Internal voltage clamp protection circuit without parasitic...

Details Internal voltage clamp protection circuit without parasitic... Internal voltage clamp protection circuit without parasitic...

1999-04-13

2001-04-10

Sherry, Michael J. (Department: 2836)

Electricity: electrical systems and devices

Safety and protection of systems and devices

Transient responsive

C361S056000, C361S091500

Reexamination Certificate

active

06215637

ABSTRACT:

BACKGROUND AND SUMMARY OF THE INVENTION
The present invention relates to integrated circuit protection circuits.
Background: ESD Protection
In MOS integrated circuits, the inputs are normally connected to drive the gate of one or more MOS transistors. (The term “MOS” is used in this application, as is now conventional, to refer to any insulated-gate-field-effect-transistor (IGFET), or to integrated circuits which include such transistors.) A longstanding problem is that electrostatic discharges (or similar externally generated voltage transients) may break down the thin gate oxide. Once the gate oxide has thus been punctured, the transistor may be permanently damaged. Thus, it has long been conventional to use protection devices on the input pins of MOS integrated circuits. Such protection devices are designed to avalanche (passing a large amount of current, and dissipating the energy of the incoming transient) before the voltage on the input pin can reach levels which would damage the gate oxide.
However, technological advances are leading to the creation of components that are increasingly smaller and faster but also more fragile. The output stages of MOS circuit which, until now, have been capable of standing up to high discharge currents, are becoming vulnerable. In particular, the advantages of the various techniques for improving the performance characteristics of integrated circuits, such as techniques for the thinning of the gate oxide layers, the reduction of the width of the conduction channels of the transistor or, again, the very low doping and small thickness of the drain regions of the transistors are offset by increased sensitivity to over-voltages or discharges, because the breakdown voltages of the junctions or punch-through voltages between drain and source of the MOS transistors become lower, and because the gate oxide is more fragile. (Flow of large currents may lead to generation of hot carriers, which can become trapped in the gate oxide and produce a long-term shift in the characteristics of the device.)
A variety of device structures for protecting integrated circuits against electrostatic discharge have been proposed. See, e.g., the following articles, and references cited therein, all of which are incorporated by reference: Duvvury et al., “ESD: a pervasive reliability concern for IC technologies,” 81 Proc. IEEE 690 (1993); Amerasekera et al., “ESD in integrated circuits,” 8 Quality and Reliability Engineering International 259 (1992); Welsher et al., “Design for electrostatic-discharge (ESD) protection in telecommunications products,” 69 AT&T Technical Journal 77 (1990); Avery, “A review of electrostatic discharge mechanisms and on-chip protection techniques to ensure device reliability,” 24 J. Electrostatics 111 (1990); Greason et al., “The effects of electrostatic discharge on microelectronic devices-a review,”
20
IEEE Transactions on Industry Applications 247 (1984); R. N. Rountree and C. L. Hutchins, “NMOS protection circuitry,”
IEEE Trans. Electron Devices,
vol. ED-32, pp. 910-917, May 1985; the entire annual proceedings of the International Reliability Physics Symposia (“IRPS”) for years 1980 to date; and the entire annual proceedings of the EOS/ESD symposia for years 1979 to date.
Overvoltage Protection
A similar problem arises in systems using multiple voltage levels. For example, an automotive system may use a system/battery power source which is nominally ~14V (and may, for a short time, go as high as ~40V), but have internal components which must be protected from overvoltage, particularly with regard to limiting the voltage applied to the gate oxide to 10-20V. In these cases, there is a risk not only of typical ESD, but that any external pin of the regulated lower-voltage system will be inadvertently shorted to the higher voltage. If this occurs, there is a significant risk of damage to the low-voltage oxide. To prevent this, it is preferable that an overvoltage protection circuit be integrated into the low-voltage system.
One way to protect internal circuitry from the application of a higher-than-normal voltage is the use of a voltage clamp, which limits the voltage passed on to the internal circuitry. This, in turn, reduces or eliminates the chance of damaging the on-die circuitry. Unfortunately, current voltage clamp circuits, which comprise a zener diode configuration, will create a parasitic device when coupled with a driver circuit.
Background: Device Isolation
One of the major problems with integrated power devices is device isolation. In particular, the voltage magnitudes and current transient which can occur with power device operation present isolation difficulties far beyond those encountered in normal small-signal integrated circuits. For this reason some smart power integrated circuits use full dielectric isolation, in which the diffusion of the power device are completely separated from the small-signal devices, i.e., there is no path through semiconductor materials to connect the two. Alternatively, only one terminal of the power device may be connected to the small-signal devices. However, this requires a much more expensive and cumbersome fabrication process.
Background: Parasitic Cross-Talk in Current Circuits
FIG. 5
shows the general context of these protection circuits: in each case a gate drive signal (G
1
or G
2
) is used to control a low-side-driver output NMOS, to drive an output pin DRN
1
or DRN
2
. In such circuits some clamping is needed between node B (the gate connection) and node C (ground) to protect the gate oxide.
Two common prior art circuits are shown in
FIGS. 6A and 6B
, which show typical zener voltage clamps.
FIG. 6A
shows a first sample prior art configuration, in which a string of base/collector-to-emitter diodes is connected between the gate drive terminal and ground. A series resistor R
1
provides current limit protection between the external node and the gate drive node.
FIG. 6B
shows a somewhat similar configuration, except that in this case the diodes used are base-to-collector/emitter connected rather than emitter-connected bipolar transistor structures.
Both of these structures have their difficulties, as shown in
FIGS. 7A and 7B
, which show a sample context in which the circuits of
FIGS. 6A and 6B
might be used. In each diode string, at least one pair of the diodes is back to back, so that conduction will not occur unless the clamp voltage is exceeded. The voltage is clamped using the using the forward voltage drop of the BC-to-E (in
FIG. 7A
) or Vbe (in
FIG. 7B
) of Q
1
, or the reverse breakdown voltage of the BC-to-E (in
FIG. 7A
) or the Veb (in
FIG. 7B
) of Q
2
and Q
3
.
One common problem in using these standard zener claims is that a parasitic device is formed in a standard N-EPI on P-substrate (or, P-EPI on N-substrate, or any junction-isolated process). In these configurations, a parasitic transistor Q
11
will exist between the NMOS drain diffusion of the output drive transistor and the collector diffusion of the diode-connected transistors; this parasitic transistor is most significant in terms of cross-talk on the transistor in the string of diode-transistors which is closest to the gate, Q
1
in this example. The parasitic device is an NPN transistor which is formed by the drain NTANK (emitter) of the NMOS and the collector NTANK (collector) of the zener, and the substrate forms the base P region, as shown in FIG.
7
A.
This parasitic NPN will turn on when one drain of a NMOS is negative and the gate node with the clamp is positive. This can cause significant cross-talk, meaning the voltage on the gate could be pulled low, switching the device off, which could then lead to operational problems. This transistor can be avoided by using a double epitaxial process or by using full dielectric isolation, but of course these technologies are very expensive and are not available in standard integrated circuit processes.
FIG. 7B
similarly shows the parasitic device in combination with the protection structure of
FIG. 6B
, where B-to-CE connected diodes are used. H

Affiliated with

Also associated with

Rating

Internal voltage clamp protection circuit without parasitic... does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Internal voltage clamp protection circuit without parasitic..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Internal voltage clamp protection circuit without parasitic... will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-2555782