voltage monitoring

Presuming that it is the current through the coil that you require to maintain constant, I'd be thinking of a current transformer as the first step.

Hi,

Would you please explain in more details about it. How would a current transformer be effective? Can you recommend one? What is the chances of putting a series of voltage dividers to get 200 volts downto 5 volts peak to peak and then use a brdidge rectifer to get the DC voltage output? A re there any high input instrument amplifiers available that can with programmable gain that can do this?

Thanks John

What's the frequency? Power? Signal source? Is one end grounded? Accuracy requirement?

What's the application?

John

hi,

Frequency is 300KHz. 200 peak to peak times 2A. Signal source is HIP

4081 driving four MOSFETS. The other end is not grounded. Accuracy requirent not known yet.

John

Is the inductor the load? Or is there somthing inside the inductor that is getting hot? IOW, Does the load look inductive or is it resistive.

You might want to rectify the siganl first and then use a resistor divider to reduce the voltage. (Less error from the diodes that way.) Though I don't know if big fat diodes are fast enough.

Apex makes some high voltage and high current opamps, but they are not cheap! say ~$50 or more each.

George H.

Hi,

The load is another coil with different number of turns plus a diode rectifier and a resistive load. The resistive load values can change that will also change the amount of current and voltage on the primary side and that is why I need to monitor the primary coil voltage and keep it constant. So, primary coil --------------> air

------------------> secondary coil --------------------> load. Need advice on the feedback monitoring system

Thanks John

Would a simple peak detector work? Transformer 20:1 or something that can be had off the shelf -> rectifier with properly set attack and decay times -> to uC or whatever needs this data.

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.

Hi,

Thanks for the reply but I did not understand the following part of your reply

"rectifier with properly set attack and decay " "times -> to uC or whatever needs this data."

Would you please explain a little more.

John

Assuming you want to measure a peak amplitude, place a rectifier, same as you would in a linear power supply. Make the capacitor large enough to tide you over the cycles but not a lot more. Then you end up with a DC voltage (with some ripple on there) that is proportional to the amplitude of your 200V spike, largely independent of the shape of that spike.

There are many other options such as sample & hold but for that we'd need to see your schematic.

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.

Hi, Can you give me the part number on the rectifier and the capacitor or method to choose them. How about using a voltage divider?

John

Well, if you don't need any isolation you can divide via resistor. But then you might as well rectify the pulses directly, into a cap, then hang a voltage divider across the cap. I normally use the BAS21 diode for such jobs, comes in a 250V variant. If the spikes can go higher than

200V I'd use one with more breakdown voltage. The cap should be a ceramic rated at 250VDC or higher. With the resistive divider being 1M a 1nF would give you a 1msec time constant. If that's too slow a reaction time drop the cap proportionately but you'll see more ripple. You can also reduce the resistor but then you'll start burning more serious power in the resistor.

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.

I typed out the following before reading the responses you already have. It's been pretty much covered by others and probably better...

=============================

That's the easiest. Use a couple of series resistors and then put a small rectifier on the voltage tap. Put a capacitor across the rectifier output to smooth out the ripply rectified output. Measure, so you KNOW it's not over five volts.

Something like a PIC or Arduino microcontroller lets you put in an analog signal, examine it, then switch or vary things according to your needs.

A pdf file:

The Arduino has ten bit resolution on the analog inputs, so you get pretty good resolution (5V / 1024 steps).

It may be a good idea to get some sort of isolation between the 200 volts and the Arduino. Maybe you could drive the base of an opto-isolator or something from the divider circuit.

Remember, this advice is worth what you paid for it.

mike

Hi All,

I am able to monitor the volatge across the coil and also can measure it with the help of an ADC on my micro. Now, I am looking for some kind of algorithm that can help to monitor the changes and act accordingly.

Specs are :

Input signal to the ADC (micro) is 2.5 volts peak (+ve only) and I would like to keep it that way. Now, if it drops to 2 volts than I need to get it to 2.5 volts again. It could be a simple if else instruction in C but I heard that algorithms like PID etc are availablr to this. How can these algorithms be helpful and what should I do to implement them.

Please advice!

Thanks John

Advise what? Not enough context? People make switching converters with PIC chips, is that what you're after? Take a look at the application notes. There are PIC chips with onboard support for PWM, including the RS latch and comparator you need for driving the switch.

Grant.

Hi,

I can do the PWM but I am confused about the control algorithm like would it be an over kill to do this with PID control algorithm. For example, right now I am doing that if the voltage is 2 volts than increase the duty cycle (using PWM) so that the voltage will increase to 2.5 volts and than when it reaches to 2.5 volts than keep the same duty cycle and keep looping to check any drop across the coil voltage. I know it also sounds like a speed control of the motor.

John