Replacing a Discontinued Instrument Without Requalifying the Whole Rack

Replacing an instrument · 3 min read · 10 cited facts

The instrument fails, or the vendor stops answering, and the question arrives fully formed: can we drop something else in without redoing the qualification? The honest answer is that it depends on four things, and only one of them is the hardware.

Check them in this order

• Command set — does your existing test code talk to the replacement without edits?
• Measurement path — is the reading taken the same way, at the same point, with the same reference?
• Electrical envelope at your operating point, not at the corner of the datasheet
• Documentation trail — can you evidence equivalence to whoever audits you?

Most programs check the third first because it is the one on the quote. It is usually the one that matters least, because two instruments that both cover your envelope can still be entirely different to integrate.

The command set is the real cost

A decade-old test executive has accumulated program-specific behavior: sequences, limit trees, error handling built around one instrument's quirks. Two instruments can both claim SCPI compliance and still disagree about defaults, timing and what happens on an out-of-range request. Audit command coverage against your actual code, not against the vendor's compliance claim.

What actually forces requalification

Changing the measurement path usually does. If the replacement reads current at a different point, references to a different ground, or applies different filtering, your data is not directly comparable to the historical set and an auditor is entitled to say so. Changing the source of stimulus while measuring identically usually does not, provided you can evidence the stimulus is equivalent.

Evidence equivalence while you still can

If the outgoing instrument still works, run a correlation study before you retire it — the same units, both instruments, same sequence. That dataset is the cheapest thing you will ever produce and the most expensive to reconstruct after the old unit is gone. Programs routinely skip it because the replacement is urgent, and then spend far longer arguing about equivalence than the study would have taken.

Mixed fleets are a legitimate answer

There is no requirement that a lab standardize. Running the outgoing instrument for existing qualified sequences while new work goes onto the replacement is often cheaper and lower-risk than a clean cutover, particularly on sustainment programs where the qualified configuration is contractually fixed.