Key takeaways
- Every spring click wrench document checked here says to store at the lowest torque setting, and the ones that address it say never to turn below it.
- The lowest mark is not a fixed 20 percent. Across seven wrenches it runs from 10 percent of maximum to 41 percent.
- In Norbar's one-month test, wrenches left at full scale moved more than wrenches wound down but all stayed in tolerance.
- Split beam and beam wrenches have no preloaded spring to relax, so they can be stored as they are.
- Recalibration defaults run from every 6 months (Harbor Freight) to 12 months or 5,000 cycles, and a dropped wrench should be checked before use.
The short answer by wrench type
The rule depends on what is inside the handle. A micrometer click wrench holds a coil spring compressed against a cam or toggle, and the number you set is how hard the spring is compressed. Winding down relaxes that spring. Other designs have nothing equivalent to relax.
| Wrench type | How to store it | Why |
|---|---|---|
| Micrometer click (adjustable handle) | Wind down to the lowest printed number, then lock | Spring is compressed at whatever value is set |
| Click, below the lowest mark | Never | Norbar says going below the minimum marking can adversely affect calibration |
| Split beam click | As it is | Precision Instruments says no wind-down is needed |
| Beam (pointer) | As it is | No adjustable spring at all |
| Click wrench in daily production use | Lowest mark if convenient, otherwise as set | Norbar calls either acceptable for frequent users |
The first two rows carry the answer. The lowest mark is where the wrench should rest, and zero is a different place.
What seven wrenches actually print
The common shorthand is that the bottom of the scale is 20 percent of maximum. Norbar's own wording is "usually 20% of maximum." It holds for two of the seven wrenches below and misses the other five. The table below takes each wrench's range from its own manual or product page and works out where the lowest mark actually falls.
| Wrench | Drive | Printed range | Lowest mark | Lowest mark as share of max | Storage instruction |
|---|---|---|---|---|---|
| Harbor Freight item 61277 | 1/4 in | 20 to 200 in-lb | 20 in-lb | 10% | Lowest setting after every use |
| Tekton 24330 | 3/8 in | 10 to 80 ft-lb | 10 ft-lb | 12.5% | Lowest setting, 10 ft-lb; do not turn below |
| GearWrench 85063M | 1/2 in | 20 to 150 ft-lb | 20 ft-lb | 13.3% | Lowest setting, but not below minimum |
| Harbor Freight Icon 57923 | 1/2 in | 40 to 250 ft-lb | 40 ft-lb | 16% | Lowest setting after every use |
| Harbor Freight Icon 59108 | 3/8 in | 20 to 100 ft-lb | 20 ft-lb | 20% | Lowest setting after every use |
| Precision Instruments C2FR100F (split beam) | 3/8 in | 20 to 100 ft-lb | 20 ft-lb | 20% | No wind-down needed |
| GoMichaelPro MP001002 | 1/2 in | 70 to 170 Nm (50 to 130 ft-lb) | 70 Nm | 41% | Lowest level, 70 Nm; do not turn below |
The spread runs from 10 percent to 41 percent. That matters for two reasons. First, "wind it to 20 percent" is a calculation nobody needs to do, because the right target is printed on the tool. Second, on the GoMichaelPro wrench the lowest mark is 70 Nm (about 52 ft-lb), so a storage setting that looks high is in fact the bottom of the scale, and someone applying the 20 percent rule would try to go below it.
Why zero is the wrong place
The wind-to-zero habit is widespread. Norbar's note on the subject opens by saying that anyone working in torque will have been taught to wind their wrench back to zero after every use. The manuals do not say zero. They say lowest setting, and three of the documents here add an explicit prohibition:
- Tekton: "Do not turn handle below lowest torque setting."
- GoMichaelPro: "Do not turn adjustment below lowest torque setting."
- GearWrench: store at the lowest setting "but not below its minimum."
Norbar is the most direct. If the wrench is wound back at all, "it should not be adjusted below the minimum scale marking (usually 20% of maximum) - never to zero as this can adversely affect the calibration of the wrench."
