Sample-and-Hold Circuit

So that the manner in which the above-recited features, advantages and objects of the invention, as well as others which will become apparent, are attained and can be understood in detail, more particular description of the invention briefly summarized above may be had by reference to the embodiments thereof which are illustrated in the appended drawings, which drawings form a part of this specification. It is to be noted, however, that the appended drawings illustrate only typical embodiments of the invention and are therefore not to be considered limiting of its scope, for the invention may admit to other equally effective embodiments.

In the drawings:

FIG. 1 is a block diagram of the crest detector and sample-and-hold circuit portions of the present invention.

FIG. 2 is a waveform diagram of the operation of the illustrated embodiment of the present invention.

Now referring to FIGS. 1 and 2, the crest detector and sample-and-hold circuits in conjunction with the desired operation, is illustrated. The top curve illustrated in FIG. 2 is the applied fluctuation signal to be detected. While the curve is at its initial low level, switch 10 is closed and the circuit is in what may be referred to as channel 1 or "track" mode operation. Switch 10 may be an FET switch or may, with the switches in the other channels to be explained hereafter, be part of a multi-input amplifier, such as Harris Semiconductor (Harris Semiconductor Division of Harris Intertype Corp.) type HA-2400 or the like. In any event, the logic signal for operating the switch, that is, supplying a "track" condition signal, is derived from a peak sensor circuit (not shown), which senses the presence of a positive or negative slope condition in excess of a predetermined value.

In this track mode, the input to high gain operational amplifier 12 is positive (or negative) at a low level. This creates an output from amplifier 12 which is very much more positive (or negative) than the input to amplifier 12 and which overcomes the forward bias of diode 14 or 16 via resistor 18. Diodes 14 and 16 are low leakage silicon diodes. The input from diodes 14 and 16 is applied to an FET input operational amplifier 20, connected as a Miller effect integrator with capacitor 22 providing 100 percent of direct feedback to the amplifier. Hence, any change in the output of amplifier 20 is due to the current input through diodes 14 and 16. The output of amplifier 20 is of opposite polarity than the input applied to amplifier 12, this output being supplied back to the input of amplifier 12 through resistor 24. This supplied back signal opposes the input, and keeps the "error" signal from amplifier 12 at a low level, within the frequency spectrum of interest. Amplifier 23 may be used to invert the output of amplifier 20, if desired.

When the peak sensor circuit detects a positive slope, a positive slope logic signal is supplied to the circuit to actuate switch 26 (and to open switch 10), or to initiate so-called channel 3 or "peak hold" operation. In channel 3 operation, this circuit continually senses the increasing level of the positivegoing signal, but holds the value of the maximum signal should the applied signal ever decrease. Now the output of amplifier 12 is supplied through resistor 28 to the negative input of amplifier 31. Diode 30, a germanium diode, operates as a negative voltage clamp for amplifier 31 in a manner to be explained. It is characteristic of a germanium diode to have a forward voltage drop that is about one-third of the voltage drop of a silicon diode. So, therefore, when diode 30 begins to conduct it begins to supply 100 percent degenerative feedback to error signal control amplifier 31, and hence, zero gain results.

Assume that the signal input to amplifier 12 is positive. The positive-going voltage applied to the amplifier results in a very much greater positive output therefrom, as with the channel 1 track operation. This produces a slight negative voltage at the output of amplifier 31 (which is clamped by diode 30, which is the input to silicon diodes 32 and 34. However, the voltage will not be sufficient to overcome the forward bias of these diodes and therefore the operation of amplifiers 12 and 20 and their associated components is the same as previously described. Namely, this operation is that the output from amplifier 20 will be determined by the input to amplifier 12.

At the crest of the peak, the positive input signal (represented by the positive curve in the top waveform of FIG. 4) becomes static and then begins to go less positive. Note that the logic signals applied to the circuit still keep the circuit operation in channel 3 operating condition in that the negative slope trigger is not yet produced. When the net input to amplifier 12 through resistors 24 and 25 goes negative, the output tries to go very much more negative because of the extremely high gain of amplifier 12. Now, extremely high gain operational amplifier 31 is permitted to rapidly go positive since germanium diode 30 is then back biased, providing no clamping feedback. Note that both amplifiers 12 and 31 are contributing to this and therefore the output is produced so fast that the time that is required for this action to take place is virtually undetectable.

