The Objectives:
The product was required to meet the heavy industrial limits for radiated immunity, but failed at a range of frequencies around 35 MHz with the temperature reading error well outside of the customers specified accuracy. At a relatively low frequency it was likely that one or more cables were responsible for this noise coupling. Cables were disconnected or suppressed one at a time until the coupling route was identified.

The Investigation:
The system could now be removed from the chamber and the PCB together with the coupling cable placed on the ETSi 4334. An external noise source was used in inject a 35 MHz signal into the cable. An X-Y plot of the PCB at the same frequency showed the areas which were affected; In this case the input stages of a differential amplifier. From this information the design engineers were able to apply suitable filtering to just those areas of the PCB that needed it.
The temperature controller was then re-assembled and found to be compliant when returned to the chamber.

Conclusion:
The use of the ETSi Scanner led to a solution which filtered the noise from the affected circuitry, and avoided expensive screened cable or impracticable external ferrites.