Command Voltage Generation
The Optopatch provides the standard range of controls for generating the command voltage (or command current when in current clamp mode), plus an oscillator for generating sine, square and triangular waves over a wide frequency range. Although not normally provided on patch clamp amplifiers, the sine wave generation facility is particularly useful for cell membrane capacitance measurement, and it will be discussed further in that section of the manual.
The controls provided are as follows. First of all, the cell holding potential can be set anywhere between -200mV and +200mV on a ten-turn potentiometer. A switch is provided to select either positive or negative operation. Step or other potential waveforms, applied externally via the command x10 and/or command x100 input sockets on the rear panel, are then superimposed on the holding potential (i.e. all three signals are summed). These inputs are attenuated by a factor of 10 and 100 respectively, so for example a signal of +500mV at the commandx10 input will generate a step in the command potential of +50mV. If superimposed on a holding potential of say -70mV, then the overall command potential will change from -70mV to -20mV. However, it may be found more convenient to include the holding potential in the external command potential, so a centre off position is provided on the holding potential polarity switch. The centre position is labelled ext, since it actually switches to another external input (hold x10 in), which is also subjected to a tenfold attenuation. This input can either be used to accept an external holding potential, with modulating potentials continuing to be applied via the command x10 and/or command x100 inputs, or the two potentials could be combined together and applied on any one of these inputs. The possible advantage of the external hold input for this purpose is that the holding potential switch allows the user to switch between an internally generated holding potential, which may be useful for setting up, and an externally generated potential that provides the entire command potential. The hold x10 and command x10 inputs have a relatively low input impedance of 10Kohms. This means that it is not necessary to ground either input if it is not used, and if greater than tenfold attenuation is preferred, this can easily be achieved by an appropriate series resistor, e.g. 90Kohms of series resistance would increase the attenuation by a further factor of ten. The command x100 input already has a correspondingly higher input impedance of 100K ohms, but its sensitivity can be further reduced if required by adding a correspondingly higher series resistance.
The other possible component of the command potential is the output from the frequency generator, which is variable from 100Hz to 100KHz in three switched ranges. The variable frequency control is not itself calibrated, but the square wave frequency can be read precisely on the meter when the selector switch is in the osc position. In order to read the frequency, square waves need to be generated as well. These are converted into constant-length pulses, which are then filtered to give a voltage directly proportional to frequency. This works very well, but there is always a risk of interference to very low-level signals whenever large rapidly-changing signals such as these are present inside the same piece of equipment. In case this is ever a problem - although it may well not be - we have provided a jumper on the main circuit board, which allows sine waves to be generated without the accompanying square waves. The only drawback of this arrangement is that sine wave frequencies can no longer be read on the meter. In any case, square waves are never generated when the osc switch is off, which would be the usual condition when very small signals, i.e. single-channel currents, are being measured. A constant scaling factor of 1V=1KHz for reading the frequency applies under all conditions. The signal amplitude is set by another ten-turn potentiometer, and the signal is always symmetrical about a zero central potential. When the osc selector switch is in the sine position, low-distortion sine wave signals of up to 100mV rms (i.e. 242mV peak-to-peak) are added to the command potential, and in the square position, signals of up to + and - 100mV (i.e. 200mV peak-to-peak) are added. These outputs are also available at a tenfold higher level on the freqgen output on the rear panel..
The Optopatch also includes an integrator, which can be driven either internally from the sine or square wave signals, and/or externally from a BNC input on the rear panel. Circuits of this type produce symmetrical voltage ramps when driven by square waves, so they are sometimes also known as ramp generators. The frequency generator output is connected to the internal input of the integrator instead of to the command potential circuit when the sine/square switch is in the current, but the external BNC input is always active if selected as an input (otherwise this socket accepts the I=0 current input signal, as described elsewhere), and the integrator will add the two inputs together if both are present. The external input is effectively attenuated by a factor of ten, so a 1V signal here is equivalent to a 100mV signal from the internal oscillator. The output of the integrator is connected to the input of the patch clamp via a small capacitor, allowing AC currents to be passed. This is the same component that is used for compensation for the input and electrode capacitance, as described in the next section, and the integrator signal is simply added to the compensation signal, so the provision of this facility does not require any additional complexity in the headstage.
The current passed through the capacitor depends on the rate of change of the voltage across it, which means that the capacitor re-differentiates the integrator output to give a current waveform of the same shape as the voltage input to the integrator. There is of course a scaling factor to be calculated, which depends on the time constant of the integrator (i.e. the rate at which its output rises for a given voltage input) and the value of input capacitor. At the nominal full-scale input, i.e. 100mV for the internal frequency generator or 1V for the external input, the integrator output changes at 10V per millisecond, and this is coupled to the input via the electrode capacitance compensation capacitor in the headstage, to give a full-scale current of 10nA, which is conveniently intermediate between the patch clamp's two current ranges of 1nA and 100nA full scale, thereby making it useful in either. The relatively high integrator gain causes its output to saturate for square wave frequencies below 200Hz (where the peak-to-peak output is 25V), but it is still possible to pass half the full-scale current at 100Hz, and for sine waves the situation is slightly better, as the clip frequency for full-scale output is about 150Hz in that case. The integrator includes a DC servo circuit to hold the average output at zero, but it has no significant effect on its signal performance over the frequency range that is of interest here.
This facility has a variety of uses. In current clamp it provides an alternative method of passing (AC) currents, but its greatest use is in voltage clamping, where it allows the bandwidth and gain of the recording system to be verified (these parameters can be modified by other controls, which will be described later). Even in voltage clamp mode, the capacitor still passes a current into the input, and in order for the input voltage to remain constant, the voltage clamp's feedback loop must correspondingly reduce the feedback current, hence this facility is sometimes referred to as a "speed test", since complementary currents will be observed on the current output, and their time course will depend on the bandwidth of the current-measuring system. In contrast, the time course of a current step caused by a change in command potential also depends on the time required for the potential change to occur at the preparation, so this is not a proper measure of the recording bandwidth.
