# How can I convert ac current to near linear dc voltage

Discussion in 'General Electronics' started by ., Jul 26, 2004.

1. ### .Guest

I want to purchase or construct 3 clip on current transformers that
will produce a small current in a parallel connected resister. This
voltage will be rectified using full wave rectifier, and then passed
to the input of a voltage monitor card that will used to monitor line
current in the cables that are running through the transformers.

First of all will this work? I am wondering if there will be a
linearity problem due to the voltage drop in the diodes or maybe
something else. The input resistance of the monitor card is 1 Mohm
and maximum 30 Vdc.

Can anyone see other problems with this design. Accuracy needs to
be within 2 or 3 %.

Chris

To send me e-mail remove the sevens

Chrisd

2. ### Robert BaerGuest

First off, you do not state the current range that you wish to
monitor.
Secondly, you do not state of you would like 2% or 3% over the whole
range (ie: from zero) or just full scale accuracy.
Therefore.
Yes, a current transformer will allow monitoring in a fashion you
suggest, and can allow isolation of the monitor from the line voltage.
The ideal case is to have the secondary shorted, and monitor that
current; a low resistance can be used and be a very good and useable
approximation; especially if a large turns ratio is used.
Say the turns ratio is 1,000 and a toroid is used; the input (primary)
is a wire going thru the hole once.
In this case, every amp in the primary would induce one milliamp in a
shorted secondary; you can see that use of a "large" resistance load in
the secondary (to develop a voltage for monitoring) can and will reduce
that current.
But as long as the load resistance is fairly low, and linear, the
device can be calibrated for use.
If the currents are low, then use of an opamp to act as a short to the
secondary and convert that current to a decent AC voltage would be of
great help.
As long as the frequencies involved are reasonable, an opamp ideal
rectifier (FW or HW) could be used to "remove" the nonlinearity of the
diode forward loss.
The net result of this, is a possible good linearity and accuracy

3. ### .Guest

I am looking to monitor wires running currents in the 10A - 200A
range. I do not know what turns ratio would be in the CT's because I
do not have them yet.

Is there a more specific name for this op-amp circuit? I would lke to
read up more on how this is done so that I can experiment a bit with
it. I did an internet search and found some information but mostly
descriptions and no diagrams. Did I mention that I was looking for
voltages of 30Vdc or less for the monitor interface card?

I appreciate you help by the way.

To send me e-mail remove the sevens

Chrisd

4. ### Robert BaerGuest

You can make your own CTs with toroids or power transformers; 1:100 to
1:1000 turns ratio would seem to be a reasonable range.
That is to say, if you could find (or wind) a toroid for 1000 turns or
find a power transformer with about that many turns for the primary (ie:
a small 5 watt 50-60Hz transformer for about 1000 turns; a 500 watt for
100 turns).
Strip off the secondary of the power transformer, to give room for the
one turn primary.
If you were to wind the transformer/toroid yourself, then you would
know the ratio for calibration correlation purposes; otherwise you would
need to use some kind of standard.
For circuit and other opamp references, check:
"Handbook of Operational Amplifier Applications" 1963 by Burr Brown
"Operational Amplifiers - Design and Applications" 1971 by Jerald
Graeme, Gene Tobey and Lawerence Huelsman (McGraw Hill)
There are undoubtedly many other good references, but these are the
best i have in my limited library.