Avalanche photodiode gain control comprising a bias circuit having a second avalanche photodiode

This application claims priority to U.S. Provisional Application 63/060,187 filed Aug. 3, 2020, titled “Avalanche Photodiode Gain Control,” which is hereby incorporated by reference in its entirety.

Photodiodes are semiconductor devices that generate an electrical current when Photodiodes are used as light-detection elements in a variety of applications. A photodiode includes a p-n junction formed in a semiconductor material. A reverse bias is applied to the p-n junction to widen the depletion layer, and an electric field is applied. Electron-hole pairs are generated in the depletion layer by the absorbed light. Under the attraction of the electric field, electrons move to the n-type semiconductor region, while holes move to the p-type semiconductor region, thereby causing a current to flow. Types of photodiodes include PIN photodiodes and avalanche photodiodes.

In one example, an avalanche photo-diode (APD) circuit includes a first APD, a reference voltage source, and a bias circuit. The first APD includes an anode. The bias circuit includes a bias output, a second APD, and a layer of metal that covers the second APD. The bias output is coupled to the anode of the first APD. The second APD includes a cathode and an anode. The cathode of the second APD is coupled to the bias output. The anode of the second APD is coupled to the reference voltage source.

In another example, an APD circuit includes a first APD and a bias circuit. The first APD is configured to detect light. The bias circuit is configured to control a gain of the first APD. The bias circuit includes a second APD, a reference voltage source, a bias voltage generation circuit, and a metal layer configured to shield the second APD from the light. The reference voltage source is configured to bias the second APD. The bias voltage generation circuit is configured to generate a bias voltage for biasing the first APD based on dark current output by the second APD.

In a further example, a light detection circuit includes an APD array and a bias circuit. The APD array is configured to detect light, and includes a first APD, and a transimpedance amplifier. The transimpedance amplifier is coupled to first APD. The bias circuit is coupled to the APD array, and is configured to control a gain of the first APD. The bias circuit includes a second APD, a reference voltage source, a current multiplier circuit, a bias voltage generation circuit, and a metal layer that is configured to shield the second APD from the light. The reference voltage source is configured to bias the second APD. The current multiplier circuit is coupled to the second APD, and is configured to provide an output current that is a predetermined multiple of the dark current output by the second APD. The bias voltage generation circuit is configured to generate a bias voltage for biasing the first APD based on the output current of the current multiplier circuit.

FIG. 1 A is a schematic diagram of a positive-intrinsic-negative (PIN) photodiode.

FIG. 1 B is a graph of the electric field distribution in the reverse biased PIN photodiode of FIG. 1 A .

FIG. 2 A is a schematic diagram of an avalanche photodiode (APD).

FIG. 2 B is a graph of the electric field distribution in the various layers of the reverse biased APD of FIG. 2 A .

FIG. 3 A is a graph of dark current in an APD.

FIG. 3 B is a graph of photocurrent in an APD.

FIG. 3 C is a graph of avalanche gain in an APD.

FIG. 4 is a block diagram of an example APD circuit that uses dark current to bias an APD.

FIG. 5 is a schematic diagram for an example bias voltage generation circuit suitable for use in the APD circuit of FIG. 4 .

FIG. 6 is a block diagram of an example second APD circuit that uses dark current to bias an APD.

FIG. 7 is a block diagram of an example third APD circuit that uses dark current to bias an APD.

In a photodiode, photocurrent (I p ) is proportional to the incident optical power (P in ). Responsivity is the ratio of photocurrent generated from incident light, to the incident light, and may be expressed as current over power.

R = I p P in

where:

I p is in units of amperes (A); and

P in is in units of watts (W).

