One of the most common calls received concerning actual experiments is that the data collected appears "noisy" or that it contains current oscillations. This phenomenon is most often caused by: 1) reference electrode, 2) improper grounding of equipment, 3) highly capacitive cells (test specimens), and 4) environmental pick-up.
The most common source of noise or oscillation problems results from reference electrode issues. Princeton Applied Research potentiostats/galvanostats utilize a differential electrometer to measure the voltage, which requires input not only from the reference electrode, but also from the working electrode. One of the benefits of using a differential electrometer is that during periods of overload of the current, the potential control of the cell is maintained. A downside to a differential electrometer design is that when the reference impedance becomes large (on the order of 50,000 Ω or more), this increased impedance coupled with the stray capacitance of the reference circuit slows down the negative feedback (stability generating) side of the electrometer's operational amplifier. The positive feedback (stability destroying) side of the electrometer is not similarly slowed, causing oscillation of the potentiostat to result.
Most of the time, this problem can be addressed by examining your reference electrode. Clogged Vycor frits, trapped air bubbles in the electrode, use of a double junction bridge tube (aka luggin
capillary), and high solution resistance in the bridge tube are a few of the causes of increased impedance in the reference. One quick test to determine if it could be a reference problem is to disconnect the reference lead from the reference electrode and connect the reference lead to the counter electrode along with the counter lead. In this two-terminal connection, the counter is serving as both the auxiliary electrode and as a pseudo-reference. If the noise is eliminated by making these connection changes, the reference electrode should be examined.
The following are recommended procedures for examining/solving problems related to the reference electrode:
- Lower the reference electrode impedance. Make sure that your frit is in good condition and replace as needed. Avoid highly resistive solutions in the bridge tube or luggin.
- Increase the systems I/E stability by applying one of the I/E filters (if available on your system). These filters place a 10 nF capacitor in parallel with the current range measuring resistor to reduce the bandwidth of the I/E converter, thus minimizing the phase shift and potential to oscillate.
- If the problem results from increased solution resistance that cannot be avoided, place a 100 nF capacitor between the counter and reference electrodes. This has the effect of slowing down the potential controlling op amp (summing amplifier) and does not allow the potential control to outrace the feedback response.
- If the experiment is a corrosion type experiment, try placing a 0.1 - 1 µF capacitor between the working and ground leads. This will allow the high frequency ac signal causing the problem to bypass the current measuring resistors. This will tend to slow down the potentiostats current measurement and would eliminate oscillations. Note: This solution is no good for impedance measurements.
- For impedance measurements, use a platinum wire in parallel with the reference electrode that is normally used. One end of the platinum wire should terminate close to the tip of the reference electrode or bridge tube, which ever is closest to the specimen. The other end of the wire is connected to the reference electrode lead, with a 0.1 - 1 µF capacitor in series with the wire. This allows the high frequency component of the signal to bypass the reference (if a 1 µF capacitor is used, anything over 2 KHz will be shunted) while the DC component will be passed through the reference.
These are a few of the suggestions that have been offered to improve the response of a differential electrometer in a cell environment which produces a high reference impedance.
When interconnecting instrumentation with other instruments and computers, it is vital that all the chassis be at the same ground plane. Believe it or not, most ground connections in outlets are not at the same potential due to differing lengths in the wiring within the walls and several hundred millivolts potential difference can exist between outlets on opposite sides of a room. If two instruments are connected together with a signal carrying cable, with each instrument plugged into a different wall outlet, a current may flow between the two instruments proportional to the potential difference and the reistance of the cable shield. This ac voltage and current developed across the shield can eventually be added to the signal you're trying to measure. How can this be avoided?
All power cords from all instruments in an experiment should be connected to the same wall outlet. This includes the potentiostat, computer, other computer attachments (Note: printers have been identified as contributing to noise problems because they are often overlooked and plugged into an outlet across the room), rotating electrodes, etc. If necessary, use an industrial type multi-outlet strip, plug all instruments into it and in turn plug it into the wall outlet. The power cords should be routed together, away from all signal bearing cables and instruments not in the system. Another approach is to reduce the amount of current flow through signal circuit braids by providing additional paths. This is often called the "strap-them-together approach". The easiest way to do this is to use a large sized (#10 or larger) copper wire, making direct connections between all the chassis in the experiment, following the shortest possible path.
Highly Capacitive Cells
All potentiostats can become unstable when connected to capactive cells. The faster the response of the potentiostat, the more pronounced the problem. For instance, a Princeton Applied Research Model 283 potentiostat/galvanostat will begin to oscillate when performing a routine corrosion experiment, the ASTM G5 (test of 430 SS in sulfuric acid), once the specimen reaches the "passive" region where a stable oxide layer has formed. This oxide layer creates a capacitive situation, adding phase shift to the potentiostat's already phase-shifted feedback signal not to mention acting as an antenna for noise. For a system with the bandwidth of a 283, this can cause the system to go into oscillations in this region of the curve.
Fortunately, the easiest fix for this particular event is to apply the I/E filters (see the second bullet in the reference electrode section above), reducing the bandwidth of the I/E converter. Also, placing a capacitor (100 nF) between the reference and counter electrode leads (see the third bullet in the reference electrode section above) can have the effect of slowing down the potential controlling op amp (summing amplifier) and does not allow the potential control to outrace the feedback response.
Power lines and cords, computer monitors, stirring motors, and fluorescent lighting can all be sources of noise, particularly when operating at low currents. The most practical way to shield your experiment is by construction of a Faraday Shield from a fine wire mesh, sufficiently large to accommodate the experiment cell. Seams should be soldered and cables from the experiment should enter and exit at one point through a circular hole no larger than necessary. The shield should be grounded to the common system ground, that is, the black ground lead on your cell cable or connection at the potentiostat. As an alternative, a grounded metal fume hood can provide good shielding and should be tried before constructing any special shielding.
One problem in particular, as of late, is the new mandatory energy efficient lighting causing excessive current noise in potentiostat/galvanostat systems. The new standard is the T-8 lamps which use less energy, are quieter, do not contain PCBs, provide truer color lighting, and operate on an electronic ballast. The electronic ballast operates at a much higher frequency, 20-25 kHz, than the electromagnetic design (which operated at 60 Hz) of the T-12 ballasts. This higher frequency results in less electronic energy being needed to provide the same level of illumination as the T-12s. Unfortunately, it is the electronic ballast that is probably causing the problem with noise pick-up in our systems. The easiest way to determine if your noise problem is from your lights is to simply turn them off while performing an experiment to see if the noise is reduced or eliminated.
In addition to eliminating the source, you can apply I/E and low pass filters to reduce the effects of the environment on your potentiostat and its response.