hi Mussawar,
Can you post details of how the resistive divider is connected to the VFD, is there a common 0v/Gnd connection between the divider and VFD?

E

I had two different schemes in my mind as I mentioned in my first post (Attachment). One with floating ground and the other with common GND. Fellows recommended a common one. So now, DC source and Micro-controller shares a common ground.

 

As per datasheet. sampling capacitor is 10PF. So 100PF should be enough to compensate high impedance and keep continue?Circuit is continuously reading ADC during operation.

I would go with 10 or 100nF just to make sure.

 

hi,
I agree a common is the easier way to go.
Have you decided to not use an OPA, but a direct resistive connection to your PIC's ADC input
At what rate are you sampling the 1kV?
.
E

 

Since now, by combinig all advice, This schematic comes in my mind. It will provide 5V for ADC.

Any views? please.

 

hi,
I agree a common is the easier way to go.
Have you decided to not use an OPA, but a direct resistive connection to your PIC's ADC input
At what rate are you sampling the 1kV?
.
E

Personally, I like to use OPA but not unnecessarily. If it works fine without OPA, then no need to increase cost and complexity. Micro is running at 16MHz and it will sample continuously.
Regards.

 

Hi @Mussawar
Will the 1000v ever exceed 1000v?
E

 

hi Mussawar,
If you needed to use isolation, i.e.: no common 0V, this is one option, using the HCNR200 analogue Opto-isolator.
E
View attachment 3

 

Will the 1000v ever exceed 1000v?

No. Normally it is between 650V to 850V

If you needed to use isolation, i.e.: no common 0V, this is one option, using the HCNR200 analogue Opto-isolator.
E
View attachment 3View attachment 339009

Another nice option and a good addition to my circuit bank. Graph is pretty straight. Use of HCNR200 along with op-amp is subject to availability and cost. A trade off between quality, safety and cost. I'll check the feasibility.
Regards.

 

Hi M,
I would advise that bottom 10k resistor be a 20k 20 turn trim pot, this will enable you to trim the readings.
If you set the trim pot to give 4.882v at 1000v, this will set the ADC count to 1000 for 1000V, makes the program neater.
E

 

Hi M,
I would advise that bottom 10k resistor be a 20k 20 turn trim pot, this will enable you to trim the readings.
If you set the trim pot to give 4.882v at 1000v, this will set the ADC count to 1000 for 1000V, makes the program neater.
E

Output voltage of divider is not a big problem because it can be manipulated in software. Also I'm using only a small window about <3.5V>.

 

using the HCNR200 analogue Opto-isolator

I have experimented with the HCNR200 but found it not good for accurate measurements.
I use "isolation amplifiers" like this. There are many different kinds. Some do +/-200mA max and some +/-2V. I think there are some that do not do negative voltages. They were built for differential ADCs. Here is an op-amp making a single ended output.
I have not tried it, but I think you could measure the voltage on Pin 7 and in software subtract Pin 6.
The outputs have some high frequency noise so that is why there is a RC filter on the putouts.

 

I have experimented with the HCNR200 but found it not good for accurate measurements.
I use "isolation amplifiers" like this. There are many different kinds. Some do +/-200mA max and some +/-2V. I think there are some that do not do negative voltages. They were built for differential ADCs. Here is an op-amp making a single ended output.
I have not tried it, but I think you could measure the voltage on Pin 7 and in software subtract Pin 6.
The outputs have some high frequency noise so that is why there is a RC filter on the putouts.
View attachment 339019
View attachment 339020

Hi,
I've simulated HCNR200 optocoupler in Proteus and the O/P is a straight line but of course simulation is performed under theoretical and ideal conditions. Since you have tried it practically. What was the result? I mean what output curve you got? Nearly straight or a lot different?
Thanks.