A METHOD OF PRECHARGING A SWITCHABLE FILTER

Abstract

A switching or switchable filter configured to switch in and out a filter with a minimal transient, the filter comprising a capacitor charged via a voltage follower. A resistor between the capacitor and the voltage follower may be short-circuited.

Claims

1.-9. (canceled)

10. A switchable capacitor assembly comprising: a first terminal, a second terminal, a first resistor, a first switch, a capacitor operatively connected between the first switch and the second terminal, a voltage follower operatively connected between the first terminal and the first resistor, the first resistor being operatively connected between the voltage follower and the first switch, and a second switch configured to short circuit the first resistor, where the first switch is configured to toggle between two states comprising: a first state where the switch operatively connects the capacitor to the first resistor and a second state where the switch operatively connects the capacitor to the first terminal.

11. The switchable capacitor assembly according to claim 10, wherein the voltage follower comprises an amplifier.

12. The switchable filter comprising: a signal path having a signal input and a signal output, a second resistor operatively provided between the signal input and the signal output and a switchable capacitor assembly according to claim 10, where the first terminal is connected to the signal path and where the second terminal is connected to a first predetermined potential.

13. An assembly for testing a device, the assembly comprises: one or more signal sources, one or more switchable filters according to claim 12, and a monitoring device configured to monitor the operation of the device, where: the device has one or more inputs and an output, each switchable filter is operationally connected between a separate source and a separate input, the output is operationally connected to the monitoring device.

14. A method of switching the switching capacitor assembly according to claim 10, the method comprising the steps of: a) connecting the second terminal to a first predetermined potential, b) providing a second predetermined voltage to the first terminal while the second switch does not short circuit the first resistor, c) operating, after step b), the second switch to short circuit the first resistor and d) operating the first switch from the first state to the second state.

15. The method according to claim 14, wherein the switchable capacitor assembly forms part of a switchable filter also comprising: a signal path having a signal input and a signal output, a second resistor operatively provided between the signal input and the signal output where the first terminal of the switchable capacitor assembly is connected to the signal path and where the second terminal is connected to a first predetermined potential, the method comprising the steps of: providing a second predetermined voltage to the signal path, and operating the first switch from the first state to the second state.

16. The method of operating according to claim 15, where one or more of the switchable filters form part of an assembly for testing a device, the assembly comprising: one or more signal sources, and a monitoring device configured to monitor the operation of the device, where: the device has one or more inputs and an output, each switchable filter is operationally connected between a separate source and a separate input, and the output is operationally connected to the monitoring device, the method comprising the steps of: at least one source feeding a varying voltage to the device via a switchable filter, the monitoring device monitoring an output of the device, maintaining the voltage output of the at least one source at a predetermined level, and operating the first switch of the switchable filter into its second state.

17. The method according to claim 15, further comprising the step of, prior to the step of operating the first switch, operating the second switch to short circuit the second resistor.

18. The method according to claim 15, further comprising the step of, prior to the providing of the second predetermined voltage to the first terminal or the signal path, varying a voltage on the first terminal.

Description

[0076] In the following, preferred embodiments of the invention will be described with reference to the drawing, wherein:

[0077] FIG. 1 illustrates a test regime of a device under test,

[0078] FIG. 2 illustrates the manner of generating a test signal and

[0079] FIG. 3 illustrates a switchable capacitor and filter according to the invention.

[0080] In FIG. 1, a device under test 10 is provided with a number, here 12, of voltages or signals in order to be tested, set up, tuned or calibrated. Often, the signals V1-V12 will be varied while an output or performance of the device 10 is monitored using a performance evaluation circuit 20. Once a suitable set of signals V1-V12 are determined for the device 10, these voltages may then be fixed and the device be used for another purpose, such as an intended purpose.

[0081] For each voltage or signal, the voltage or signal is provided by (see FIG. 2) a voltage or signal source 6 and fed through a switchable low-pass filter 8 to the device 10. The source 6 will often be a source of high-frequency noise caused by active component usually present in the source 6. The source 6 may also comprise a current sensor if desired. In usual operation of the device 10, this high-frequency noise is desired filtered so as to not affect the operation of the device.

[0082] Clearly, any swift variation, as described above, of a voltage will be affected by the low pass filter when provided in the signal path. Thus, it is desired that the filter can be coupled-in but also coupled-out. Then, the tuning may be performed with the filter coupled out, so that the voltage may be varied more swiftly.

[0083] Once the voltage is set as desired, the low pass filter is coupled in. On the other hand, the coupling-in or switching-in of the filter should be performed without excessive transients, as transients may destroy the tuning or state of the device 10.

