The $740 Transmitter That Wasn't Broken—and the Sartorius Support Call That Fixed My Checklist

By the time I called Sartorius support, I'd already done the expensive thing. I'd replaced a pressure transmitter that wasn't broken with a spare that cost us $740. The spare did exactly what the old one did: for the first twenty minutes it read beautifully, and then the pressure on the HMI started to float while the calibrated reference gauge stayed rock steady.

I've been handling process instrumentation and lab support orders for seven years. I've personally made—and documented—fourteen significant mistakes, totaling roughly $46,000 in wasted budget. That's why I maintain our team's checklist. This story is why rule one exists.

The setup: single pressure sensors & transmitters on a small TFF skid

We were supporting a client site in September 2024. The suite had a small tangential-flow filtration rig for an mRNA purification process. They used a few Sartorius single pressure sensors & transmitters, one on the feed line and one on the permeate line. Nothing exotic. You put a calibrated gauge next to them and forget they're there.

Until the feed pressure reading started to drift.

I remember the day clearly because the process had to be paused. The output from the feed transmitter was jumping from 15 psi to 23 psi and then back again. Our calibrated gauge on the same port never moved. That's not a normal process change; that's a signal problem.

I checked the connector, the cable, and the calibration certificate. All looked fine. Then I made the assumption: it's the transmitter. I grabbed a spare from the shelf, installed it, and watched the same thing happen. That spare cost $740 on the last purchase order. It was not the answer.

What I skipped: a five-minute voltage check

I didn't call Sartorius support until the second failure. I don't have a clean explanation. The client needed the skid running quickly, and replacing the sensor looked like the fastest fix. It wasn't.

When I finally called, the support engineer asked a simple question: 'Did you measure the supply voltage at the transmitter while the pump was actually running?'

I hadn't. I had measured it with the pump off, seen 23.9 V, checked the calibration certificate, inspected the connector, and decided the sensor was bad. But I hadn't tested the loop under load.

The support engineer was patient. She told me to set a multimeter across the transmitter terminals, run the pump until the reading started drifting again, and record the voltage at that moment.

So here's how to use Fluke 117 true RMS multimeter for this check. Put it on DC volts. Red lead to the VΩ jack. Black lead to COM. Touch the transmitter terminals while the circuit is energized. Then watch the reading long enough for the system to warm up, not just long enough to say you did it. If the voltage drops by more than a couple of volts when the load kicks in, you have a loop problem, not necessarily a sensor problem.

When I did that, the voltage started at 23.8 V and then dropped to 11.2 V as the fault kicked in. A 24-volt loop doesn't work properly if the device only sees 11 volts. The transmitter wasn't the problem. The problem was a terminal screw on the 24 V feed that had backed off. It passed idle current, but under load it didn't.

The thermal camera made the rest obvious. Our diagnostic bag is simple: a Fluke 117 true RMS multimeter, 1 thermal camera—an entry-level model—and a small screwdriver set. I scanned the terminal strip while the process was faulted, and one terminal on the 24 V feed was noticeably warm. The bad connection was turning voltage into heat. We re-terminated the wire, put the original transmitter back, and the reading held exactly where the reference gauge said it should.

What the support call actually taught me

Here's something vendors don't put on the return authorization form: a lot of pressure transmitters that get swapped in the field are not faulty when they reach the bench. They fail under load. The loose connection, dying power supply, or bad ground stays in the panel, and the next transmitter repeats the same failure.

The same logic applies on the analytical side. A method that doesn't work is often a symptom of something upstream: wrong mobile phase pH, column not equilibrated, or an outdated protocol. Before I redesign an assay, I now look for an application note or call Sartorius support to ask what the official method conditions are. That habit has caught 47 potential process errors in our lab over the last 18 months, and it has saved far more than $740.

What this has to do with HPLC double stranded RNA

The week after the pressure transmitter incident, I was on another call about mRNA analytics. We were using a dot-blot method for dsRNA impurities, and the result didn't give us the resolution we needed. The support engineer mentioned BIA Separations, whose CIM monolithic columns are now part of the Sartorius portfolio. I didn't know much about that product family, so that night I typed 'hplc double stranded rna bia sartorius' into a search box, all lowercase, because I wasn't sure what BIA was.

It was the right question, even if it looked odd. I found an application note for an HPLC double stranded RNA method on a small CIMac DEAE analytical column. Instead of spending half a day on a blot that was hard to quantify, we switched to a chromatography run with a peak we could integrate.

That doesn't mean every old method is wrong. But support gave me a better option, just as it gave me a better way to check a pressure transmitter.

The point is not that I'm a hero. I'm the person who swapped a good transmitter because I skipped a voltage check. The checklist I maintain now starts with the boring steps: compare with a reference, test under load, and call Sartorius support before you spend money on a spare.