Circuit for High Impedance Broad Band Probe

The present invention relates generally to the field of testing devices, more particularly, devices designed to test micro circuits or the like where the capacitance of the testing instrument is relevant since the instrument itself would significantly affect the test results.

With the development of large-scale integrated circuits (LSI), there has been a serious problem in the testing of these circuits. Because the components are so small and so close together, the introduction of any test equipment into the circuit would so strongly interfere with the characteristics of the circuit, that the test could not be effectively performed. Specifically, a mechanical probe which would be applied to at a point within an integrated circuit might add so much capacitance or "load up" the circuits so substantially, that the measurement would be meaningless. Furthermore, since the circuits may operate up to 400 MHz, the problem of creating a high impedance probe becomes even more accute at these higher frequencies.

The only prior art method known to be effective in making LSI measurements involves a stroboscopic effect and requires extremely expensive equipment. It uses an electron beam to indirectly measure currents within the circuit.

The present invention overcomes the problem of probe interference without the significant cost and complexity of existing methods.

The present invention relates to a high impedance broad band probe and circuit therefore, which is used in making electrical measurements. The probe is characterized particularly by its low capacitance and, therefore, its lack of disturbance of the circuit upon which a test is being performed. Specifically, the circuit includes an active circuit element having a substrate and layers corresponding to emitter base and collector, and an inherent base-to-substrate capacitance and base-to-emitter capacitance. The base-to-emitter capacitance is reduced by adjusting the emitter gain of the active element to be close to but not less than one and a positive number.

According to another aspect of the invention, the base-to-collector capacitance is reduced by maintaining the collector and emitter voltage increments generally equal by providing a DC voltage source with low AC impedance between the collector and emitter.

According to further aspects of the invention, the base-to-substrate capacitance is reduced by supporting the substrate atop a second substrate layer having a metalized layer thereon so that the original substrate layer is elevated above ground.

According to further aspects of the invention, a positive feedback loop between input and output provides an AC current flow toward the input thereby creating a negative capacitance and cancelling the input capacitance.

Various advantages and features of novelty which characterize the invention are pointed out with particularity in the claims annexed hereto and forming a part hereof. However, for a better understanding of its use, reference should be had to the drawings which form a further part hereof, and to the accompanying descriptive matter in which there is illustrated and described preferred embodiments of the invention.

In the drawings, specific embodiments of the invention are shown in good detail and explained in the detailed description of the drawings as follows:

FIG. 1 is a circuit schematic of a model capacitance circuit showing the effective location of input capacitances;

FIG. 2 is a circuit schematic showing a circuit for reducing base-to-collector capacitance;

FIG. 3 is a circuit schematic like FIG. 2 showing an alternate arrangement;

FIG. 4 is a schematic cross-sectional view of a solid state device according to the present invention;

FIG. 5 is a circuit schematic employing the solid state device in FIG. 4; and

FIG. 6 is a circuit schematic showing a negative impedance circuit.

The present invention is a device which allows direct measurement of a circuit at very high frequencies without adversely affecting the test results. Although the invention is called a "Zero Capacitance Probe", it could be characterized as a high impedance buffer circuit which places a high impedance between the test equipment (such as an oscilloscope) and the integrated circuit. With the buffer circuit in place, the oscilloscope does not influence the integrated circuit.

The problem with conventional circuitry is that capacitances result from the physical presence of the circuitry itself. We can model in where the capacitance in a circuit might exist by drawing representations of physical capacitors in those places. Looking at FIG. 1 of the drawings attached hereto, the input to a conventional transistor amplifier will have three main sources of capacitance. C.sub.BS (i.e., the base to substrate capacitance of the transistor), C.sub.BC (i.e., the base-to-collector capacitance) and C.sub.BE (i.e., the capacitance from base to emitter).

Capacitance creates a problem because it is a factor in the calculation of impedance (Z=1/(WC) where C is the capacitance and W is the radian frequency) so an increase in the capacitance or the frequency will result in a decrease in the impedance, which tends to load the circuit.

In cases involving an active element such as a transistor, the Miller effect must be considered. The Miller effect results in the capacitance from base to collector to be increased by a factor of one minus the gain of the transistor. Thus, the input capacitance of the circuit shown in FIG. 1 would be effectively:

As can be seen from the above equation, to reduce the input capacitance, it is necessary to reduce each of its three sources.

The present invention proposes a separate solution to reduce each capacitance.

The word "gain" is meant to mean a ratio of an output measurement over an input measurement. The output point can be either emitter or collector or equivalent.

Reducing the capacitance from the base emitter is dependent upon the gain of the transistor. If gain (K) is less than but very close to 1, then (1-K.sub.1) will be an extremely small number thereby reducing C.sub.BE.