When the output of amplifier 31 goes positive, diode 34 conducts and supplies via resistor 36 an input to amplifier 12 to oppose the net negative input to amplifier 12 through resistors 24 and 25. The output of amplifier 12 is reduced to virtually zero. That is, the bucking signal is so complete that even though the input to amplifier 12 through resistor 25 becomes significantly less positive than the peak value, the output of 31 increases at virtually the same pace to keep the voltage applied to resistor 18 at a near zero level. This may be observed by viewing the "error signal" waveform following crest detection. When the error signal is zero, neither diode 14 nor 16 is forward biased and the output from amplifier 20 "holds" at a constant level. It should be further observed that, as no active switching is done at the summing point input to amplifier 20, no error is introduced by switching transient energy.

Connected to the output of amplifier 31 is a clamped high-speed operational amplifier 44 connected as a Schmitt trigger and operating as the crest detector. The almost instantaneous output from amplifier 31 produces the crest logic signal for operation of the control voltage channel.

When the peak sensor circuit detects a negative slope, then so-called absolute "hold" or channel 4 operation is initiated. At this time, all switches around amplifier 31 are opened and amplifier 31 is permitted to follow absolutely the input applied thereto, up or down. It is during this period that the analog-to-digital converter is converting the held crest value to a digital format and performing other logic functions, as is hereafter explained. In any event, channel 4 operation may be thought of as overriding all other modes.

The input of amplifier 20 is a current summing point 46. That is, the point is not tied to ground, but it is adjusted to near zero by zero adjust variable resistor 48. The gain of the operational amplifier is extremely high, on the order of 100,000, and therefore the voltage at the input is extremely low, on the order of 10 microvolts. This type of circuit is well known and may be referred to as a Miller effect integrator. Characteristic of such an integrator is that the voltage run up of the output is linear with a linear increase in applied current. Also it is characteristic that the device is not sensitive to a change in voltage at the input, only to the applied current. Therefore, when the error voltage applied to diodes 14 and 16 is zero, there is not enough bias voltage to cause these diodes to conduct. Hence, the input to amplifier 20 is effectively disconnectd without any switching having taken place. The output voltage remains constant (or nearly so), because of the near perfect capacitor action of capacitor 22 and the extremely low input current of operational amplifier 20. The leak-off or drop rate may readily be made less than 100 microvolts per second, which for purpose of this circuit operation is so long as to be non-significant.

When a related ADC (not shown) is through, it produces a signal that returns the circuit shown in FIG. 1 to the initial "track" conditions, or channel 1 operation. It should be noted that the output of amplifier 20 has been kept from tracking the input until channel 1 conditions return. When these conditions return, the operation of amplifier 31 is immediately terminated and the error signal rapidly rises. This produces a rapid catch up of the output of amplifier 20 with the input to amplifier 12 in the manner previously explained for the track or channel 1 operating conditions.

The control voltage signal is stored for comparison to a digital value in a sample-and-hold circuit operation similar to that which has been described except that there is no crest detection and the hold or channel 4 operation is initiated by the crest logic signal from amplifier 44, as described above. That is, amplifiers similar to amplifiers 12 and 20 as described above track the control voltage until a crest signal puts the sample-and-hold circuit into channel 4 condition. After the control voltage is converted to a digital value in either a separate ADC or in a properly time shared ADC, as in the preferred embodiment, for handling both the fluctuating signal and the control voltage signal, the operation returns to track or channel 1 operating condition.

Note that there is a channel 2 operation for the circuit shown in FIG. 1 which has not been described. Such operation would include the closing of switch 40 connected to germanium diode 42, diode 42 being connected to produce a clamping action on amplifier 31 on the application of a negative error signal via resistor 28. Operation would be identical to channel 3 operation, except that minimums or valleys would be sensed rather than maximums or crests. A valley comparator 38, similar to crest comparator 44, would produce a logic-type signal at the instant of minimum occurrence.

While only particular embodiments of the invention have been shown, it will be understood that the invention is not limited thereto, since many modifications may be made and will become apparent to those skilled in the art.