The Best Torque Wrench Calibrators and Testers

Tester, calibrator, or calibration service?

The three words get used interchangeably in listings, and they are not the same thing. A tester or checker measures what your wrench actually delivers and shows you the error. A calibration system applies known torque and produces a certificate. A calibration service is a laboratory that does the second one for you and issues the paperwork.

CDI puts its own instrument on one side of that line: "Unlike a torque calibration system, which can be used to generate a torque wrench certification, the ETT is a portable, battery powered stand-alone torque tester." Every product on this page is a tester in that sense. They will tell you a wrench has drifted. None of them produces the certificate.

That is still most of the value. A check takes minutes, it catches the wrench that has been dropped or left wound up, and it tells you whether the annual lab visit is due early. Our calibration guide covers the procedure, including the three test points and the error arithmetic.

How accurate does the tester have to be?

More accurate than the wrench, by a margin. The long-standing rule of thumb is a 4:1 test uncertainty ratio, which puts the reference about four times inside the tolerance of the tool being checked: a ±1% tester against a ±4% click wrench, or a ±0.5% instrument against the ±2% and ±3% wrenches. It is a rule of thumb rather than a requirement — accredited laboratories now work from a documented uncertainty budget instead — but it is a sound way to shop.

Read the small print on where the rating applies. GEARWRENCH rates the bench testers here at ±1% clockwise from 20% to 100% of the range, so on the 25–250 ft-lb model that is 50 ft-lb upward; its touchscreen models are rated differently again. CDI rates the ETT at ±0.5% of indicated value, clockwise and counter-clockwise, from 10% to 100% of full scale. Below those points the tester is not lying to you, but it is no longer inside its stated tolerance, and neither is the wrench you are judging.

• Pick a tester whose range brackets your wrench's own range, rather than the biggest one you can get.
• Check the direction: the GEARWRENCH testers here are rated clockwise only.
• A wrench tested at the very bottom of a big tester is being judged in that tester's worst zone.

First peak is the mode that matters

A click wrench releases and the torque immediately falls away. Ordinary peak mode records the highest value it saw, which on a click wrench can be the overshoot after the break rather than the break itself. First-peak mode captures that first release instead, which is what you set the wrench to.

All the electronic instruments here have it: GEARWRENCH calls the modes track, peak and first; CDI's checker offers first peak, peak hold and tracking, and its ETT adds a power-tool mode. A digital torque adapter is the simplest way to get a reading, but adapters generally offer peak and tracking only, which is the main reason they are a sanity check rather than a test.

The tester needs calibrating too

A reference instrument drifts like any other. GEARWRENCH warrants the calibration accuracy of its bench testers for 90 days and the parts for a year, which is a fair signal of how the maker views the interval. CDI ships its digital checker with a certificate of calibration in the box, and Precision Instruments sells a certificated version of its comparator for buyers who need the paperwork.

So treat the tester as the thing you protect: keep it bolted down, do not overload it, and have it recalibrated on a schedule. A reference that has quietly drifted is worse than no reference, because it makes a bad wrench look fine.

When to use a laboratory instead

Send the wrench out when the answer has to be documented or when the wrench is outside tolerance. A calibration laboratory applies known torque, adjusts the wrench where the design allows it, and issues a certificate with the readings on it. That is what an employer, an auditor or an airworthiness process wants to see, and no bench tester substitutes for it.

Send it out, too, when a check says the wrench is off. Unless the maker publishes an adjustment procedure, do not open the handle: on most click wrenches the spring preload is set at the factory and re-set against a reference you do not have. Use your tester to decide which wrenches make the trip, and how urgently.

What we left out, and why

Several things that come back under these search terms are not torque wrench testers at all, and a couple that are were not buyable:

• Bottle-cap torque testers. A large share of the "torque tester" listings measure the opening torque of caps and lids for packaging work. They do not take a square drive.
• Dial-indicator torque wrenches. These are indicating wrenches, useful tools in their own right, but they are the thing being checked, not the reference.
• Used instruments. One professional torque verifier came back described by its seller as used; a reference tool with no known history is not a reference.
• Unbranded calibrators with no maker documentation. Several listings quote an accuracy figure with nothing behind it that we could read.
• CDI's older DTT testers, which were unavailable or on long lead times across the models we checked, and which one distributor describes as discontinued and replaced.

Where the guidance comes from

The specifications and statements on this page come from these documents. Product sources are listed under each pick.

• GEARWRENCH product pages for the 89631, 89634 and 89635 bench-top torque testers.
• CDI Torque Products catalogue, for the ETT electronic torque tester's accuracy, modes and scope.
• Torque Products Canada's specification sheet for the CDI DTC digital torque checkers.
• Precision Instruments' torque comparator pages, for the TC3F175F's capacity and accuracy.
• Amazon listings for each pick, checked against the maker's figures.
• Our own calibration guide for the test-point and error method.