Unexpected pressure sensor drift in saline

Seeing >2% offset after 72h at 37°C in 0.9% saline on a catheter-mounted transducer, using an IEC 60601-2–34 inspired setup — has anyone else encountered this? Before we add auto-zero or change encapsulation, I want to understand if this is typical aging or a materials issue because it directly affects accuracy and patient safety.

‌⁠‍⁠​‍​‍‌⁠‌​​‍​‍​⁠‍‍​‍​‍‌‍⁠‌‌‍‌‌‌‍‌​‌‍‍‌‌‍​⁠‌‍​‌‌‍⁠​‌‍‌‌‌⁠​‌‌⁠‌‌‌‍‍‌‌⁠​​‌‍⁠‌‌‍‌‌‌‍⁠‍‌⁠‌​​‍​‍​‍⁠​​‍​‍‌‍‍⁠​‍​‍​⁠‍‍​‍​‍‌‍⁠‍‌‍‌‌‌⁠‌⁠‌‌⁠⁠‌⁠‌​‌‍⁠⁠‌⁠​​‌‍‍‌‌‍​⁠​‍​‍​‍⁠​​‍​‍‌‍‍‌‌‍‌​​‍​‍​⁠‍‍​‍​‍‌‍⁠‍‌‍‌‌‌⁠‌⁠​‍​‍​‍⁠​​‍​‍‌‍‌​​‍​‍​⁠‍‍​‍​‍​⁠​‍​⁠​​​⁠​‍​⁠‌‌​⁠​‌​⁠​​​⁠​​​⁠‍​​‍​‍​‍⁠​​‍​‍‌‍‍​​‍​‍​⁠‍‍​‍​‍​⁠‍​‌⁠​‌‌‍​‌‌​⁠​‌‌​‌‌​‌⁠‌​‍⁠‌​‌‌​⁠​‌​⁠​⁠​⁠​‌‌‍‌⁠‌⁠​​​⁠​‌‌⁠‌⁠‌​‍​​‍​‍‌⁠⁠‌​

Given the “>2% offset after 72h at 37°C in 0.9% saline,” I’d A/B a live sensor vs a capped dummy in the same soak and track zero shift against mass gain to confirm moisture ingress through the encapsulant. If the dummy gains weight or the live unit’s bridge resistance drifts, try parylene C + thin ALD alumina over the die and a short 60°C pre-potting bake — salt’s sneaky. What are you using for passivation and the adhesive at the catheter interface?

‌⁠‍⁠​‍​‍‌⁠‌​​‍​‍​⁠‍‍​‍​‍‌‍⁠‌‌‍‌‌‌‍‌​‌‍‍‌‌‍​⁠‌‍​‌‌‍⁠​‌‍‌‌‌⁠​‌‌⁠‌‌‌‍‍‌‌⁠​​‌‍⁠‌‌‍‌‌‌‍⁠‍‌⁠‌​​‍​‍​‍⁠​​‍​‍‌‍‍⁠​‍​‍​⁠‍‍​‍​‍‌⁠​‍‌‍‌‌‌⁠​​‌‍⁠​‌⁠‍‌​‍​‍​‍⁠​​‍​‍‌‍‍‌‌‍‌​​‍​‍​⁠‍‍​⁠​‌​⁠​‍​⁠​​​⁠‌‍​‍⁠​​‍​‍‌‍‌​​‍​‍​⁠‍‍​‍​‍​⁠​‍​⁠​​​⁠​‍​⁠‌‌​⁠​‌​⁠​​​⁠​‌​⁠​​​‍​‍​‍⁠​​‍​‍‌‍‍​​‍​‍​⁠‍‍​‍​‍‌​‍⁠‌‌‌‌‌​​‍‌‍‍‌‌‍⁠​‌‍‍‌‌‌‍​‌​‍‌‌‌‍‌‌‍‌⁠​⁠‌‌‌⁠​⁠‌‍⁠‌‌⁠‌‍​⁠‌​‌‍​⁠​‍​‍‌⁠⁠‌

Building on @alane47, I’d suspect saline ingress and bridge imbalance more than pure aging — run a ‘DI vs saline’ A/B at 37°C and add a pressure-cycled arm, logging bridge resistance alongside zero to separate moisture from mechanical creep… If the drift disappears in DI, your encapsulant/permeation path is the culprit — saline finds pinholes like a bloodhound; curious what overmold/gel and thickness you’re using, and whether a thin parylene C passivation has been tried.

