[PIC] Ac voltage monitor using pic 16f886

[Scolioza]

Hi all.

I want to design an AC (230V 50Hz) voltage monitor using PIC 16f886 uc. I read a lot of topics about this problem and i want to ask here some questions. Because one picture worths more than 1000 words i will attach a picture with my calculations:



1. It is ok to use the circuit in the picture? I know that there is no insulation. If i will use a transformer how it will affect the precision?

2.Using a 20Mhz crystal osc. will result 1 instruction in 0.2us. Using Frc for adc will result ~6us/bit. TAD=6X11=66 us. I dont understad if Tacq is the value for 1 bit (it will multiply by 11) or for the entire convertion.

3. Assuming that Tacq is the value for one bit (worst case) and it is ~4us and ts is 200us (according to picture) between two sampling will remain 200-(6x11+4*11)=90us for instructions. Hence 1 instruction is executed in 0.2us will result 450 instruction to execute before taking the next sample. The RMS voltage is :

√((V1²+V2²+...+V100²)/100).

I am not good in assambler so i am using mikroC. Suppose that i am using the following algorithm:

a) generate a timer1 interrupt every 200us
b)after adc will reamain 450 instructions to execute before the next interrupt

Is there enough time to convert digital value back in analog domain, raise to the power of two and add to the previous value?

Thanks in advance!

 

[udayan92]

1.If you are using the full wave rectifier circuit as shown..then the only problem will be the maximum magnitude of voltage available to you on the output terminals will be reduced by 0.7V.
2.Tacq is for the entire 11 bits of the ADC...I think.It is for the entir conversion.

 

[Scolioza]

Not with 1.4 V? Because there are 2 diodes conducting.
Thank you.

 

[schmitt trigger]

Fortunately for Scolioza, the diode bridge is located prior to the resistor divider, and therefore the 1.4 volt loss is also divided down by the resistor ratio: 1.4/100 =0.014v.

However, the way I would do it is to use a transformer (for safety) and use a precision rectifier which in its simpler configuration, only takes an opamp package (LM358), two diodes and 3 resistors.

 

[deepeshmishra]

Using transformer with precision rectifier is safe and good, but it increases cost of one transformer and (two / one) opamp circuit.

 

[Scolioza]

Thank you all.
But what about timing? I mean PIC software?

 

[schmitt trigger]

Using transformer with precision rectifier is safe and good, but it increases cost of one transformer and (two / one) opamp circuit.

My personal policy when providing advice, is to give best engineering practices, and allow the person receiving the advice how much of it is relevant to his particular view.

 

[deepeshmishra]

My personal policy when providing advice, is to give best engineering practices, and allow the person receiving the advice how much of it is relevant to his particular view.

I also like to work on same policy but when the cost matters we have to think for cheaper in this competitive market.