Rapid method of digital automatic gain control

Claim 2

A rapid method of digital automatic gain control in a signal channel having a digitally programmable variable attenuator that adjusts gain and an analog-to-digital converter to sample the channel output and convert signal samples to digital values, comprising the steps of: (a) providing to an automatic gain control computer the digitized output signal X whose exponential average with time constant T is to be kept at a desired signal level L by adjusting the gain G presented to said variable attenuator; (b) determining and storing the integer constants N, NT, L, -NT and -L where L and T are selected and the others computed, and N is the smallest integer such that NT is greater than or equal to a given maximum value of gain; (c) setting arbitrary variables A and B to initial values, where A is an accumulated difference over a plurality of sampling periods between the output signal X and desired output signal level L and B changes by the current gain G when he magnitude of A becomes larger than L; (d) in each sampling period, adding to A the difference between the current digitized output signal X and the desired average level L, and (i) when A is between L and -L, taking no further action to adjust current gain G; (ii) when the output signal X has been large for sufficient time above the desired average signal level, and A becomes larger than L, decrementing A by L and incrementing B and G, and decreasing current gain G only if B is now greater than NT; (iii) when the output signal X has been small for sufficient time, below the desired signal level L, and A becomes less than -L, incrementing A by L and decrementing B by G, and increasing current gain G only if B is now less than -NT; and (e) retaining the values of A, B and G for use in the next sampling period and passing the current gain G to said variable attenuator.

Claim 3

The method of claim 1 further comprising, in step (dii), comparing A to L and repeating as needed until A is less than L.

Claim 4

The method of claim 1 further comprising, in step (diii), comparing A to -L and repeating as needed until A is greater than -L.

Claim 5

The method of claim 1 further comprising, in steps (dii) and (diii), adjusting gain G by N.

Claim 6

A rapid method of digital automatic gain control to cyclically adjust the gain of a signal channel having an analog input signal x and digitized output signal X so as to maintain the exponential average of the output signal X with a time constant T at a desired average signal level L, comprising the steps of: (a) determining and storing in a computer the constants N, NT, L, NT and -L, where L and T are given and N is the smallest amount by which gain is changed and the smallest integer such that NT is greater than or equal to a given maximum digital gain G max ; (b) setting arbitrary variables A and B to initial values, where A is the accumulated difference over a plurality of sampling cycles between the output signal X and the desired average output level L, and B changes by current gain G when the magnitude of A becomes larger than L; (c) in each cycle, adding to A the difference between the current digitized output signal X and the desired average level L, and taking no action to change B and G when A is between L and -L; and (d) retaining the values of A, B and G for use in the next cycle and presenting the current gain G to digitally programmable variable gain means in said signal channel.

Claim 7

A rapid method of digital automatic gain control to cyclically adjust the gain of a signal channel having an analog input signal x and digitized output signal X so as to maintain the exponential average of the output signal X with a time constant T at a desired average signal level L, comprising the steps of: (a) determining and storing in a computer the constants N, NT, L, NT and -L, where L and T are given and N is the smallest amount by which gain is changed and the smallest integer such that NT is greater than or equal to a given maximum digital gain G max ; (b) setting arbitrary variables A and B to initial values, where A is the accumulated difference over a plurality of sampling cycles between the output signal X and the desired average output level L, and B changes by current gain G when the magnitude of A becomes larger than L; (c) in each cycle, adding to A the difference between the current digitized output signal X and the desired average level L, and taking no action to change B and G when A is between L and -L; and (d) retaining the values of A, B and G for use in the next cycle and presenting the current gain G to digitally programmable variable gain means in said signal channel.

Claim 8

The method of claim 5 further comprising, in step (c), determining that A is greater than L, decrementing A by L and incrementing B by G, determining that B is not greater than NT and A is not greater than L, and proceeding to step (d).

Claim 9

The method of claim 5 further comprising, in step (c), determining that A is greater than L, decrementing A by L and incrementing B by G, determining that B is now greater than NT, decrementing B by NT and decrementing G by N, and proceeding to step (d) after confirming that A is not greater than L.

Claim 10

The method of claim 7 further comprising, when A is still greater than L, repeating step (c) as needed until A is less than L.