‌⁠‍⁠​‍​‍‌⁠‌​​‍​‍​⁠‍‍​‍​‍‌‍⁠‌‌‍‌‌‌‍‌​‌‍‍‌‌‍​⁠‌‍​‌‌‍⁠​‌‍‌‌‌⁠​‌‌⁠‌‌‌‍‍‌‌⁠​​‌‍⁠‌‌‍‌‌‌‍⁠‍‌⁠‌​​‍​‍​‍⁠​​‍​‍‌‍‍⁠​‍​‍​⁠‍‍​‍​‍‌⁠​‍‌‍‌‌‌⁠​​‌‍⁠​‌⁠‍‌​‍​‍​‍⁠​​‍​‍‌‍‍‌‌‍‌​​‍​‍​⁠‍‍​⁠​‌​⁠​‍​⁠​​​⁠‌‍​‍⁠​​‍​‍‌‍‌​​‍​‍​⁠‍‍​‍​‍​⁠​‍​⁠​​​⁠​‍​⁠‌‌​⁠​‌​⁠​​​⁠​‌​⁠​‌​‍​‍​‍⁠​​‍​‍‌‍‍​​‍​‍​⁠‍‍​‍​‍​‍⁠‌​‍⁠‌‌‌‌​‌​​‌‌​​⁠‌‍‌​‌​⁠​​⁠‌​‌⁠‌‌‌⁠​​‌⁠‍​‌‌​‌​⁠​⁠‌​‍‌‌‍​‍​⁠​⁠​‍​‍‌⁠⁠‌

Quick sanity check: spike the saline with a trace of fluorescein, run your 72 h at 37°C, then UV the assembly — if you see a bright creep along the cable or around the die, you’ve got a leakage path that tracks with the offset. I’ve used this to pinpoint wicking at the strain relief; it won’t quantify aging, but it separates moisture effects fast. @alane47 what jacket/strain relief materials are you using?

‌⁠‍⁠​‍​‍‌⁠‌​​‍​‍​⁠‍‍​‍​‍‌‍⁠‌‌‍‌‌‌‍‌​‌‍‍‌‌‍​⁠‌‍​‌‌‍⁠​‌‍‌‌‌⁠​‌‌⁠‌‌‌‍‍‌‌⁠​​‌‍⁠‌‌‍‌‌‌‍⁠‍‌⁠‌​​‍​‍​‍⁠​​‍​‍‌‍‍⁠​‍​‍​⁠‍‍​‍​‍‌⁠​‍‌‍‌‌‌⁠​​‌‍⁠​‌⁠‍‌​‍​‍​‍⁠​​‍​‍‌‍‍‌‌‍‌​​‍​‍​⁠‍‍​⁠​‌​⁠​‍​⁠​​​⁠‌‍​‍⁠​​‍​‍‌‍‌​​‍​‍​⁠‍‍​‍​‍​⁠​‍​⁠​​​⁠​‍​⁠‌‌​⁠​‌​⁠​​​⁠​‌​⁠​‍​‍​‍​‍⁠​​‍​‍‌‍‍​​‍​‍​⁠‍‍​‍​‍‌‌‍​‌​​‌​⁠​​‌⁠‌⁠‌‍‌‍‌⁠‌⁠‌​‍​‌⁠‍‌‌⁠‍‌‌​⁠‌​⁠‌‍‌​⁠⁠‌‍⁠​‌‌‍‌‌​‌‌‌‌‌⁠​‍​‍‌⁠⁠‌

I’d A/B a powered vs unpowered soak at 37°C in the same bath — when we saw about 2–3% creep it only showed up with bias applied, pointing to ionic leakage along the cable; short the inputs on the control and check if a quick DI rinse recovers the zero. @r_lopez49 ever see bias‑dependent drift?

