Adding a second fixed gear output does not halve the first output’s turn count. Each branch follows its own tooth-count ratio to the common input. In our new Dual Output Gear Bench, one turn of the central 24-tooth gear drives the 8-tooth output three turns and the 40-tooth output three fifths of a turn. Both outputs turn opposite to the input. What a second load can change is the effort required, or the actual input speed if the drive cannot maintain it.
Separate a gear’s speed from its strength
Brick Experiment Channel’s credited demonstration asks how much load an 8-tooth gear can handle. Its expanded description identifies two versions of that gear and reports results for a particular powered stress-test rig. The demonstration footage shows assembled gear arrangements; the inspected description does not provide a complete assembly tutorial. Those results belong to that creator’s setup. They are not a load rating for our model or an invitation to break parts.
Our original Dual Output Gear Bench asks a gentler question: can a small fast output and a large slow output share one input without changing each other’s turn ratio? It has 37 compatible parts, fourteen assembly steps and an estimated twenty-minute build time. Its geometry and arrangement are our educational design, not a reconstruction of the video’s stress machine. You can answer the question by turning gently and counting, without a motor, weights or destructive tests.
Three vertical shafts sit in separate bearing bridges across a flat frame. The middle 24-tooth gear directly touches the front 8-tooth gear and the rear 40-tooth gear. The outputs do not touch each other. This is a branching drive: input to fast output, and input to slow output. Compare Gear Friends, where motion travels through a serial chain, or the Compound Counting Bench, where reductions multiply through shared shafts. Drawing the connections first prevents you from multiplying ratios along a path that does not exist.
Predict both branches before turning
For an external spur pair, each passing input tooth requires a matching output tooth to pass the contact point. Divide input teeth by output teeth to get output turns per input turn. Our first branch gives 24 divided by 8, or three. The second gives 24 divided by 40, or 0.6. Each direct external mesh reverses the input, so the two outputs turn in the same direction as each other.
Enter three in the prediction field, then check it. Move the input-angle slider to 360 degrees. The signed readouts show −3 fast-output turns and −0.6 second-output turns. The minus sign means opposite to the chosen positive input direction; it does not mean negative speed or a broken gear. Move to 1,800 degrees: five input turns produce fifteen fast turns and three slow turns. That is a convenient complete return to all three starting orientations.
Interactive: Two-output prediction lab. Predict turns in two independent fixed gear branches: a 24-tooth input drives an 8-tooth output three turns and a 40-tooth output 0.6 turns per input turn, both reversed. Compare teeth, toggle a branch and reset; counts do not establish load or power sharing.
Toggle the second branch out of the diagram while leaving the input and first output unchanged. The first branch still needs three turns for each input turn. Now vary only the second branch’s tooth count and predict what stays constant. Its own count changes; the other branch’s ratio does not. These are hypothetical diagram experiments. Different gears need different shaft spacing and a new clearance review; do not swap them into the authored bench at its fixed hole positions.
Count actual rotation, not assumed motor speed
A fixed mesh constrains angles. If an additional load makes a motor slow down, both branches may complete fewer turns in ten seconds because the input completed fewer turns. That does not mean their turns per actual input turn changed. Counting for a fixed time and counting for a fixed input rotation are different experiments. Our hand-powered bench uses the latter, with no claim about rpm or maximum load.
The ratios also do not specify how power divides. A lightly loaded branch may need very little torque, while another may resist strongly. Friction and real loads matter. This is not a differential: either stopped output prevents the common input from turning while its teeth remain engaged. Do not hold an output still and force the input. The gears explanation introduces direction and tooth counts; the speed and torque guide helps keep speed, effort and strength separate.
Make a fair counting experiment
Place the model on a level table and hold the stationary frame. Use the input shaft above the central gear and turn slowly. Choose a visible tooth or shaft face as a reference, and sketch the starting orientations on separate paper kept clear of the gears. Keep your viewing direction the same. Count one input revolution, then repeat with five. Record three trials before changing anything. If a starting mark is uncertain, write that down rather than rounding the observation to the prediction.
| Actual input turns | Predicted fast output | Predicted slow output | Observed counts, three trials |
|---|---|---|---|
| 1 | 3 reverse turns | 0.6 reverse turn | Record / record / record |
| 2 | 6 reverse turns | 1.2 reverse turns | Record / record / record |
| 5 | 15 reverse turns | 3 reverse turns | Record / record / record |
We have not run physical trials; the observation cells are intentionally blank. Direction is as useful as count. Small counting errors can come from losing the fast reference, overshooting the input mark or confusing a partial turn with a complete one. Repeat slowly with one observer assigned to each shaft. Avoid adding friction by gripping a moving shaft.
Check what keeps every shaft in place
Each bridge is fixed by two separated friction pins, so it cannot simply hinge around one connection. Its centre round hole allows the axle to turn and resists sideways displacement. Half bushes directly above and below the bridge prevent axial escape. A further cap retains the gear on its upper spacer. Four retained corner feet carry the base to the table. The axles guide explains the difference between a round bearing hole and a cross hole that keys a gear to its shaft.
If a gear lifts out of mesh, check the underside bush and the cap before turning again. If the bridge rocks, seat both pins. If teeth meet point to point, turn the whole output shaft gently so its teeth enter the input’s gaps; keep its keyed parts together. Do not force it down. If anything binds, stop and check gear height, spacers and shaft alignment. Holding a loose output by hand would hide the missing support rather than fix it.
The animated instructions and booklet show the same retention and insertion order. Model checks cover the complete operating cycle, tooth proximity and unwanted sideways, lift, drop and tilt motions. A deliberately omitted underside retainer must allow an escape in the regression test, showing why a smooth assigned animation alone is insufficient. These finite geometry checks are useful design evidence, not measurements of friction, tolerances, strength or real performance. Finish by explaining which connection fixes each output’s ratio and why the second branch cannot change that count just by existing.
Interactive: Check yourself. A few quick questions on the ideas in this guide, each with an explanation.
- One central 24-tooth input directly drives an 8-tooth output. How many output turns per input turn?
- A second 40-tooth output also meshes directly with the input. Does this halve the 8-tooth branch’s count?
- Which way do the two outputs turn relative to each other?
- What does a turn-count calculation tell you about the power split?