The quantum efficiency of the photodiode may be defined as:

Responsivity may be expressed based on quantum efficiency as:

Positive-intrinsic-negative (PIN) photodiodes and avalanche photodiodes (APD) are commonly used to detect light. FIG. 1 A is a schematic diagram of a PIN photodiode. FIG. 1 B is a graph of the electric field distribution in the reverse biased PIN photodiode. The PIN photodiode includes an intrinsic (or lightly doped) region sandwiched between a p-type region and an n-type region. The intrinsic region increases depletion region width, where photo-generated carriers (electrons and holes) travel by drift, rather than by diffusion as in the p-type and n-type regions. The depletion region creates capacitance that is inversely proportional to the width of the depletion region. Because response time of the photodiode is a function of the capacitance, increasing depletion region width decreases response time. PIN photodiodes are reverse biased well below the breakdown voltage. Table 1 lists various characteristics of typical PIN photodiodes.

FIG. 2 A is a schematic diagram of an APD. FIG. 2 B is a graph of the electric field distribution in the various layers of the reverse biased APD. The APD is similar to the PIN photodiode, and includes an additional p-type region that operates as a multiplication layer. Secondary carriers are generated in the multiplication layer via impact ionization. The responsivity of the APD is expressed as:

where:

M is the multiplication factor provided by addition of the multiplication layer; and

R is responsivity of the photodetector without the multiplication layer.

The multiplication factor M may be expressed as:

M = 1 - k ⁢ A exp ⁡ [ - ( 1 - k ⁢ A ) ⁢ α e ⁢ d ] - k ⁢ A

where:

α h is a hole impact ionization constant;

α e is an electron impact ionization constant;

d is thickness of the multiplication layer, and

The APD is reverse biased at a higher voltage than the PIN photodiode to enable impact ionization (M>>1), With low reverse bias voltage; the APD behaves like a PIN photodiode (M=1). Table 2 lists characteristics of typical APDs.

APD gain is dependent on temperature and reverse bias voltage. Gain should be controlled to meet system performance metrics, such as sensitivity, signal-to-noise ratio, linearity, etc. in an array of APDs, the different APDs may receive different incident optical power, and control of the APD gain should be independent of incident light. The effects of ambient or stray light should be compensated.

FIGS. 3 A, 3 B, and 3 C are graphs of dark current, photocurrent, and avalanche gain in an APD respectively. Dark current is the small current flowing in an APD when no light is incident on the APD. FIG. 3 A shows that for a dark APD avalanche breakdown is dependent on temperature, with higher temperature requiring a higher reverse bias voltage. The slope in the breakdown region is steep. FIG. 3 B shows that photocurrent behavior of the APD is similar to the dark current behavior of the APD (e.g., higher temperature requiring higher reverse bias voltage). FIG. 3 C shows that for different temperatures the reverse bias voltage must be adjusted to maintain a desired gain. The APD bias control circuits described herein take advantage of this similarity in dark current and photocurrent behavior, and apply dark current to bias the APD and control the gain thereof.

FIG. 4 is a block diagram of an APD circuit 400 that uses dark current to bias an APD. The APD circuit 400 may be a light detection circuit, a light detection and ranging (LIDAR) receiver circuit, or other optical transduction circuit. The APD circuit 400 includes an APD array 402 and a bias circuit 404 . The APD array 402 may include one or more APDs. Some implementations of the APD array 402 may include hundreds or thousands of APDs. The APD array 402 includes an APD 406 and a transimpedance amplifier (TIA) 408 . The APD 406 and the TIA 408 are coupled to form pixel circuitry of the APD array 402 . The APD 406 , and other APDs of the APD array 402 , may be fabricated in silicon, germanium, silicon-germanium, or other semiconductor materials. The APD 406 is exposed to and detects incident light, and produces a photocurrent that is proportional to the incident light. The TIA 408 converts the photocurrent to voltage. An input 408 A of the TIA 408 is coupled to the cathode 406 C of the APD 406 . The anode 406 A of the APD 406 is coupled to the bias circuit 404 for receipt of a bias voltage that controls the gain of the APD 406 . In practice, the anode of each APD of the APD array 402 is coupled to the bias circuit 404 for receipt of the bias voltage.