The calibration procedure explains why the bottom mark is a meaningful number rather than an arbitrary one. Under ISO 6789-2:2017, as Atlas Copco summarizes it, a click wrench of classes A, D and G is checked at its minimum capacity, at 60 percent of maximum and at 100 percent of maximum, with five results at each point and a 4 percent deviation limit. The minimum mark is one of the points the wrench was verified at. Below it, the wrench is outside anything its certificate describes.
Zero is not on the scale of most of these wrenches in the first place. The Tekton 24330 scale starts at 10 ft-lb and its manual states the wrench "does not measure torque below 10 ft-lb (13.6 Nm)." There is no zero setting to store it at, only a direction the manual tells you not to turn.
Does storing it wound up actually hurt it?
This is the real source of the argument, and the honest answer is that it hurts less than the folklore suggests.
The manuals give the reason for winding down. Tekton says storing at the lowest setting "relieves extra tension on the internal spring, eliminating fatigue that can adversely affect accuracy." Harbor Freight and GoMichaelPro give the instruction without a reason.
Norbar then tested the claim. Four wrenches came off its production line and went through its calibration lab. Two were left wound to maximum and two to their minimum scale setting. The lab took results again after 24 hours, after a week and after a month. The outcome, in Norbar's words: "all of the wrenches remained within tolerance, regardless of whether they were left at the minimum scale setting or at their maximum. The wrenches left wound to maximum did move more but the change happened quickly, within 24 hours, and then they stabilised."
So both circulating positions hold part of the truth. Spring set is real: the wound-up wrenches moved more. It is also small and quick, not a slow creep that ruins a wrench over a weekend. Norbar's resulting advice splits by user. The occasional user should wind back to the minimum mark after use. In a commercial garage where the wrench is in use all day, "the choice is yours," and Norbar notes it has thousands of production wrenches left at their setting for months on end without a problem.
The practical reading for anyone who owns one or two wrenches and uses them weekly or less: wind down. It costs ten seconds, every manual asks for it, and the test that supports leaving it wound up covered one month on one maker's wrenches. The wrench that sits in a drawer from one brake job to the next is exactly the occasional use Norbar has in mind.
The wrenches that do not need winding down
Split beam. Precision Instruments builds its click wrenches around a split beam mechanism. The listing for the C2FR100F states there is "no need to wind down after use" and that the internal split beam mechanism "allows for immediate storage without the requirement to wind the tool down to its lowest setting after use." It is the one wrench in the table where the storage instruction is to do nothing.
Beam. A pointer beam wrench is a bar that bends and a needle that does not. There is no adjuster and no spring, so no storage setting exists.
Production users of any click wrench. As above, Norbar treats leaving a wrench set as acceptable when it is used constantly. That is a statement about Norbar wrenches in heavy use, and the cheaper wrenches in the table carry their own instruction to wind down after every use.
Storage beyond the setting
The setting gets the attention, but the other storage instructions in these manuals are worth as much.
| Instruction | Who says it |
|---|---|
| Store in its case | Harbor Freight (both manuals), GoMichaelPro ("protective packaging") |
| Indoors, clean, dry location | Harbor Freight; Tekton says clean, dry environment |
| Wipe with a clean dry cloth, no liquid, do not immerse | Tekton and Harbor Freight |
| After long storage, operate several times at a low setting to recoat moving parts with lubricant | Tekton |
| If dropped, test or calibrate before putting it back into use | Harbor Freight (both manuals) |
| Avoid temperature extremes, high humidity, fluid immersion and corrosive environments | FenderBender care guidance |
Three of those translate directly into where the wrench lives in a box.
Dry side of the chest. A torque wrench belongs with the meters and gauges, away from lubricants and brake fluid, which is the wet side and dry side split in how to lay out tool chest drawers. The no-liquid cleaning rule is the reason: Tekton warns that immersing the wrench in any liquid or cleaner may damage its internal components.
