A while ago I received an email from an electrochemist who lamented:
“We have some problems with the 173, which we still prefer to use occasionally because of its analog nature. … (The) potentiostat goes into oscillations.”
Although the M173 has a reputation for ‘stability’ it has always had problems with oscillation! These problems tend to be most troublesome when the more sensitive current ranges are used and when the cell capacitance is large. PAR had a ‘noise filter’ in their catalog for a long time that was really a ‘stability aid’ more than a ‘noise filter.’ It consisted of a capacitor that was placed between the counter electrode lead and the input jack of the M178 electrometer. This acts as a shunt for the higher frequency, oscillation-producing signals. A capacitor value of 0.01 µF is a good place to start. I think this stands the best chance of stabilizing your system.
This is one of the tricks outlined in the PAR Application note TN 200. That note can be http://www.gamry.com/application-notes/reference-electrodes/requested from their website. Although that note talks specifically about the M273, some of the comments apply to nearly all potentiostats. You will find many of the same suggestions on the GAMRY website in their Tech Tip on ‘Reference Electrode Effects on Potentiostat Performance‘.
Another trick which is often successful with high capacitance cells is to add resistance to the COUNTER electrode lead. This makes the potentiostat work harder by requiring a greater potential at the potentiostat’s counter electrode lead. This has the effect of slowing down the potentiostat, making it less prone to oscillate.
There is a description of potentiostat stability (written by DK Roe) in the Kissinger & Heineman book (See Section 7.V in the 2nd Ed.). It’s a bit heavy on the math. However, the manual for the Solartron 1286 (and I presume the 1287 as well) has a nice description in their chapter on selecting ‘Stability and Bandwidth’ parameters.