The circuit in FIG. 1 is an emitter follower which has a gain K as measured at the emitter as being approximately less than or equal to 1, but a positive number.

The base-to-collector capacitance problem is not as easily resolved. Again, due to the Miller effect, the capacitance C.sub.BC =(1-K.sub.2).multidot.(C.sub.BC). In this case, K.sub.2 is the gain measured at the collector. Typically, the gain at the collector is a negative number less than minus 1, which would result in a very substantial increase in capacitance by the Miller effect. The present invention solves the problem by tying the gain K.sub.1 very close to K.sub.2.

Turning to FIG. 2, resistor R1 is replaced with a current source I.sub.K2 to provide a fixed current through the collector of the transistor. In order to prevent the gain K.sub.2 from being substantially different from K.sub.1, we can tie the collector and emitter together by placing a battery across the terminals. A battery or zener diode shifts the voltage and provides a high resistance for purposes of direct current but a very low impedance to AC. Voltage at the point marked K.sub.1 is thus (V.sub.in .multidot.K.sub.1). The voltage at K.sub.2 then will be:

and the AC voltage at K.sub.2 will be:

Instead of using a battery, a zener diode provides the same conditions. Thus, the voltage across the capacitor C.sub.BC is the difference of V.sub.in -(V.sub.in .multidot.K.sub.1) (ignoring the battery or diode which is insignificant). Thus, the voltage across C.sub.BE =(1-K.sub.1).multidot.V.sub.in. Assuming that the incremental (or AC) currents I.sub.K2 and I.sub.K1 are approximately equal, the capacitance C.sub.BC effectively becomes equal to:

One could by the way also speak of incremental voltage instead of current, which translates to incremental current by determining the voltage across the capacitance divided by the product of the capacitance and frequency.

As mentioned above, K.sub.1 is already a number less than but very close to 1, (say, 0.999). Thus, C.sub.BC will also be reduced substantially.

In reality, currents I.sub.K1 and I.sub.K2 as shown in FIG. 2 are not equal. Thus, it is necessary to provide a means for reconciliation of this current flow.

FIG. 3 shows such a circuit. I.sub.K2 here is meant to be substantially larger than I.sub.K1. By providing an additional transistor TR2, a shunt path is provided as a current I.sub.Kd (the differential current) flows through zener diode CR1, thus balancing the equation. This solves the problem of the unequal current flows or potential inequality and makes the solution to C.sub.BC practical.

The solution to the base to substrate capacitance (C.sub.BS) problem is different from the above solutions. Eliminating the emitter and collector capacitances involved circuitry whereas the base capacitance is removed by a mechanical solution. FIG. 4 shows a cross sectional view of a multilayered signal amplifying device in the form of an integrated circuit modified as per this invention. The structure shows a base layer atop a collector layer which itself is on a substrate. So far, this is a standard procedure. The present invention inserts a metalized layer under the first substrate and sets that atop a second insulating substrate which is an insulator (see FIG. 4). By elevating the first substrate above ground, C.sub.BS is effectively reduced by the formula C.sub.BS (1-(K3.multidot.K2), where K.sub.2 and K.sub.3 are the gains of the transistors TR.sub.1 and TR.sub.2 which are operated as voltage followers. Their product will be slightly less than 1, perhaps as high as 0.99, yielding a deficiency factor (1-K.sub.1 .multidot.K.sub.2) of 0.01, thereby reducing C.sub.BS by a factor of 100. The circuit in this configuration is represented in FIG. 5.

As an alternate or perhaps an additional solution to the problem, it is possible to provide a so-called "negative impedance" Z.sub.f which has a negative capacitance C.sub.f (i.e. a capacitance with a negative or reverse current flow) which will effectively cancel other capacitances in the circuit. The circuit shown in FIG. 6 provides such a solution. In this case, an additional operational amplifier A1 is provided having a gain of K.sub.4.

Resistors R.sub.1 and R.sub.2 are provided at the output of the amplifier to permit adjustment of the gain K.sub.4 to a value between 1 and 2. Looking at the current through C.sub.f, I.sub.4, it will be equal to:

and this is achieved by adjustment of the potentiometer R.sub.1. If K.sub.z is greater than 1, then I.sub.4 (the current through C.sub.f) will be flowing toward the input (as shown in FIG. 6) making the impedance less than zero, i.e., negative impedance. This will provide means to cancel completely input currents into C.sub.BS, C.sub.BC and C.sub.BE.

Numerous characteristics and advantages of the invention have been set forth in the foregoing description together with the details and structure and function of the invention, and the novel features thereof are pointed out in the appended claims. The disclosure, however, is illustrative only, and changes may be made in detail, such as substitution of active circuit elements with existing or other devices developed in the future which are employed within the principle of the invention to the full extent indicated by the broad general meaning of the terms expressed in the appended claims.