Dear Sir,

Thank you for your interest. The designs you want are the -G units for
ground fault.

However, the design for 100 uA of current sensing would probably not be
something that can work for these products. 100 uA is extremely small and
generates very small magnetic field. Typically you can expect the ground
fault CTs to measure 3-4 mA of imbalance. The reason for this is every CT
with a core must have a certain amount of energy to excite the core. The
design of the ground fault CTs uses 80% nickel, 20% iron type cores. This
type of core has relatively a small amount of iron that is very easy to
excite, and therefore a 3-4 mA signal will be able to produce a measurable
output. But 100 uA of current would probably produce little if any of
output signal as the majority of the energy would go to exciting the core.

Another issue is at even 3-4 mA, the output at best will be in the 50 mV or
less range, irregardless of burden resistor. Again the reason for this is
that the core at these levels is only partially excited, and the percent of
signal to excitation is very low. As the signal gets larger, the core
becomes fully excited, and the output capability increases dramatically.
The result is that the transfer function is very non-linear until the core
becomes fully saturated. Typically the ground fault CTs are applied such
that a single point of sensing is desired. For human protection, that is
3-4 mA.

This leads directly to your question on using the same CT for ground fault
levels and 100 Amp sensing. You will probably need 2 different CTs for the
two different applications. I have attached an updated spec sheet that
fully explains this. CTs can sense currents up to the point of saturation.
A CT secondary can be shorted, and the current flowing through the secondary will be the primary divided by the turns ratio. This is the most accurate point of a CT as losses are minimal. A resistance is then added to the secondary. As this resistance is increased, the voltage that the CT must
develop to keep the current flowing according to the turns ratio also increases. This results in higher magnetic field densities in the core,
which also results in more losses. You can keep increasing the resistance
until you reach saturation, at which the magnetic field strength in the core
can no longer be increased. This saturation point is a function of core
size and the number of turns on the secondary.

A ground fault CT has minimal iron in it so it excites fairly easily.
However this also has the effect of decreasing dramatically the saturation
point. A general purpose CT uses silicon steel core and is 90% iron. Its
excitation is significantly higher, but its magnetic saturation is much
higher as well. These are typically used to measure above 500 mA.

This is the main difference between ground fault and general purpose CTs.

Irregardless of CT chosen, the smaller the output voltage required, the more
accurate the part.

You can see on the spec sheet that the output saturation level is very much
lower on the ground faults than on the general purpose.

My recommendation is to use the CR8420-1000-G for your ground fault, and if possible, wrap the wire pair through the hole more than once. This will
give you the best chance at sensing below 3 mA. Please note again that you
will need to empirically determine the burden resistor on this application
as the transfer function of a CT at this low is non-linear. You may need an
amplifier to get it to a level that can be read by your A to D.

I would use the CR8449-2000 for the 100 amp sensing. If you keep the
voltage output low enough, it will provide a linear response at 100 amps
input. (You should be ok to run it at 100 Amps, because all the CTS are
designed to handle 4 X Ir continuously.)