The bias circuit 404 generates the bias voltage based on dark current, and is therefore insensitive to ambient light and differences in optical power across the APD array 402 . The bias circuit 404 includes a bias output 404 A, an APD 410 , a current multiplier circuit 412 , an amplifier 416 , an APD 418 , a bias voltage generation circuit 422 , and a reference voltage source 424 . The bias output 404 A is coupled to the anode 406 A of the APD 406 . The APD 410 and the APD 418 may be instances of the APD 406 . The APD 410 and the APD 418 may be located near the APD array 402 (on an integrated circuit) for process and temperature tracking. A layer of metal 414 (or other opaque material) covers the APD 410 , and a layer of metal 420 (or other opaque material) covers the APD 418 . For example, the metal 414 and the metal 420 may be same metal layer of an integrated circuit that isolates the APD 410 and APD 418 from light. Thus, the APD 410 and the APD 418 generate dark current and do not generate photocurrent.

An anode 410 A of the APD 410 is coupled to the reference voltage source 424 . The reference voltage source 424 provides a bias voltage that biases the APD 410 for non-avalanche operation. The bias voltage applied to the APD 410 may relatively low (e.g., 0 to −20 volts). The dark current (I D1 ) generated by the APD 410 is multiplied by a predetermined multiplier value in the current multiplier circuit 412 . The predetermined multiplier value is determined based on the desired gain of the APD 406 . For example, the predetermined multiplier value may be in a range of 10-500. A cathode 410 C of the APD 410 is coupled to an input 412 A of the current multiplier circuit 412 . The current multiplier circuit 412 may be implemented as a current mirror circuit to provide the desired multiplication of the dark current received from the current multiplier circuit 412 . The multiplied dark current (MIDI) produced by the current multiplier circuit 412 is provided to the amplifier 416 . An input 416 B of the amplifier 416 is coupled to the output 412 B of the current multiplier circuit 412 .

The anode 418 A of the APD 418 is coupled to the bias output 404 A. Thus, the same bias voltage is applied to the APD 418 and the APD 406 , to bias the APD 418 and the APD 406 for avalanche mode operation. The dark current ( 102 ) generated by the APD 418 is provided to the amplifier 416 . A cathode 418 C of the APD 418 is coupled to the input 416 A of the amplifier 416 .

The amplifier 416 compares the dark currents generated by the APD 410 and the APD 418 to produce a difference voltage (the difference of the two currents (MIDI and I D2 ). The difference voltage output by the amplifier 416 is provided to the bias voltage generation circuit 422 .

The bias voltage generation circuit 422 applies the difference voltage received from the amplifier 416 to generate the bias voltage applied to the APD 406 and the APD 418 . An input 422 A of the bias voltage generation circuit 422 is coupled to an output 416 C of the amplifier 416 . The bias voltage generated by the bias voltage generation circuit 422 biases the APD 418 to produce dark current 102 , where I D2 =MI D1 . If I D2 <MI D1 , the APD reverse bias is increased. If I D2 >MI D1 , the APD reverse bias is decreased. If I D2 =MI D1 , the APD reverse bias is not changed. Thus, the bias voltage applied to the APD 406 , and other APDs of the APD array 402 , is based on the dark current output by the APD 410 and the APD 418 as function of temperature and process.

FIG. 5 is a schematic diagram for an example bias voltage generation circuit 422 . The amplifier 416 is also shown in FIG. 5 for reference. The bias voltage generation circuit 422 includes a transistor 502 and a capacitor 504 . The transistor 502 may be a p-type field effect transistor (FET). A gate 502 G of the transistor 502 is coupled to the input 422 A of the bias voltage generation circuit 422 and the output 416 C of the amplifier 416 . A drain 502 D of the transistor 502 is coupled to a negative voltage source 506 . The negative voltage source 506 may provide voltage in a range of −20 to −250 volts for biasing the APD array 402 . A source 502 S of the transistor 502 is coupled to the output 422 B and to terminal 504 A of the capacitor 504 . A terminal 504 B of the capacitor 504 is coupled to ground. The difference voltage received at the gate 502 G of the transistor 502 controls the transistor 502 , and the bias voltage generated at the source 502 S of the transistor 502 .

