Clamp diodes are widely used in
analog electronic circuits. Like all real-world components, these diodes have
some non-ideal characteristics; this paper investigates some of these
characteristics for a few diode candidates. The AMETRIX® Instruments’ Model 101
Picoammeter is used for measuring these small currents.
Reverse biased leakage current
is a characteristic of concern for a clamp diode. Ideally the leakage would be
zero but this is the real world, so what can one live with? In a typical
application there might be two clamp diodes, one to +15 V the other to -15 V
and a common connection to the signal to be clamped. The leakage current drops
as the voltage drops, so if that signal is nearer to +10 V, then the lower clamp
diode would have more leakage than the upper one. So in such an application, a
diode with both low leakage and a fairly flat leakage vs. voltage would be
desirable.
Figure
1 is plotted data of
leakage measurements versus reverse voltage. A quiet DC supply was used along
with an AMETRIX Instruments’ Model 101 Picoammeter to measure the current (and
yes, that is single-digit femptoamps you are seeing).
Comments about reverse leakage:
1.
The SST4391 is rated for 35 V
gate-drain and gate-source
2.
The SST4391 costs $0.057 each
3.
The 2n3904 has a Vcb maximum of
60 V
4.
The MMBT3904LT1G is $0.0231
each
5.
The FLLD261 is $0.0275 each
6.
The BAS16 costs $0.0151 each
7.
The BAT54 costs $0.0204 each
8.
The BAS70 costs $0.0323 each
9.
The MMBD4148CC costs $0.0194
each
So even though there is a 3:1
price ratio between the most and least expensive, they are all < 6¢, so cost
is probably not a serious part of the decision process.
The forward voltage
drop is another diode characteristic to be considered for a voltage clamp.
These measurements were made as follows (the data is presented graphically in figure 2):
1.
Forward bias the diode
with a quiet DC supply that has fine resolution; in this case an EDC calibrator
2.
With function
generator set to pulse, 1 Hz, 1% duty cycle, drive a mercury wetted reed relay.
Connect the relay contacts across DUT such that the DUT is shorted for 990 msec
and unshorted for 10 msec; this is to minimize self heating effects.
3.
Measure forward
voltage with Tektronix MSO4104 scope with Model 100 Series' probes to
minimize the low forward current errors. Set the scope’s bandwidth limit to 20
MHz to minimize noise. Use the scope’s vertical channel’s offset mode and 10 mV
range, adjusting the offset knob until trace is in center of screen, then
document the offset voltage.
In conclusion, some
simple measurements can provide much insight into the components that you think
might be ideal in a given application; and they might reveal some things that
are unexpected.
No comments:
Post a Comment