The advantage of using a capacitor to pass current is that essentially instantaneous current steps can be generated by applying voltage ramps, whereas a step voltage across a current-passing resistor is likely to cause an initially larger current transient, caused by the presence of some parallel capacitance across the resistor. Therefore the capacitor method provides a reliable speed reference for the recording system. For the same reason, it may also be a somewhat better way of producing fast current steps in current clamp mode, although in the Optopatch the two methods should be more nearly equivalent than in other designs.
When the patch electrode is in the bath, a potential of zero should ideally be recorded in I=0 mode and no current should flow in Vclamp mode, but in practice, electrode and bath potentials will result in a residual voltage. Although this could be taken into account by the applied potentials, it is clearly preferable to have some means of independent correction, which is provided by the junc control and switch. This is a centre-zero control that can provide up to + or - 200mV of additional voltage offset when the switch is in the on position. The switch also provides a "search" facility for use in voltage clamp mode. The search facility automatically adjusts the junction potential offset so as to drive the electrode current towards zero, but with a relatively long time constant, so that transient currents are not significantly attenuated. As its name suggests, this facility allows patches to be searched for at high current gain, while preventing the relatively large currents that would otherwise flow through the relatively low electrode resistance in response to only small potential differences (once a patch has been formed, the resistance of course increases dramatically, so this is no longer a problem). The applied junction potential offsets that are automatically generated in search mode can be read on the meter when the selector switch is in the junc position (when the switch is in the on position, the actual junction potential set by the control is read). When a patch has been found, the meter reading can be used as an additional guide to selection of the appropriate control setting before the switch is moved to the on position.
Although under most conditions the experimental bath will be connected directly to ground via a low-resistance bath electrode, there are also circumstances which may require external potentials to be imposed here, or in which the zero potential of the bath cannot be guaranteed. A bath potential input is therefore provided on the rear panel. If any signal that is present in the bath is also applied to the bath potential input, then the patch clamp command potentials will be generated with reference to this signal rather than ground, so that the operation of the patch clamp will not be affected by it. Once again, the input is of low impedance (10Kohms), so there is no need to ground it if it is not used. It is a unity-gain input, so signals should be applied at their original level. Because of the low input impedance, it is not suitable for direct connection to a reference electrode, but in any case it is greatly preferable to use such electrodes with a buffer amplifier sited close to them, rather than with a long unbuffered connection to the recording equipment (but provision for connecting and powering such an amplifier is made by a five-way circular connector on the rear panel). However, there is one useful exception. If the bath electrode is of low impedance (so as to be able to support a reasonable length of signal connection without coupling mains hum or interference into the bath, and to drive the bath effectively), then it can be connected directly to the bath potential input, in which case it is earthed by the 10Kohms input resistor (or an even lower resistor could be connected in parallel between this input and ground). External signals can then be imposed on the bath by applying them at this point, where they will drive both the bath electrode and the bath potential input.
The total command potential thus consists of some combination of an internal or external holding potential, an external command potential, an internal oscillator signal, an internal junction offset potential, and an external bath potential. (The command potential that is actually applied to the electrode may be further modified by the compensation facilities described later.) A command x10 output is available on the rear panel, i.e. at a tenfold higher level than the command potential applied to the cell. This output does not, however, contain the bath or junction potentials, since they are effectively not seen by the preparation either. The same signal combination is displayed on the meter at unity gain when the meter selector switch is in the vcommand position.
The bandwidth of the command potential signal pathway is relatively high, with an overall time constant of less than a microsecond. Under some recording conditions it can be useful to reduce the command potential bandwidth in order to minimise the transient errors that occur when the command potential suddenly changes. A command rounding switch is therefore provided for this purpose. When this switch is on, the command potential (again, apart from the bath and junction potential components) is filtered with a ten microsecond time constant, and it also affects the command potential output on the rear panel.
In current clamp mode, the command potential instead becomes a command current. However, junction and bath potentials are no longer summed into this signal, and they are instead added to the headstage output, which now represents the electrode potential (actually amplified tenfold), in order to act in the same way as in voltage clamp mode. A full-scale current of 1nA or 100nA requires a voltage input of 10V at the input to the current-passing circuit, and the external holding and command inputs supply this signal at unity gain, so the external x10 command sensitivities are 100pA/V and 10nA/V for the two current ranges. To retain the same relation between internal and external sensitivities as in the voltage clamp modes, these equate to 10pA/V and 1nA/V for the holding and frequency generator potentials.
Sometimes it may be desirable to switch from voltage clamp to current clamp (or vice versa) while maintaining the cell at some specified potential. This means changing from one value of command potential to another value of command current, at the same instant that the system is switched from voltage clamp to current clamp. Although that can in principle be done by a computerised control system, we decided to offer a simpler alternative method. We have provided a completely independent command input, which can pass a command current in I=0 mode ONLY, and which also has no effect in voltage clamp mode. Therefore, an appropriate voltage can be applied here, so that the required transition can be made by switching from voltage clamp to I=0 mode (which is thus no longer a true I=0 mode under these conditions, but an inaccurate name is a small price to pay for this useful facility). Its sensitivity is also x10, to match that of the external hold input in voltage and current clamp modes. The regular command input signals can then be changed appropriately if required, and the system then switched into current clamp in order to apply them, or the command currents can be added to the I=0 input signal instead (yes, a front panel control to set the I=0 current would be useful, but we'd run out of space...).