Claim 11

The method of claim 5 further comprising, in step (c), determining that A is less than -L, incrementing A by L and decrementing B by G, determining that B is not less than -NT and that A is not less than -L, and proceeding to step (d).

Claim 12

The method of claim 5 further comprising, in step (c), determining that A is less than -L, incrementing A by L and decrementing B by G, determining that B is now less than -NT, incrementing B by NT and incrementing G by N, and proceeding to step (d) after confirming that A is not less than -L.

Claim 13

The method of claim 10 further comprising, when A is still less than -L, repeating step (c) as needed until A is not less than -L.

Claim 14

The method of claim 5 wherein said sampling cycles correspond to sampling periods of an analog-to-digital converter which presents said digitized output signals to said computer.

Claim 15

A rapid method of digital automatic gain control in a machine tool monitor having a vibration sensor to provide a signal representative of vibrations generated by interaction of a cutting tool and workpiece, an analog signal processing channel to condition the vibration signal to an analog output voltage in a restricted frequency range to an analog output voltage in a restricted frequency range and which has a digitally programmable variable attenuator that adjusts gain, and means to sample and convert the analog output to a digitized output signal X which is analyzed by a computer that calculates digital gain G, said method adjusting at every sampling period the digital gain G presented to said variable attenuator by said computer in order to maintain the exponential average of the digitized output signal X with a time constant T at a desired signal level L, comprising the steps of: (a) determining and storing in said computer the constants N, NT, L, -NT and -L, where L and T are given and equal to a given maximum gain G max ; (b) setting arbitrary variables A and B to initial values, where A is the accumulated difference over a plurality of sampling periods between the output signal X and the desired average output level L, and B changes by current gain G when the magnitude of A becomes larger than L; (c) in each sampling period, adding to A the difference between the current digitized output signal X and the desired average level L, and (i) when A is between L and -L, taking no further action to adjust current gain G; (ii) when A is greater than L, decrementing A by L and incrementing B by G, and decreasing current gain G only if B is now greater than NT; (iii) when A is less than -L, incrementing A by L, and decrementing B by G, and increasing current gain G only if B is now less than -NT; and (d) retaining the values of A, B and G for use in the next sampling period and presenting the current gain G to said variable attenuator.

Claim 16

A rapid method of digital automatic gain control in a machine tool monitor having a vibration sensor to provide a signal representative of vibrations generated by interaction of a cutting tool and workpiece, an analog signal processing channel to condition the vibration signal to an analog output voltage in a restricted frequency range to an analog output voltage in a restricted frequency range and which has a digitally programmable variable attenuator that adjusts gain, and means to sample and convert the analog output to a digitized output signal X which is analyzed by a computer that calculates digital gain G, said method adjusting at every sampling period the digital gain G presented to said variable attenuator by said computer in order to maintain the exponential average of the digitized output signal X with a time constant T at a desired signal level L, comprising the steps of: (a) determining and storing in said computer the constants N, NT, L, -NT and -L, where L and T are given and equal to a given maximum gain G max ; (b) setting arbitrary variables A and B to initial values, where A is the accumulated difference over a plurality of sampling periods between the output signal X and the desired average output level L, and B changes by current gain G when the magnitude of A becomes larger than L; (c) in each sampling period, adding to A the difference between the current digitized output signal X and the desired average level L, and (i) when A is between L and -L, taking no further action to adjust current gain G; (ii) when A is greater than L, decrementing A by L and incrementing B by G, and decreasing current gain G only if B is now greater than NT; (iii) when A is less than -L, incrementing A by L, and decrementing B by G, and increasing current gain G only if B is now less than -NT; and (d) retaining the values of A, B and G for use in the next sampling period and presenting the current gain G to said variable attenuator.

Claim 17

The method of claim 13 further comprising, in steps (cii) and (ciii), adjusting G by N.

Claim 18

The method of claim 13 further comprising, in step (cii), decrementing B by NT and decreasing current gain by decrementing G by N.

Claim 19

The method of claim 13 further comprising, in step (ciii), incrementing B by NT and increasing current gain by incrementing G by N.