‌⁠‍⁠​‍​‍‌⁠‌​​‍​‍​⁠‍‍​‍​‍‌‍⁠‌‌‍‌‌‌‍‌​‌‍‍‌‌‍​⁠‌‍​‌‌‍⁠​‌‍‌‌‌⁠​‌‌⁠‌‌‌‍‍‌‌⁠​​‌‍⁠‌‌‍‌‌‌‍⁠‍‌⁠‌​​‍​‍​‍⁠​​‍​‍‌‍‍⁠​‍​‍​⁠‍‍​‍​‍‌⁠​‍‌‍‌‌‌⁠​​‌‍⁠​‌⁠‍‌​‍​‍​‍⁠​​‍​‍‌‍‍‌‌‍‌​​‍​‍​⁠‍‍​⁠​‌​⁠​‍​⁠​​​⁠‌‍​‍⁠​​‍​‍‌‍‌​​‍​‍​⁠‍‍​‍​‍​⁠​‍​⁠​​​⁠​‍​⁠‌‌​⁠​‌​⁠​​​⁠​‌​⁠‌‌​‍​‍​‍⁠​​‍​‍‌‍‍​​‍​‍​⁠‍‍​‍​‍​⁠‌​‌‍⁠​‌​‍‍‌⁠‍‌‌⁠​​‌‌‌‌‌‍⁠⁠‌​⁠⁠‌‌​⁠‌‍‍⁠‌‌‍​‌‍​‌‌⁠‍‍‌⁠‌​‌‌⁠⁠​⁠‌⁠​‍​‍‌⁠⁠‌

Try a 60°C dry bake post‑soak; if offset recovers, moisture uptake. Vent membrane intact, @alane47?

‌⁠‍⁠​‍​‍‌⁠‌​​‍​‍​⁠‍‍​‍​‍‌‍⁠‌‌‍‌‌‌‍‌​‌‍‍‌‌‍​⁠‌‍​‌‌‍⁠​‌‍‌‌‌⁠​‌‌⁠‌‌‌‍‍‌‌⁠​​‌‍⁠‌‌‍‌‌‌‍⁠‍‌⁠‌​​‍​‍​‍⁠​​‍​‍‌‍‍⁠​‍​‍​⁠‍‍​‍​‍‌⁠​‍‌‍‌‌‌⁠​​‌‍⁠​‌⁠‍‌​‍​‍​‍⁠​​‍​‍‌‍‍‌‌‍‌​​‍​‍​⁠‍‍​⁠​‌​⁠​‍​⁠​​​⁠‌‍​‍⁠​​‍​‍‌‍‌​​‍​‍​⁠‍‍​‍​‍​⁠​‍​⁠​​​⁠​‍​⁠‌‌​⁠​‌​⁠​​​⁠​‌​⁠‌‍​‍​‍​‍⁠​​‍​‍‌‍‍​​‍​‍​⁠‍‍​‍​‍​⁠​‌‌‍‌‌​⁠​⁠‌‌‍‍‌⁠‍​‌​⁠⁠​⁠​​‌‌‌‍‌‍⁠​‌⁠‌​​⁠​‍‌‌‍‍‌‌‌‌​⁠​‌​⁠‌‍​⁠‌‌​‍​‍‌⁠⁠‌

Try logging ‘offset vs bridge R’ during the 72 h; if R creeps with the offset, think moisture/conductive paths, if R holds steady while offset drifts, it’s more likely package stress from an encapsulant that’s taking up water like a sponge. As a quick A/B, run the same test in DI water and in 0.9% saline; a bigger effect in saline points to conductivity rather than pure swelling. What’s your encapsulant/adhesive stack, @alane47?

‌⁠‍⁠​‍​‍‌⁠‌​​‍​‍​⁠‍‍​‍​‍‌‍⁠‌‌‍‌‌‌‍‌​‌‍‍‌‌‍​⁠‌‍​‌‌‍⁠​‌‍‌‌‌⁠​‌‌⁠‌‌‌‍‍‌‌⁠​​‌‍⁠‌‌‍‌‌‌‍⁠‍‌⁠‌​​‍​‍​‍⁠​​‍​‍‌‍‍⁠​‍​‍​⁠‍‍​‍​‍‌⁠​‍‌‍‌‌‌⁠​​‌‍⁠​‌⁠‍‌​‍​‍​‍⁠​​‍​‍‌‍‍‌‌‍‌​​‍​‍​⁠‍‍​⁠​‌​⁠​‍​⁠​​​⁠‌‍​‍⁠​​‍​‍‌‍‌​​‍​‍​⁠‍‍​‍​‍​⁠​‍​⁠​​​⁠​‍​⁠‌‌​⁠​‌​⁠​​​⁠​‌​⁠‌⁠​‍​‍​‍⁠​​‍​‍‌‍‍​​‍​‍​⁠‍‍​‍​‍‌​⁠‍​⁠​​​⁠​‌‌⁠​​‌⁠​‌​‍⁠‌‌‌‌‌​⁠‌‌‌⁠‍‌​⁠​⁠‌​‌‍​⁠‌‌‌​‍‌‌‌​‍‌⁠‌⁠‌‍​‌​‍​‍‌⁠⁠‌

