HCPL-7510 Datasheet, PDF (15/16 Pages) Agilent(Hewlett-Packard) – Isolated Linear Sensing IC

FREQUENTLY ASKED QUESTIONS ABOUT THE HCPL-7510
1. THE BASICS
1.1: Why should I use the HCPL-7510 for sensing
current when Hall-effect sensors are available which
don’t need an isolated supply voltage?
Available in an auto-insertable, 8-pin DIP package, the
HCPL-7510 is smaller than and has better linearity, offset
vs. temperature and Common Mode Rejection (CMR)
performance than most Hall-effect sensors. Addition-
ally, often the required input-side power supply can be
derived from the same supply that powers the gate-drive
optocoupler.
2. SENSE RESISTOR AND INPUT FILTER
2.1: Where do I get 10 mΩ resistors? I have never seen
one that low.
Although less common than values above 10 Ω, there
are quite a few manufacturers of resistors suitable for
measuring currents up to 50 A when combined with the
HCPL-7510. Example product information may be found
at Dale’s web site (http://www.vishay.com/vishay/dale)
and Isotek’s web site (http://www.isotekcorp.com) and
Iwaki Musen Kenkyusho’s website (http://www.iwaki-
musen.co.jp) and Micron Electric’s website (http://www.
micron-e.co.jp).
2.2: Should I connect both inputs across the sense
resistor instead of grounding VIN- directly to pin 4?
This is not necessary, but it will work. If you do, be sure to
use an RC filter on both pin 2 (VIN+) and pin 3 (VIN-) to
limit the input voltage at both pads.
2.3: Do I really need an RC filter on the input? What is
it for? Are other values of R and C okay?
The input anti-aliasing filter (R=39 Ω, C=0.01 µF) shown in
the typical application circuit is recommended for filtering
fast switching voltage transients from the input signal.
(This helps to attenuate higher signal frequencies which
could otherwise alias with the input sampling rate and
cause higher input offset voltage.)
Some issues to keep in mind using different filter resistors
or capacitors are:
1. (Filter resistor:) The equivalent input resistance for
HCPL-7510 is around 700 kΩ. It is therefore best to
ensure that the filter resistance is not a significant
percentage of this value; otherwise the offset voltage
will be increased through the resistor divider effect.
[As an example, if Rfilt = 5.5 kΩ, then VOS = (Vin * 1%)
= 2 mV for a maximum 200 mV input and VOS will vary
with respect to Vin.]
2. The input bandwidth is changed as a result of this
different R-C filter configuration. In fact this is one of
the main reasons for changing the input-filter R-C time
constant.
3. (Filter capacitance:) The input capacitance of the HCPL-
7510 is approximately 1.5 pF. For proper o p e r a t i o n
the switching input-side sampling capacitors must
be charged from a relatively fixed (low impedance)
voltage source. Therefore, if a filter capacitor is used
it is best for this capacitor to be a few orders of
magnitude greater than the CINPUT (A value of at least
100 pF works well.)
2.4: How do I ensure that the HCPL-7510 is not
destroyed as a result of short circuit conditions which
cause voltage drops across the sense resistor that
exceed the ratings of the HCPL-7510’s inputs?
Select the sense resistor so that it will have less than 5 V
drop when short circuits occur. The only other require-
ment is to shut down the drive before the sense resistor
is damaged or its solder joints melt. This ensures that the
input of the HCPL-7510 can not be damaged by sense
resistors going open-circuit.
3. ISOLATION AND INSULATION
3.1: How many volts will the HCPL-7510 withstand?
The momentary (1 minute) withstand voltage is 3750 V
rms per UL 1577 and CSA Component Acceptance Notice
#5.
4. ACCURACY
4.1: Does the gain change if the internal LED light
output degrades with time?
No. The LED is used only to transmit a digital pattern.
Avago Technologies has accounted for LED degradation
in the design of the product to ensure long life.
5. MISCELLANEOUS
5.1: How does the HCPL-7510 measure negative
signals with only a +5 V supply?
The inputs have a series resistor for protection against
large negative inputs. Normal signals are no more than
200 mV in amplitude. Such signals do not forward bias
any junctions sufficiently to interfere with accurate