Limit and Force Verification After a Repair

Purpose

Every setting that decides where a garage door stops also decides how hard it pushes before it gives up. Move one and you have moved the entrapment protection, whether you meant to or not. This SOP makes the reversal test part of the repair rather than a courtesy at the end, and gives the shop one acceptance every tech runs the same way whichever repair sent them there.

Without it a cable job, a board swap or a seal change ends with a door whose reversing behavior nobody has checked since the parts changed, and no record either way.

Scope

Runs at the end of any visit that moved a travel limit, changed a force setting, changed anything in the drive train, changed the door's weight or the friction it runs against, or replaced the board or motor.

Does not cover commissioning a newly installed operator, owned by garagedoor-opener-replacement-and-safety-reversal-test. Does not cover finding and fixing a photo-eye fault, owned by garagedoor-photo-eye-alignment-and-verification; this SOP only proves the pair works on the finished job. Does not cover commercial operators with hold-to-close constant-pressure controls, which have no inherent reversal to test.

Roles and responsibilities

What the limit sets and what the force sets

The travel limit tells the operator where the door is finished. The force setting tells it how much resistance counts as an obstruction rather than as normal running. They fail into each other, which is why they are verified together and never one at a time.

• Down limit past true floor contact. The operator keeps driving after the door is down, reads the floor as an obstruction every cycle, and the tech turns close force up to stop the nuisance reversals. The door now closes on a body with the force needed to fight a slab.
• Down limit too shallow. The door stops short, the seal never compresses, and the callback is about daylight and rodents.
• Close force raised to get through a bind. Travel completes, the mechanical fault is hidden, and so is the reversal that would have happened at the original setting. Force is never the fix for a hard spot.
• Auto-learning heads. Many current operators learn force during a travel-learn cycle rather than on a screw. That cycle runs against whatever friction exists that day, so it happens after the mechanical work, never before.

Procedure

1. Gate the whole procedure on the counterbalance. Unplug the operator, keep the plug in your pocket per the cord-and-plug exemption at 29 CFR 1910.147(a)(2)(iii)(A), release the trolley with the door DOWN, and work the door by hand. Acceptance is moderate one-hand lift and the door holding within 2 in at mid travel over 10 seconds. Stop rule: a door failing this gets no limits or forces set, because every setting is calibrated to a counterbalance about to change. Hazard: never pull the release with the door up, since the door drops under its own weight.

2. Clear the opening and keep it clear throughout. Move the vehicle out and get people, children and pets outside the door plane. Acceptance is an empty opening confirmed aloud before every powered cycle, not once at the start. Wrong looks like a customer standing in the opening to watch. Stop rule: the cycle does not start until they move. Hazard: this procedure deliberately drives a door onto an obstruction, so the opening is a moving hazard for its whole duration.

3. Re-engage the trolley and restore power from beside the head. Push the trolley into the latch until it snaps audibly, then plug in standing to the side of the head rather than under it. Acceptance is a latched carriage confirmed by hand pull, and one full cycle with no nuisance reversal before any setting is touched. Wrong looks like a trolley that re-latches only if you hold the lever. Stop rule: an unlatched carriage releases the door under power, so it is fixed first. Hazard: stand clear of the door plane at first power-up.

4. Set the up limit so the door stops clear without bottoming the trolley. Acceptance is a fully open door with the bottom section clear of the opening and the trolley not hitting its stop at the header end. Wrong looks like a bang at the end of open travel, or a bottom section still hanging in the opening where a truck mirror catches it. Stop rule: re-set it rather than absorbing it with force. Hazard: keep clear under the door running to full open, since a top section that has never gone that far can foul the rail.

5. Set the down limit to true seal contact at the slab. Acceptance is a closed door with the bottom seal just compressing evenly across the full width and no chain or belt slack jumping at the end of travel. Wrong looks like the door driving into the floor after contact, or daylight at one end. Stop rule: a limit past floor contact is corrected here, before force is touched, because setting force against a bad limit calibrates the operator to fight the slab. Hazard: hands and feet out from under the bottom rail while cycling to check seal contact.

6. Set close and open force by the manual's procedure for that model. On a screw-adjust head follow the manual's turns; on an auto-learning head run the travel-learn now the mechanical work is complete. Acceptance is full travel both ways with no nuisance reversal at the manufacturer's setting or learned value, and a pass at step 7. Stop rule: if the door needs an elevated setting to finish travel, that is a mechanical bind and it goes back to the rail, roller or counterbalance work rather than being settled with a screw. Hazard: fingers clear of the drive and out of the door's path during force trials, which reverse without warning.

