The electronics have evolved somewhat since the early days & have been kept as simple as possible.
I decided to use a 27.8kHz square wave, rather than the more common sine wave, to give a cleaner switch-over of voltage.
The original design used a long-tailed pair differential amplifier but it was found that a 741 op-amp was superior, particularly in common mode signal rejection.
The precision rectifier is from a design by Analog Innovations with a x 10 gain amplifier added. It gives a precisely linear response. There is a zero offset voltage of 50 mV, which could not be eliminated by using the offset null pin on the amplifier 741. The precision rectifier is required because the output voltage from the long tailed pair (or OPAMP) is too low for the use of a simple diode rectifier, i.e. it is often below the threshold voltage of a diode.
The AC bridge components were wired with shielded cable with equal cable lengths for the two cells & the two sides of the 1k pot. [Wire-wound 10-turn pot can be replaced with a simple inexpensive version].
POWER SUPPLY 12V. The main power supply is a stabilised 12V source but I have also used a Li ion battery pack recovered from a laptop power supply that gives power for 8 hours running of the apparatus. Power consumption is a max of 12 watts during warmup. A useful power supply is from Electrocomponents RS [413-667] that gives 1.3A, 12V & costs £8.94.
A drawback to the use of this particuar precision rectifier design is that a 9V dual rail power supply is required, supplied by 2 x PP9 batteries, the current draw is modest, so the batteries last for several weeks of use. Designs of rectifier are available that use a single rail, see link above. However the 741 OPAMP differential amplifier is now also powered from this source.
The temperature of the system is controlled to +/- 0.02 deg but can drift upwards slowly if the ambient temperature is close to 25 deg. The drift can be 0.5 - 1 deg over 8 hours but this has little effect on initial rate measurements. Only the addition of a cooling system can overcome this, as far as I am aware. Normally the drift is below 0.04 deg/hour.
The Thermostatic Control System
The Square Wave Generator
The Precision Rectifier
NOTE THAT IT IS FINE TO USE A SIMPLE 1-TURN POT IN PLACE OF THE 10-TURN SHOWN HERE
The 1k pot shown here is used with the equivalent conductivity [2mS/cm] of 0.1% NaCl. 100 ohms with 1% NaCl equiv., 10k with 0.01% NaCl, 100k with 0.001% NaCl & 1M with 0.0001% NaCl - see Calibration (sub-page from RESULTS) for further details
It would be better to use a 6.7k resistor & 1k pot for ease of adjustment of the output voltage.
There was previously an error in the circuit below. The feedback resistor (220k [67k can also beused]) at the final amplifier was connected to the non-inverting (positive +) input - it should be connected to the inverting (negative -) input as now shown.