Vin+ and IIn+ pins. A brief look at the CS5463 datasheet reveals that the analogue voltage should be less than VA+0.3V max.

In your circuit, those pins are connected to the Opamp OPA2277PA (IC1B and IC2B) outputs and these Opamp output pins can briefly go up to Opamp supply voltage level when the circuit is first powered ON or whatever, which is near +15V.

I think this could be a possible cause of the destruction for CS5463.

Hi eblc1388,

Thank you very much for your reply. I think you are right and I will follow your instructions.

By the way, should I protect the Vin+ and IIn+ pins against only positive over-voltage or against both positive and negative over-voltages?

Could I use zener diodes plus resistors to do that (resistor to limit current and zener to cut over-voltage signals), or do you suggest me something else?

I thought in supplying the op-amps with the same positive power supply applied to the CS5463 but I guess it may not be enough. What do you think?


Thank you,
Rodrigo

 

[eblc1388]

By the way, should I protect the Vin+ and IIn+ pins against only positive over-voltage or against both positive and negative over-voltages?

Both.

Could I use zener diodes plus resistors to do that (resistor to limit current and zener to cut over-voltage signals), or do you suggest me something else?

It would be better to use a series resistor plus two diodes. Connects resistor to the output pin of the Opamp and the other end to the CS5463 input, and diodes from CS5463 pin to VA+ and 0V respectively to limit the swing of the possible input voltage.

This is a common and effective means to limit the swing of input voltages between the power supply and 0V of the down stream CS5463 IC. Since the inputs are CMOS with impedance of min. 2Mohm, you can try 100K and two 1N4148 silicon signal diodes. You'll need to repeat this on the other input pin.

I thought in supplying the op-amps with the same positive power supply applied to the CS5463 but I guess it may not be enough. What do you think?

Correct. One wants sufficient supply headroom on the opamp supply so that it can perform its job more effectively.

 

[rprodrigues]

Both.



It would be better to use a series resistor plus two diodes. Connects resistor to the output pin of the Opamp and the other end to the CS5463 input, and diodes from CS5463 pin to VA+ and 0V respectively to limit the swing of the possible input voltage.

This is a common and effective means to limit the swing of input voltages between the power supply and 0V of the down stream CS5463 IC. Since the inputs are CMOS with impedance of min. 2Mohm, you can try 100K and two 1N4148 silicon signal diodes. You'll need to repeat this on the other input pin.



Correct. One wants sufficient supply headroom on the opamp supply so that it can perform its job more effectively.

Hi again,


Thank you for your advices!!!!


I implemented it and it seems to be fine because all the tests with over-voltages were ok!

I used the following diagram, as you told me:

**broken link removed**







Since the input range for both channels is 500 mVpp (peak-to-peak) and the diode threshold is about .6V (above the input range of +/- 250 mVp), I also thought in using the following idea:

**broken link removed**



What do you think about this idea? What one could be better?



Thank you very much,
rprodrigues

 

[Nigel Goodwin]

What do you think about this idea? What one could be better?

I would say the first one is better, it's the standard way of doing it.

 

[eblc1388]

Since the input range for both channels is 500 mVpp (peak-to-peak) and the diode threshold is about .6V (above the input range of +/- 250 mVp), I also thought in using the following idea:

Well, any range limiter on the signal would possibly affect accuracy and performance. The two diodes in the second circuit is a limiter, with threshold very close to the signal range. Also remember that a diode might start to conduct below 0.6V because the conduction process with forward applied voltage follows a curve unlike that of Zener diode.

The first circuit is not limiting to signal as such but placed there to "prevent damage" to the chip. The diodes are always reverse biased so leakage current is small compare to the second circuit.

 

[rprodrigues]

I would say the first one is better, it's the standard way of doing it.

Thank you Nigel!!!!

 

[rprodrigues]

Well, any range limiter on the signal would possibly affect accuracy and performance. The two diodes in the second circuit is a limiter, with threshold very close to the signal range. Also remember that a diode might start to conduct below 0.6V because the conduction process with forward applied voltage follows a curve unlike that of Zener diode.

The first circuit is not limiting to signal as such but placed there to "prevent damage" to the chip. The diodes are always reverse biased so leakage current is small compare to the second circuit.

I am going to use the first circuit.

Thank you very much!

 

[mpo9999]

I have the Problem with 4-20mA to 1-5V conversion

Hi rprodrigues
I have the Problem with 4-20mA to 1-5V conversion and I search on internet for solve this problem That I see some topic in this website that topic is post by rprodrigues .URL of this topic is
"https://www.electro-tech-online.com/threads/possible-4-20ma-to-1-5v-conversion-problem.32377/"

Please You help me for solve this problem
I want to know How you fix this problem on your circuit

Thank you very much