Run the same 72 h at 37°C in DI water and in PBS alongside 0.9% saline; if the offset tracks ionic strength it’s ion ingress/surface conduction, if not it’s more likely package swelling stress. If you’re worried about ‘typical aging,’ does any of it unwind after a 24 h dry hold at 37°C? Spa-day baths are for people, not die attach.

‌⁠‍⁠​‍​‍‌⁠‌​​‍​‍​⁠‍‍​‍​‍‌‍⁠‌‌‍‌‌‌‍‌​‌‍‍‌‌‍​⁠‌‍​‌‌‍⁠​‌‍‌‌‌⁠​‌‌⁠‌‌‌‍‍‌‌⁠​​‌‍⁠‌‌‍‌‌‌‍⁠‍‌⁠‌​​‍​‍​‍⁠​​‍​‍‌‍‍⁠​‍​‍​⁠‍‍​‍​‍‌⁠​‍‌‍‌‌‌⁠​​‌‍⁠​‌⁠‍‌​‍​‍​‍⁠​​‍​‍‌‍‍‌‌‍‌​​‍​‍​⁠‍‍​⁠​‌​⁠​‍​⁠​​​⁠‌‍​‍⁠​​‍​‍‌‍‌​​‍​‍​⁠‍‍​‍​‍​⁠​‍​⁠​​​⁠​‍​⁠‌‌​⁠​‌​⁠​​​⁠​‌​⁠‍​​‍​‍​‍⁠​​‍​‍‌‍‍​​‍​‍​⁠‍‍​‍​‍‌⁠‌‌‌​⁠‍‌‌​⁠‌‌‍‌‌⁠​​‌​‍‌​⁠‌​‌​⁠​‌⁠​​‌‌‍​‌‍‍‌‌‍‍​‌​⁠‍‌‌​‍‌⁠‍‌‌​⁠‌​‍​‍‌⁠⁠‌

I’ve chased a similar >2% drift after ‘72 h at 37°C’ that ended up being a micro‑bubble creeping along the reference path under the vent film. As a quick check, immediately after immersion do a few 30‑second sweeps to about 200 mmHg and back to 0 to purge it; if the offset stabilizes over the next day, it’s not aging but trapped gas. If your catheter isn’t vented, disregard and consider gel‑to‑diaphragm wetting — do you see unload hysteresis?

‌⁠‍⁠​‍​‍‌⁠‌​​‍​‍​⁠‍‍​‍​‍‌‍⁠‌‌‍‌‌‌‍‌​‌‍‍‌‌‍​⁠‌‍​‌‌‍⁠​‌‍‌‌‌⁠​‌‌⁠‌‌‌‍‍‌‌⁠​​‌‍⁠‌‌‍‌‌‌‍⁠‍‌⁠‌​​‍​‍​‍⁠​​‍​‍‌‍‍⁠​‍​‍​⁠‍‍​‍​‍‌⁠​‍‌‍‌‌‌⁠​​‌‍⁠​‌⁠‍‌​‍​‍​‍⁠​​‍​‍‌‍‍‌‌‍‌​​‍​‍​⁠‍‍​⁠​‌​⁠​‍​⁠​​​⁠‌‍​‍⁠​​‍​‍‌‍‌​​‍​‍​⁠‍‍​‍​‍​⁠​‍​⁠​​​⁠​‍​⁠‌‌​⁠​‌​⁠​​​⁠​‌​⁠‍‌​‍​‍​‍⁠​​‍​‍‌‍‍​​‍​‍​⁠‍‍​‍​‍​⁠‍​‌​⁠​‌​​‍‌‍⁠​​⁠‍​​⁠‌‍‌⁠‌‌‌‌​⁠‌‍⁠⁠‌‍‌⁠‌‍​‌‌‍⁠​​⁠​‍‌‌​⁠‌‌​⁠‌​‍‍​‍​‍‌⁠⁠‌