RADIATION TOLERANT ELECTROSTATIC DISCHARGE PROTECTION NETWORKS
An ESD network. The ESD network including a redundant voltage clamping element in series with a first voltage clamping element between two voltage pads. The ESD network may be connected to a power voltage pad or a signal voltage pad either directly or through a dummy voltage pad. The voltage clamping elements may further comprise an array of unit cells wherein the array is electrically equivalent to single large transistors currently used in ESD networks. By creating an ESD network as an array of unit cells, benefits greater than those obtained by using a single transistor as a clamping or a trigger element are realized such as increased ballast resistance and less overall damage to the circuitry resulting from cosmic rays and particles.
This application is a division of copending U.S. patent application Ser. No. 11/162,999 filed on Sep. 30, 2005.
FIELD OF THE INVENTIONThis invention relates generally to the field of protection of semiconductor circuits from radiation effects, and in particular relates to creating electrostatic discharge (ESD) circuits that are tolerant of single event upsets.
BACKGROUND OF THE INVENTIONSemiconductor circuits are composed of a regular arrangement of atoms, usually silicon. In semiconductors, the arrangement of the atoms is often in a lattice or mesh arrangement and the electrons and the holes created by the lack of electrons are free to move within planes of the lattice. As semiconductor devices become smaller and smaller, fewer and fewer atoms are used to create an electronic device. Transistors of semiconductor materials often store sufficient charge to create digital information by storing a 0 or a 1 state depending upon the threshold voltage of the transistor.
The primary component of nature is space and so there is mostly space between the atoms in a semiconductor lattice. Also, as in all of nature, space is not quiet but rather active with high energetic particles including heavy ions and cosmic rays from solar or galactic origins moving about, colliding, transferring energy. These particles and/or rays collide with atoms in the semiconductor lattice and through elastic, inelastic collisions, and/or field interactions between the energetic particles and the atoms and/or electrons in the lattice, the atoms and/or electrons in the lattice may either change energy and/or position such that the electronic state of the semiconductor device is no longer reliable. For instance, when the intruding particle is near a p-n junction, it may induce a soft error, or single-event upset because of the excess electron-hole pairs generated. If the electromagnetic field in the neighborhood of the p-n junction is sufficiently strong, the charged electrons and holes separate to a nearby device contact and when or if the collected charge exceeds a critical threshold value, a random signal is registered. This is not an unusual occurrence. The Van Allen belt is a small region of high-energy particles held captive by the magnetic influence of the Earth approximately 4000 miles or so above the Earth's surface. The Van Allen belt consists mainly of high-energy protons, ten to fifty million electron volts (10-50 MeV), which are by-product of cosmic radiation. The probability of a nuclear hit between an energetic neutron interacting with a silicon large scale integrated (LSI) circuit is one out of 40,000 incident neutrons will interact within 10 microns of the circuit; further calculations reveal that at sea level almost every silicon-neutron hit within one micron of a LSI circuit results in a soft error of single-event upset. During a quiet sun period, the primary flux of particles averages seventy percent protons and thirty percent neutrons, but during an active sun period the number of solar particles hitting the outer atmosphere increases a million fold, and is larger than the flux of intergalactic cosmic rays. Thus, soft-errors or single event upsets are a real concern for satellites and satellite-based communications and technology.
A single event upset (SEU) is defined by NASA as radiation-induced errors in microelectronic circuits caused when charged particles lose energy by ionizing the medium through which they pass, leaving behind a wake of electron-hole pairs. SEUs are transient soft errors and are non-destructive. A reset or rewriting of the device results in normal device behavior thereafter. SEUs typically appear as transient pulses in logic or support circuitry, or as bit flips in memory cells or registers. A multiple-bit SEU occurs when a single ion hits two or more bits causing simultaneous errors. Multiple-bit SEU is a problem for single-bit error detection and correction where it is impossible to assign bits within a word to different chips, such a dynamic random access memories (DRAMs) and certain static random access memories (SRAMs). A severe SEU is the single-event functional interrupt in which an SEU in the device=s control circuitry places the device into a test mode, halt, or undefined state.
A single event latchup (SEL) is a condition that causes loss of device functionality because of a single-event induced current state. SELs are hard errors, and may cause permanent damage to the device. SEL results in a high operating current above the device specification and the latched condition can destroy the device, drag down the bus voltage, or damage the power supply. An SEL can be cleared by a power off-on reset or power strobing of the device, but if the power is not removed quickly, catastrophic failure may occur because of excessive heat, or metallization or bond failure.
Single event burnout (SEB) is a more severe condition that causes device destruction because of high currents in a power transistor. SEBs include burnout of power MOSFETs, gate rupture, frozen bits, and noise in charge-coupled devices. An SEB can be triggered in a power MOSFET biased in the OFF state when a heavy ion passing through deposits enough charge to turn the device on. SEBs can also occur in bipolar junction transistors.
