Pancake Load Cells

Two things, both structural. The first is direction: the DBSL, the DBSL-XS and the TC414 are compression load cells, and only the DBSL can be converted for tension, through its optional tension base. The load path exists while the load presses the rim down onto its seat, so a reversed load does not read negative, it unseats the part. If the structure can lift (wind on a vessel, torque from a mixer, a specimen that snaps back in a test frame), the uplift has to be carried somewhere else.

The second is the mounting surface. A beam spans between a fixed end and a loaded end, and the arm between them absorbs some imperfection in how it was bolted down. A pancake has no arm, so the plate underneath sits inside the measuring system and its flatness is a specification.

The DBSL sheet draws the series in three bands (1 to 5 t, 10 to 20 t and 30 t), and each one has its own body. The centre thread goes M16×1.5, M32×1.5, M40×1.5, and the base pattern from eight 7.5 mm holes to eight 11 mm, so two capacities that look adjacent on a price list can need two different mounting plates.

The bridge changes too. At and below 20 t the DBSL is 387 ±20 ohm in and 350 ±5 ohm out; above that it is 775 ±20 ohm and 700 ±5 ohm. Two bridge resistances cannot be summed in one junction box and trimmed sensibly, so check resistance as well as capacity when replacing one load cell out of a set.

The DBSL-XS has two datasheets of its own, one for 5 and 10 t and one for 20 and 30 t. Performance matches the DBSL: 3.0 mV/V ±0.25%, ±0.03% nonlinearity, hysteresis, repeatability and 30-minute creep, 150% safe overload, compensated −10 to +40 °C, on a 6 m four-conductor shielded cable. The bodies do not: the 5 and 10 t are 154 mm across and 45 mm tall on an M24×1.5 thread, the 20 and 30 t 141 mm and 50.8 mm on M40×1.5. The bridge splits at a different point too: 387 ±20 ohm in and 350 ±5 ohm out at 5 and 10 t, 775 ±20 ohm and 700 ±5 ohm at 20 and 30 t.

Where the requirement is “hold this load, in this height, for the least money”, the XS is the sensible buy. Where a plate is being machined to a drawing, work from the sheet for the capacity being bought: DBSL and DBSL-XS bodies, threads and bolt patterns differ.

The TC414 is built as a drop-in for that series and the capacities map one to one: TC414-300LBF for the LC414-300, and so on through 600, 1.2K, 2.4K, 6K, 12K, 24K and 30K lbf. The mount is the same flange arrangement and load goes in through a 1in-8 UNC-2B central thread.

Electrically it is 3 mV/V from a 387 ±20 ohm bridge at 10 V rated excitation, on an integrated four-conductor 28 AWG shielded cable, with nonlinearity and hysteresis of ±0.05% and repeatability of ±0.02% of rated output. Confirm those against the instrument you already have, and recalibrate afterwards.

The top face is not a platform. Only the central boss carries load: the DBSL takes M16×1.5 on the 1 to 5 t body, M32×1.5 from 10 to 20 t and M40×1.5 at 30 t; the TC414 takes 1in-8 UNC-2B.

Screw the rod, dome or button into that thread and make sure it is the only thing touching the top face. A plate laid across the whole top will find the rim as soon as anything deflects, and load that reaches the rim without passing through the web never gets measured. A hardened dome against a flat pad is the usual arrangement, because it delivers force on the axis without carrying a moment into the body.

The load cell assumes force along the axis of the boss. Anything off that axis is partly measured and partly reacted through the mounting, and the split is not constant: it moves with the load, with temperature and with how the structure has settled. The usual sources are a vessel that expands thermally and drags its feet sideways, a frame that is not square to its base, and piping bolted rigidly to a weighed structure. Take lateral force out with check rods anchored to something that is not weighed, and confirm the loading axis is perpendicular to the mounting plate rather than merely parallel to the frame, then confirm it again with the vessel full.

Only the DBSL, and only with its optional tension base; all three are compression load cells. The load path exists because the load presses the rim onto its seat, so a reversed load does not read negative, it unseats the part. For pulling or hanging loads use an S-beam.

Flat enough that the rim bears all the way round, which is a practical test rather than a number: blue the plate and look at the contact pattern, or check it with a straightedge before the load cell goes on.

No. It is a separate low-profile part sold from 5 to 30 t, with its own datasheets. Output, accuracy and overload match the DBSL (3.0 mV/V, ±0.03%, 150%), but the bodies, threads (M24×1.5 and M40×1.5) and bridge resistances differ, so one is not a drop-in for the other.

That is what it is built for. The capacities map one to one from 300 through 30,000 lbf, the mount is the same flange, and load goes in on 1in-8 UNC-2B. Recalibrate afterwards: a replacement load cell never inherits the old calibration.

None of these three carries an NTEP certificate. Where the reading has to be a legal weight, the SBS single-ended shear beam is certified under CoC 95-146 from 1,000 through 10,000 lb.