Bank Elektronik - Intelligent Controls GmbH

Some Rules for Wiring and Shielding

Something you don't learn at any school is to make correct ground connections. Where 1 V is a very low voltage, and 1 mA is a small current, there are no big problems about ground lines.

Electrochemical measurements deal with millivolts and microamperes (or much smaller currents), and with amplifiers having extreme high amplification (about 1 million or much more), so you are bound to follow some strict rules. Many people think that their noise problems arise from bad shielding. That may occur, but in most cases the noise comes from elsewhere.

Rule 1

Never form spider - webs of current lines! Instead of, connect them star - like: One single point shall be the ground - star - point.

Rule 2

(is a consequence of rule 1) Never form a loop by connecting ground wires!

Rule 3

Use a faradayic cage shielding your cell, if either - very low currents shall be measured - the electrolyte is badly conductive - the reference electrode system has high impedance Shielding of the cell is absolutely necessary when measuring very low currents or when using high - impedance reference electrodes. It is important to locate the cell away from power control panels and power cables. Electrical equipment producing powerful magnetic stray fields, such as magnetic stabilisers or regulating transformers should never be placed near the cell. Iron parts adjacent to the cell may have the effect of concentrating magnetic stray fields within the laboratory on to the cell. Screening the cell against electric fields is relatively simple; hum induced by magnetic fields can in practice be avoided only by correct location of the equipment. When using high - impedance reference electrodes (appreciably above 10 k) it is important to screen the cell. A bent sheet of aluminium, left open at the front, is usually quite sufficient. This sheet is then grounded to the grey plug on the cell cable together with the supporting stand.

Rule 4

Do not use the reference electrode cable shield for grounding the faradayic cage! (that is the reason why the reference cable shield is clipped). This shield lies on virtual ground only. A connection of this shield to the earth ground might produce rare results! The reference electrode is kept at virtual ground. Do not try to connect it to its shield or the ground plugs of the potentiostat, otherwise the potentiostat will not operate!

Rule 5

If a computer is interfaced to the potentiostat, please let it have a distance of at least 1 meter to the potentiostat, the cell and the other analogue instrumentation and cables. Especially its monitor (the screen) may cause RF radiation.

Rule 6

Avoid power cables in parallel to the cell cable or any other cable feeding control voltages into the potentiostat. (Power cables near the analogue output lines also induce some noise, but this noise acts on a rather low - ohmic line and - moreover - it will never pass the cell, so it may be regarded as less harmful)

Rule 7

Avoid power cables in vicinity of the cell.

Rule 8

Any metal parts in the neighbourhood of the cell which are not already grounded should be grounded to the potentiostat. Any metal stand in particular should be grounded through the grey earth plug of the cell cable, taking care to ensure a reliable electrical contact. (Here, the formation of ground loops is very easy: be careful!) Who informs you about which instrument is grounded internally and which is not? There are several hints. Instruments which have isolated "Lo" inputs and separate protection earth sockets usually allow a separation between protection earth (PE) and GND. On the other hand, many oscilloscopes have, for reasons of shock prevention, always GND internally connected to PE. The same applies to A/D - D/A - boards mounted in personal computers: In computers, also PE is connected to their internal ground. A/D - D/A converter boards have to use this ground, so it is connected to measurement ground (GND). However, the analogue inputs of these boards usually can be switched to "differential input" to avoid such grounding problems. More problems arise when the D/A outputs are connected to two different instruments. In that case it may be necessary to use the computer's ground as star point (alas, it is not the best one) and disconnect PE from GND at all other instruments. On all our instruments made by Bank Elektronik you will find a removable ground bridge. So you are free to decide where the ground - to earth connection shall be established.

Further reading for people who try to do their best in measuring small signals in noisy environments: Keithley Instruments, "Low Level Measurements"