The trend towards reducing device size and power; increasing the line resolution, memory, and speed of electronic devices only heighten SEU susceptibility. Integrated circuits already include electrostatic discharge networks, the purpose of which is to dissipate any static electricity or other high voltage problems during manufacturing, construction, handling, shipping, etc. In space, however, ESD networks actually create problems and concerns, especially when the ESD networks are between the power rails, because of the shorting and physical damage as presented above when bombarded by particles and/or cosmic rays in space. Bipolar circuits in electronic space applications have a high probability of being hit by a cosmic ray and therefore require high reliability. This invention addresses the need to create reliable ESD networks for space applications.
SUMMARY OF THE INVENTIONThese needs and others are satisfied by first realizing that electrostatic discharge networks pose a hazard to the integrity of the circuitry in space. Upon that realization, then an electrostatic discharge network has been invented, the network comprising: a trigger element; a voltage clamping element connected to and initialized by the trigger element; and a redundant element in series with the voltage clamping element in which the trigger element, the voltage clamping element and the redundant element are located between a first voltage pad and a second voltage pad. The voltage clamping elements and/or the redundant elements may be a bipolar transistor, or a MOSFET transistor. There may also be a redundant trigger element either in series or in parallel with the first trigger element. Additional protection can be realized by connecting the ESD network to a dummy voltage pad or voltage rail not connected to processing circuitry.
In another embodiment of the invention, the clamping element further may comprise an array of unit cells, each unit cell comprising at least one transistor, the sum of the transistors of the array being electrically equivalent to a single clamping element, each unit cell separated from another unit cell by a distance at least as great as the distance a colliding proton could displace in the network. There may be barriers or trenches between the unit cells of the array.
A fault tolerant ESD network adapted to reduce the incidence of a system failure caused by a single event upset, comprises a first trigger circuit element; a first voltage clamp element connected to the first trigger circuit element; and a redundant voltage clamp element coupled in series with the first voltage clamp element. The first voltage clamp element and/or the redundant voltage clamp element and/or the trigger element and/or a redundant trigger element may actually be an array of unit cells, each unit cell comprising a plurality of transistors, the electrical sum of the unit cells being electrically equivalent to a single redundant voltage clamping transistor. Each unit cell may be different than another unit cell in the array.
The invention is also considered a method to design an ESD protection network, comprising the steps of: defining a reliability requirement of an integrated circuit for a single event upset; defining an electrostatic discharge requirement using a technology database file and ESD data; defining circuit topology of a redundant ESD element in series with a voltage clamping element of the ESD protection network to be applied between a first voltage and a second voltage; defining a unit cell to comprise the redundant ESD element using a SEU tool; and evaluating the probability of failure of the ESD protection network.
The invention is also a method to protect electronic circuitry in space, by first realizing that electrostatic discharge networks can cause circuitry to experience single event upsets or worse in space, and then including a redundant voltage clamping element in series with a first voltage clamping element between two voltage pads in the electrostatic discharge networks; and connecting at least one trigger element to the redundant voltage clamping element and the first voltage clamping element. The advantages of the invention set forth will further be realized upon a careful reading of the Description of the Invention in conjunction with the included Drawing.
As discussed earlier, the collision of heavy ion particles often leads to secondary breakdown events in space caused by high energy protons and neutrons colliding with the silicon lattice of electronic circuits leading to fission fragments and damage to the electronic devices. Electrostatic discharge networks (ESD) can be used to quickly remove power from a circuit resulting from bombardment by cosmic rays but when the ESD network itself is hit, then catastrophic failure may occur because of excessive heat, or metallization or bond failure. Bipolar-based ESD networks, moreover, actually fail in space applications. Hereinafter the distinction between SEUs, SELs, and SEBs shall be merged and the event, whether it be a latch-up or a burn-out, shall simply be referred to a single event upset (SEU). Until now, the failure of ESD networks in space from SEUs has not been addressed.
An ESD network having redundant elements can be used in a variety of semiconductor technologies, typically of silicon but also any Type III/V, Type II/IV semiconductor combinations, including but not limited to gallium arsenide, silicon germanium, indium phosphide, silicon-germanium-carbon, silicon-on-insulator, silicon bipolar etc. In fact, it is also realized that such an ESD network with a redundant element as described herein is applicable to an ESD network of MOSFET transistors because they inherently have a parasitic silicon bipolar transistor. SEUs are particularly problematic with bipolar transistors, and homo- and heterojunction transistors because these transistors have narrow regions that make it easy for the collector to short with the Vdd, or the emitter to short to ground.
One embodiment of a radiation tolerant ESD network has an ESD redundant circuit element in series or in parallel with the trigger element. Consider the simplified circuit diagram of
Similarly, trigger element 120 would also be considered a first stage, with clamping elements 130, 432, 442 as a second stage. The redundant second stage is in a series cascode arrangement with the clamping second stage, i.e., each redundant element is in series with a voltage clamping element between the pad, power voltage Vdd 110, and its respective clamping element 130, 432, 442. Redundant trigger element 410 is connected to the second stage SEU redundant elements 330, 430, 440. In this embodiment, the redundant trigger element 410 is parallel to the trigger element 120 for the clamping elements 130, 432, 442. In some circumstances, it is desirable to bias the SEU redundant element, for example, the redundant trigger element 410 can be grounded with resistor 412. Clamping elements 130, 432, 442 may have substantially the same or may have different voltage clamping capabilities from each other and from their respective redundant elements 330, 430, 440.