[0084] In FIG. 3, a preferred circuit or filter 8 is illustrated in which a signal path is seen between Vin and Vout. A low pass filter is desired switched into this signal path without detrimental transients. The actual low pass filter 8 is formed by R1, provided in the signal path, and C, which may be provided between the signal path and ground.

[0085] A switch S1 is provided which is capable of connecting the capacitor to the signal path and thus couple-in the low pass filter in the signal path. The low pass filter is coupled-out when the switch S1 is in the position seen in FIG. 3.

[0086] In FIG. 3, a voltage follower, here in the shape of an operational amplifier A, is connected to the signal path. The operation of the voltage follower is to follow the voltage on the signal path.

[0087] The amplifier A preferably is a precision amplifier with a low bias current, a high amplification and a low offset.

[0088] The output of the voltage follower is provided to the capacitor C in order to charge the capacitor to the same voltage as seen in the signal path, here Vout. The operation of the voltage follower is to charge the capacitor C to the voltage Vout before flipping the switch S1.

[0089] The voltage follower thus will charge and discharge the capacitor as the voltage Vout and thus Va, varies.

[0090] A resistor, R2, is provided between the amplifier output and the capacitor C so as to allow a difference between the voltage, Vc, over the capacitor and the output voltage, Va, of the amplifier.

[0091] This allows the voltage Va to follow the voltage Vout even though the capacitor will charge and discharge with a time constant. Thus, when the voltage of the signal path varies swiftly, the voltage follower will perform its operation and follow that voltage, resulting in a corresponding charging and de-charging of the capacitor, but without obtaining a voltage difference between Vout and Va sufficient to exceed the threshold voltage of the parasitic diodes D1 and D2 of the amplifier. These diodes are indicated in hatched lines, as they exist within the amplifier.

[0092] When the voltage on the signal path is fixed or constant, the voltage Va of the voltage follower will match Vout and the voltage Vc will move toward Va and thus Vout. The time required for this will depend on the component values. It is desired that this takes no more than 5 seconds.

[0093] It is noted that the capacitor C will have a parasitic leak resistance, R1, indicated in hatched lines. Thus, as R1 and R2 will form a voltage divider, Vc will not completely reach Va, whereby a transient of some size will be seen when flipping the switch S1. The size of this transient and thus of the voltage Vc will depend on the relative sizes of R2 and R1.

[0094] If C1 is 1 ?F and R2 is 10 k?, the RC time constant of the charging is 10 ms so that Vc would be within 0.5V of Vout within 30 ms. However, with an R1 of 300 M? typical for a high-quality 1 ?F capacitor, the DC difference between Vout and Vc would be 333 ?V when Vin is 10V. Thus, a transient of >333 ?V would be seen when flipping S1. This high a transient would be unacceptable for many applications.

[0095] This transient may be reduced further, if the switch S2, which has a much lower resistance than R2, is flipped before the switch S1 is flipped. In that manner, Vc will reach Va and thus Vout within ?V. Thus, S2 should be flipped before S1 so that Vc has time to reach Va. It is noted that when the voltage different, with S2 open, between Vc and Va/Vout is lower than the threshold voltage of the parasitic diodes D1/D2, the closing of S2 will not exceed that threshold voltage.

[0096] If A delivers 10 mA independently of the voltage difference on its inputs, the capacitor will be charged with a dvdt of 10 mA/10 ?F=10 kV/s. in the case of the voltage difference being 0.5V, the capacitor would be fully charged in 50 ?s. Then, the switch S1 may be flipped with hardly any transients.

[0097] The time required for the final charging of the capacitor clearly will depend on the component values and the initial voltage difference between Vc and Vout, but usually a time delay of 1 ms would suffice for Vc to reach Vout after flipping of S2.

[0098] The flipping of the switches S2 and S1 may even be controlled by the same signal if the switches are selected as different in type. A FET transistor-based switch is swifter in operation than a latched relay, so that when feeding the same flipping signal to S2 as a FET transistor, S2 will close in time before S1 being a latched relay would have time to react and switch. In that situation, a single signal may be used for controlling both switches when the desired voltage of Vin has been reached and has allowed to persist for a sufficient period of time for the voltage follower to charge the capacitor to the value defined by the relation between R2 and R1.

[0099] It is noteworthy that the main insight of this invention is that the typical parasitic leak resistance R1 and the typical parasitic diodes D1/D2 are making it necessary to add the switch S2 to the circuit and use the above mentioning switching timing, to achieve the lowest possible transient when switching S1. Without these parasitics, which unfortunately are present in the components used for most applications, the extra switch S2 would not be needed, and S1 could be switched without transients.