7. Run the contact reversal test, using the smaller object where the manual permits. Lay the obstruction flat under the center of the opening and close from full open. Acceptance is contact, reversal, and full return to open within the time the manual states for that model, commonly 2 seconds from contact. Most manuals name a 2x4 laid flat, a 1.5 in object; the compliance test in 16 CFR Part 1211 uses a 1 in object. The taller board is contacted earlier in travel, with more distance left before the down limit, so the operator has more room to read the resistance as an obstruction rather than as the end of travel, which makes the 2x4 the easier test to pass. Stop rule: stopping on the object is a FAIL, not a marginal pass, and so is a pass on the 2x4 with a fail on the 1 in object; both send you back to steps 5 and 6. Hazard: nobody's hand, foot or tool is ever used as the obstruction, and the tech stands outside the door plane while it closes.

8. Run the photo-eye test on a moving door, three times. Close from full open and break the beam from outside the opening with a broom handle, at a point clear of the door's path, at a different point in travel each time. Acceptance is immediate reversal to full open on all three. Wrong looks like reversal on two of three, which is an alignment or wiring fault, not a marginal pass. Stop rule: any failure routes to garagedoor-photo-eye-alignment-and-verification and the door does not go back until it passes here. Hazard: break the beam from outside the opening only, never by reaching a hand or foot into the door's path.

9. Re-run both tests after any further adjustment, then record or red-tag. Any change to a limit, a force, the seal or the drive after step 8 invalidates both results and both run again. Acceptance is a final pair of results with observed values written down, taken after the last adjustment of the visit. The pair is the point: UL 325 lists a residential head with an inherent contact reversal plus a secondary means, and 16 CFR Part 1211 requires both, so one passing result is half a job. Stop rule: a door that cannot pass both is red-tagged, the trolley disconnected, the door left on manual, and the customer told plainly it must not be run by the opener until fixed. Hazard: leave the door DOWN and the operator unplugged when you red-tag, so nobody restores it and finds out the hard way.

The record this produces

• Fields: operator make, model and serial; what repair triggered this verification; balance drift in inches over 10 seconds; up and down limit as-left described by observable state including seal contact across the width; force method used, either the manual's turns or a completed travel-learn; contact reversal result with the object height used and the observed time from contact to reversal; whether full return to open occurred; photo-eye result as three separate outcomes with their break points; and any red-tag with its reason.
• Where it lands: the door's equipment record against the operator serial, plus a photo of the test object in position under the closing door.
• Who reads it later: the next tech comparing today's settings against a nuisance-reversal complaint, the manager on a callback, and the owner or their insurer after an incident, because a recorded reversal time beside a serial is a verification rather than an assertion.

Worked pass: after a lift cable replacement, one failed step

Single 9 by 7 steel door, chain drive, in for a cable that came off the drum. Cable work complete under its own SOP, and this verification ran before the door went back.

Step 1 passed: released with the door down, hand-worked, held at mid travel with about 1 in of drift over 10 seconds, inside the 2 in acceptance. Step 2: vehicle out, homeowner moved to the side door, confirmed aloud before every cycle. Step 3: trolley latched audibly and held on a hand pull, power restored from beside the head, one clean cycle.

Steps 4 to 6: up limit clear of the opening with no bang at the header end; down limit set to even seal compression across the full width, which took two attempts because the first left daylight at the left jamb; the head was an auto-learning model, so the travel-learn ran after the cable work and completed with no nuisance reversal.

Step 7 failed and took its stop rule. With a 2x4 laid flat the door contacted the board and reversed cleanly to full open inside the manual's stated time. With a 1 in object at the same spot the door contacted it, hesitated, and stopped on it. Under the step that is a FAIL, and the mechanism is the one the step names: at the lower object the door was nearer the down limit at contact, so the operator read the resistance as the end of travel. The tech went back to step 5, found the down limit still a fraction past true seal contact from the pre-repair setting, corrected it, re-ran the travel-learn under step 6, and retested. Second attempt: contact on the 1 in object, reversal starting inside about a second, full return to open. The 2x4 result was then re-run for the record and also passed.

Step 8: three closes, beam broken with a broom handle from outside the opening at three different points in travel, immediate reversal to full open on all three. Step 9: no further adjustment after the last test, so the final pair stood, and both were recorded with the object height, the observed reversal time and the three break points rather than as "OK". No red-tag needed.

References

• 16 CFR Part 1211 and the UL 325 listing it incorporates, binding manufacturers of operators made on or after January 1, 1993. Its compliance test object is 1 in high; the reversal timing, force procedure and permitted test object for a given head are in the manual shipped with that listed unit, and the manual governs at the door.
• 29 CFR 1910.147(a)(2)(iii)(A), the cord-and-plug exemption relied on in step 1.
• See related: garagedoor-opener-replacement-and-safety-reversal-test, garagedoor-photo-eye-alignment-and-verification, garagedoor-door-balance-test-standard, garagedoor-cable-off-the-drum-correction-standard.