Its own pocket. A case or a cut foam pocket stops it from rolling against other tools, which is the drop-and-knock damage Harbor Freight treats as grounds for recalibration. A wrench laid flat needs about 2 in (50 mm) of foam, per the depth table in the Kaizen foam guide. If the drawer has a liner instead, avoid plasticized PVC under a painted scale; drawer liners compared explains how plasticizer migration softens painted scales over time.
Not the truck box. The manuals say indoors, clean and dry. A truck box swings through temperature and humidity every day, and the condensation that follows is the subject of rust prevention in tool storage. A wrench that rides to jobs should come back inside with the batteries, for the same reason given in tool batteries in a truck box or cold garage.
Calibration interval, by document
Storage is the small variable. Use, overload and impact move calibration more, which is Norbar's own conclusion: how wrenches are used and sometimes abused "will have a far greater influence on calibration" than whether one is left wound up. So the interval matters more than the storage setting.
| Source | Interval stated | Other trigger |
|---|---|---|
| GoMichaelPro MP001002 manual | After about 5,000 cycles or 12 months | None stated |
| FenderBender care guidance | 5,000 cycles or up to one year, whichever comes first | If you suspect it was dropped |
| Harbor Freight 61277 manual | Every 6 months | After a drop, test before use |
| Harbor Freight Icon 57923 and 59108 manual | Every 6 months | After a drop, calibrate before use |
| Harbor Freight Icon warranty | Out-of-box accuracy covered for 90 days | Not a recalibration interval |
| GearWrench 85063M | Calibration accuracy warranted for 90 days | Not a recalibration interval |
Twelve months or 5,000 cycles is the common default. Harbor Freight's 6 months is stricter, and for those wrenches the manual is the document to follow. The 90 day figures are warranty periods. They describe how long the maker stands behind the factory calibration, not how long the wrench stays accurate.
To make the interval stick, put a calibration due date on the wrench's record. The tool inventory template in shadow boards, labels and tool inventory control has a cal_due field for exactly this.
Accuracy is not specified at the bottom of every scale
One more reason not to treat the lowest mark as a working setting. GearWrench specifies the 85063M at plus or minus 3 percent clockwise "in the upper 80% range of maximum torque." On a 150 ft-lb wrench, the upper 80 percent starts at 30 ft-lb, so the stated accuracy covers 30 to 150 ft-lb while the scale is printed from 20. The bottom 10 ft-lb is on the scale and outside the accuracy statement.
Tekton adds a practical note for the low end: at low settings the click can be subtle, so pull slowly and use the wrench in a quiet environment. If the fastener spec sits near the bottom of a wrench's range, a smaller wrench is the better tool. A 1/4 in drive 20 to 200 in-lb wrench covers the 36 in-lb J-hook bolts worked out in crossover tool box J-hooks. A 3/8 in drive 10 to 80 ft-lb wrench would need to be set to 3 ft-lb for the same job, below its first mark.
Where this guidance stops applying
- Your manual overrides this page. These are seven specific wrenches. A wrench whose maker gives a different storage instruction should be stored the way its maker says.
- The Norbar test is small. Four wrenches, one month, one manufacturer. It shows the wound-up shift is quick and was within tolerance. It does not show that any wrench can be left at full scale for a year.
- Electronic and dial wrenches are a different case. They measure torque with a sensor or a deflecting element rather than a preset spring, and none of the documents here gives them a storage setting.
- Storage does not fix overload. A wrench used past its maximum, used to break fasteners loose, or dropped needs a calibration check whatever setting it was stored at.
The ten-second routine
- After the last fastener, unlock the handle.
- Turn down to the first number on the main scale. Not zero, not a percentage.
- Lock the handle so the setting cannot drift in the case.
- Wipe it with a dry cloth. No solvent, no dunking.
- Into its case or foam pocket, in a dry drawer indoors.
- If it has sat for months, click it several times at a low setting before trusting it on a real fastener.
- If it hit the floor, it gets checked before it torques anything that matters.