Whether the radiation tolerant ESD circuit is shown as in
With respect to
The advantages of such an array 700 replacing each trigger, clamping, and/or redundant element is that the charge collected by each transistor is less, the probability of a single transistor experiencing an collision is less, and the increase in resistance is greater; these advantageous factors combine to reduce the probability of a SEU event (σ(π,Σl), σ(ν,Σl) event) in an ESD network. The self-ballasting effect of the array 700, moreover, protects more than merely adding redundancy and further limits the current of a given element if a SEU-induced failure occurs. The ESD array 700 further allows customization and personalization for the space environment, and parts per million (PPM) reliability requirement of space application and ESD requirements. The ESD circuit still has the requisite resistance but it is digitated into a plurality of resistances to decrease the probability of being hit and creating a natural series resistance that does not draw as much current as a large single transistor.
An ESD array 700 contains a plurality of unit cells 710, 720. The size of and the distance between the unit cells 710, 720 affect the probability of failure of more than one unit cell. The distance between the unit cells is preferably larger than the range of particle distribution and the length of track of the particles. For example, an alpha particle may travel up to 80 microns, so it is preferred that the distance be on the order of or greater than 80 microns.
The size of the unit cell is actually dependent upon many considerations as will be discussed with respect to
It will be appreciated that variations of some elements are possible to adapt the invention for specific conditions or functions. The concepts of the present invention can be further extended to a variety of other applications that are clearly within the scope of this invention. Having thus described the present invention with respect to preferred embodiments as implemented, it will be apparent to those skilled in the art that many modifications and enhancements are possible to the present invention without departing from the basic concepts as described in the preferred embodiment of the present invention. Therefore, what is intended to be protected by way of letters patent should be limited only by the scope of the following claims.
Claims
1. A fault tolerant electrostatic discharge network, comprising:
- a first trigger element;
- a first array of unit cells initialized by said first trigger element, each unit cell of said first array comprising a plurality of transistors, an electrical sum of all unit cells of said first array being electrically equivalent to a first single voltage clamp element connected to said first trigger circuit element;
- a second trigger element, said first trigger element coupled in series with said second trigger element; and
- a second array of unit cells initialized by said second trigger element, each unit cell of said second array comprising a plurality of transistors, an electrical sum of all unit cells of said second array being electrically equivalent to a second single voltage clamp, said second array coupled in series with said first array.
2. The fault tolerant electrostatic discharge network of claim 1, wherein every unit cell of said first array is different.
3. The fault tolerant electrostatic discharge network of claim 1, wherein every unit cell of said first array of unit cells is the same.
4. The fault tolerant electrostatic discharge network of claim 1, wherein at least one unit cell of said first array is different from at least one other unit cell of said first array.
5. The fault tolerant electrostatic discharge network of claim 1, wherein each unit cell of said second array is different.
6. The fault tolerant electrostatic discharge network of claim 1, wherein each unit cell of said second array of unit cells is the same.
7. The fault tolerant electrostatic discharge network of claim 1, wherein at least one unit cell of said second array is different from at least one other unit cell of said second array.
8. The fault tolerant electrostatic discharge network of claim 1, wherein every unit cell of said first array and every unit cell of said second array of unit cells are different.
9. The fault tolerant electrostatic discharge network of claim 1, wherein every unit cell of said first array and every unit cell of said second array is the same unit.
10. The fault tolerant electrostatic discharge network of claim 1, wherein at least one unit cell of said first array or of said second array is different from at least one other unit cell of said first array or said second array.
11. The fault tolerant electrostatic discharge network of claim 1, wherein the said unit cells of said first array and said unit cells of said second array are separated by a charge barrier.
12. The fault tolerant electrostatic discharge network of claim 11 wherein the charge barrier is a trench.
13. The fault tolerant electrostatic discharge network of claim 11 wherein the charge barrier is a set of N-wells.
14 The fault tolerant electrostatic discharge network of claim 11 wherein the charge barrier is a substrate contact.
15. A method, comprising:
- (a) defining a reliability requirement of an integrated circuit for a single event upset;
- (b) defining an electrostatic discharge requirement using a technology database file and electrostatic discharge data;
- (c) defining circuit topology having a redundant electrostatic discharge element in series with a voltage clamping element of the electrostatic discharge protection network to be applied between a first voltage and a second voltage in the integrated circuit;
- (d) defining a unit cell to comprise the redundant electrostatic discharge element using a single event upset simulation tool;
- (e) evaluating the probability of failure of the electrostatic discharge protection network; and
- (f) adding a design of said electrostatic discharge protection network to a design of said integrated circuit.
16. The method of claim 15, further including:
- between steps (e) and (f), repeating steps (b) through (e) until said design of said electrostatic discharge protection network meets said reliability of said integrated circuit.
Type: Application
Filed: Aug 13, 2007
Publication Date: Jul 3, 2008
Inventor: Steven H. Voldman (South Burlington, VT)
Application Number: 11/837,633
International Classification: G06F 17/50 (20060101); H02H 9/00 (20060101);