FIG. 6 is a block diagram of a second APD circuit 600 that uses dark current to bias an APD. The APD circuit 600 may be a light detection circuit, a light detection and ranging (LIDAR) receiver circuit, or other optical transduction circuit. The APD circuit 600 includes the APD array 402 and a bias circuit 604 . The bias circuit 604 includes the bias output 404 A, the APD 410 , the current multiplier circuit 412 , the APD 418 , a transimpedance amplifier 602 , a unity gain buffer 606 (a buffer circuit), and a variable current source 608 . As explained with regard to the bias circuit 404 , the metal 414 and the metal 420 prevent light from reaching the APD 410 and the APD 418 .

The APD 410 is biased at a low reverse bias by the reference voltage source 424 . The dark current generated by the APD 410 is multiplied by M in the current multiplier circuit 412 . The multiplied dark current is provided to the variable current source 608 .

In the APD circuit 600 , the transimpedance amplifier 602 , the unity gain buffer 606 , the variable current source 608 , and the APD 418 are part of a bias voltage generation circuit. A control input 608 A of the variable current source 608 is coupled to the output 412 B of the current multiplier circuit 412 . The multiplied dark current received at the variable current source 608 controls the current output by the variable current source 608 . An input 608 C of the variable current source 608 is coupled to a negative voltage source 610 . Current MIDI output by the variable current source 608 is forced into the APD 418 to generate the bias voltage for the APD 406 . An output 608 B of the variable current source 608 is coupled to the anode 418 A of the APD 418 and the input 606 A of the unity gain buffer 606 . The cathode 418 C of the APD 418 is coupled to an input 602 A of the transimpedance amplifier 602 to provide the same cathode voltage as the APD 406 .

The unity gain buffer 606 buffers the voltage at the anode 418 A of the APD 418 . An output 606 B of the unity gain buffer 606 is coupled to the bias output 404 A to provide bias voltage to the APD 406 , and other APDs of the APD array 402 .

FIG. 7 is a block diagram of a third APD circuit 700 that uses dark current to bias an APD. The APD circuit 700 may be a light detection circuit, a light detection and ranging (LIDAR) receiver circuit, or other optical transduction circuit. The APD circuit 700 includes the APD array 402 and a bias circuit 704 . The bias circuit 704 includes the bias output 404 A, the APD 410 , the current multiplier circuit 412 , and a variable current source 708 . As explained with regard to the bias circuit 404 , the metal 414 prevent lights from reaching the APD 410 .

The APD 410 is biased at a low reverse bias by the reference voltage source 424 . The dark current generated by the APD 410 is multiplied by M in the current multiplier circuit 412 . The multiplied dark current is provided to the variable current source 708 .

In the APD circuit 700 , the variable current source 702 is part of a bias voltage generation circuit. A control input 708 A of the variable current source 708 is coupled to the output 4126 of the current multiplier circuit 412 . The multiplied dark current received at the variable current source 708 controls the current at output 708 B of the variable current source 708 . An input 708 C of the variable current source 708 is coupled to a negative voltage source 710 . Current MIDI output by the variable current source 708 is forced into the APD 406 to generate the desired bias voltage. An instance of the variable current source 708 is coupled to each APD of the APD array 402 with a control input coupled to the output 412 B of the current multiplier circuit 412 and an input coupled to the negative voltage source 710 .

In this description, the term “couple” may cover connections, communications, or signal paths that enable a functional relationship consistent with this description. For example, if device A generates a signal to control device B to perform an action: (a) in a first example, device A is coupled to device B by direct connection; or (b) in a second example, device A is coupled to device B through intervening component C if intervening component C does not alter the functional relationship between device A and device B, such that device B is controlled by device A via the control signal generated by device A.

Modifications are possible in the described embodiments, and other embodiments are possible, within the scope of the claims.