Thermal cycler checks after a power cut

After a power cut, the number on a thermal cycler screen is a claim about a sensor, not proof that your programme finished. This explainer is the check. Did the block complete the run, did the programme start again at cycle one, and is the heated lid still at the set-point the method named? Only then do you decide whether a precious plate can go on that machine. The programme you are trying to protect is the cycle in how polymerase chain reaction works.

The photograph shows a gloved hand closing the lid of a benchtop cycler on a strip of tubes, with city lights beyond the glass. That lid is part of the instrument you have to re-trust. Instruments and plastic are specified through the molecular biology catalogue and the quote request. This page does not assign a calibration number.

What actually stops when the power stops

A cycler is three systems that must agree. The block heats and cools the tubes through a listed series of holds. A counter remembers which cycle and which step it is in. The heated lid, a separate heater, stays warmer than the sample so water does not park on the caps and leave the reaction. A power cut removes energy from all three. Fans stop, displays go dark, and any step that was in progress is incomplete.

When power returns, instruments do not share one behaviour. Some resume the step they were in. Some abort and sit at a safe hold. Some reboot and wait for a person. Some can be set to start a stored protocol from the first cycle. You cannot see which of those happened from a single temperature flashing on the screen. The manual for the model you have is the authority for that model's resume logic. Read it once, before the next outage, and write the expected behaviour into the laboratory note.

Did the block finish?

Look at the run log, the cycle counter and the clock, not only at the block temperature. If the log shows the final extension or the final hold starting before the outage, the thermal programme may have finished, and the remaining question is whether the tubes sat hot or cold for a long time afterwards. A long hot soak can degrade product or dry a poor seal. A long cold sit is usually kinder to the DNA and still does not repair a run that had already failed.

If the log stops mid-cycle, the block did not finish. Samples experienced an unknown partial cycle: perhaps a denaturation without the matching extension, or an extension cut short. Endpoint product may be weak, mixed, or absent for a reason that has nothing to do with the template. Do not interpret those lanes as biological negatives. Repeat them once the instrument has been checked.

Write down what the display said at the moment power returned, including any error text. Memories of "it looked normal" are how a restarted cycle disappears from the record.

Did the programme restart at cycle one?

A restart is a different failure from an abort. The counter returns to the beginning and the samples receive extra cycles. Endpoint PCR already stops being quantitative once reactions plateau, because primers, nucleotides and enzyme are finite. Extra cycles push tubes deeper into that plateau, grow primer-dimer, and can turn a barely specific anneal into a visible side band. A real-time run that restarts has a trace you cannot line up with the samples that ran uninterrupted.

Compare the log's start time with the time of the outage. A start time after the power returned means this execution began again. A cycle count that is higher than the programme you loaded means something added cycles. Either finding is enough to quarantine the plate's interpretation. The tubes might still contain usable product for a low-stakes check. They are not a precious result.

Is the lid still at its set-point?

The heated lid has its own set-point, often planned somewhere around 100 to 105 Celsius on many instruments, always as the manufacturer specifies for that model. After a reboot the lid heater may be off while the block heater reports a familiar sample temperature. Caps then run cooler than the liquid. Water condenses under the cap, and in a small volume a share of the reaction can leave the bottom of the tube. If cycling continues in that state, magnesium, primers and dNTPs concentrate in whatever liquid remains. Late cycles run in a different mixture from the one you pipetted.

Read the lid temperature on the display and compare it with the set-point stored in the programme you intended to run. Compare it also with the laboratory's calibration or service record for that instrument. If the display and the set-point disagree, do not "correct" the lid by typing a number you invented. Service the instrument, or follow the manufacturer's recovery steps, and then prove the block with controls.

Two cyclers, one displayed temperature

Blocks differ in geometry, sensor placement, lid pressure and wear. Plastic that fits one block rocks in another. A calibration certificate, where the institution has one, is a record of a test done on a date, by a method, with an outcome. It is not a universal offset you can copy from this article onto a second machine. National measurement practice, of the kind a body such as NIST describes for weights and measures, is a reminder that a number without a record is not a calibration. Your laboratory's service sheet is the record that counts.

So a cycler that "shows 58 degrees, same as the other one" can still anneal differently. After an outage, the functional check is a known positive and a no-template control, in the plastic you will use for the real samples, on the programme you will use, on the block that lost power. The positive shows that this mix and this block can still amplify. The no-template control shows that you did not declare the cycler healthy by contaminating the test. If the positive fails, the precious set waits. If the blank shows the assay product, the outage is no longer your only problem, and the reagents need the same suspicion as the block.

What you may conclude

A log that shows a completed programme, a lid that matches its set-point, a positive that amplifies, and a no-template control that stays clean are the conditions for trusting the next precious set on that block. Any one failure sends you back to a repeat, not to a corrected story about the samples that were on the machine during the cut.

A UPS, a generator transfer, or a policy of never starting a precious plate in a storm window are institutional equipment decisions. This explainer does not require them and does not forbid them. It asks you to treat the outage as a break in the thermal history until the log and the controls say otherwise.

Safety and the limit of the claim

Blocks and lids stay hot after a short outage and can be burn hazards when power returns unexpectedly. Do not lean on a lid that is recovering. Research results remain research results. A cycler that passed today's checks is still not a diagnostic platform.

In buildings where supply interruptions are part of ordinary working life, write the resume behaviour of each cycler into the induction note while the power is on. The note is free. Reconstructing it from a dark screen is not.

What to send with an enquiry

If you are specifying a cycler or a replacement block, state that you need a run log a person can read after an interruption, a heated-lid set-point you can display, and plastic that matches the block. Name endpoint or real-time use, vessel profile, and the control habit you will use to re-qualify the instrument: a known positive and a no-template control. The nucleic acid analysis pathway is the workflow around the instrument. Ask whether a quotation is possible. Do not ask the seller to invent a calibration offset the institution has not measured.