How to Time an NC Servo Feeder to a Press Stroke?
The line was rated at 300 strokes a minute and it would not hold feed length above 180. Nothing was broken.
The feed window had simply run out of milliseconds, and the setting sheet had been copied from a 150 SPM job.
Timing an NC servo feeder to a press is arithmetic before it is a setting.
The window is fixed by the press, the feed length is fixed by the part, and the only variable left is how hard the rolls are allowed to accelerate.
This is the six-step method used on our own run-offs, with the numbers worked through at 200 SPM, plus the point where the arithmetic stops being useful.
The full error budget that governs feed accuracy from coil to die is set out in the coil line guide library. The machine range is listed under our machines.
Read the Feed Window Before You Touch a Setting
The feed window is the crank angle during which strip can move without the pilots standing in the die. Everything else is derived from it.
It opens when the pilot has cleared the strip and closes when the pilot re-enters. On a progressive die with pilot release, that band is typically 200 to 240 degrees of crank rotation.
The mistake that costs the most time is starting from the feed length. Feed length tells you how far the strip travels.
The window tells you how long it has to do it in, and that second number is the one that decides whether the job is possible.
Write both numbers down before adjusting anything. A feeder that looks slow is often a feeder with a window that closed earlier than the operator assumed.
The Four Numbers You Need From the Press
Four figures come off the press and the die before the feeder is programmed. All four are usually available without a stopwatch.
- Strokes per minute at production rate. This sets the whole cycle length. At 200 SPM a full revolution takes 300 milliseconds.
- The crank angles that bound the window. From the die drawing or by measuring pilot release, in degrees. Typically around 40 to 260 degrees after top dead centre.
- Feed length per stroke. The advance, not the part length. On a two-out die those are different numbers, and using the wrong one halves or doubles the required speed.
- Settling time before pilot entry. How long the strip needs to be still before the pilot touches it. Usually 10 to 20 milliseconds, and it comes out of the window.
Two of these four are often wrong in the first draft. Strokes per minute gets quoted at maximum rather than at production rate, and the feed length gets quoted as the part pitch.
Working the Numbers at 200 Strokes a Minute
With the four figures in hand the arithmetic takes about two minutes, and it tells you whether the job fits before any hardware is touched.
Start with the cycle. At 200 SPM one revolution is 300 milliseconds. A 200 degree window is therefore 200 divided by 360, multiplied by 300, which is 167 milliseconds.
Take out 15 milliseconds of settling time and 152 milliseconds remain for the move itself.
Now apply the feed length. For a 250 mm advance in 152 milliseconds, average strip speed is about 1.6 metres a second, and a triangular velocity profile puts the peak at roughly twice that.
The acceleration that implies is in the region of 35 metres per second squared, close to 3.6 g. That is at or beyond what a roll pair can transmit on oily 1.2 mm steel without slipping.
The arithmetic is not telling you the feeder is unsuitable. It is telling you the window has to grow, the profile has to be softened, or the press speed has to come down.
On this job the answer was a combination of all three.
Setting the Release Signal Against Pilot Entry
The release signal is where a correctly sized feeder still fails, because the strip is ready but the die is not.
The feeder issues a feed-complete signal, and the press uses it to enable the pilot. The gap between those two events has to cover the settling time plus any response delay in the press control.
A 24 volt signal through a solid-state output typically adds 2 to 5 milliseconds of its own. On a 300 millisecond cycle that is small, and on a 150 millisecond cycle at 400 SPM it is worth counting.
Wire the feed-complete signal as a hardware interlock as well as a logic input. A software-only handshake that is bypassed during commissioning tends to stay bypassed.
Then check the strip, not the screen. Pilot marks that appear as a light witness on the strip mean the pilot arrived early. Marks that appear only on some strokes mean the timing is marginal rather than wrong.
Proving the Timing Without Scrapping a Coil
A new timing is proved in stages, and each stage costs a few minutes rather than a coil.
| Signal or setting | What it gives the cycle | Where it costs you |
|---|---|---|
| Widening the feed window | More milliseconds for the same move, so lower peak acceleration | Needs pilot release in the die, which is a tooling change rather than a feeder setting |
| S-curve acceleration profile | Lower peak demand and less strip slip at the start of the move | Adds a few milliseconds of its own, so it only helps when the window has slack |
| Earlier feed-complete signal | Longer settling time before the pilot enters | Eats into the window, and taken too far the strip is still moving when the pilot lands |
| Hardware feed-complete interlock | A press that cannot cycle on a missed feed | One more circuit to maintain, and it must be proved rather than assumed |
Run the first proving pass at 80 per cent of production speed for three coils. Measure the feed length with a gauge on the finished strip, not from the HMI counter.
Check the strip for pilot witness marks, then look at the scrap bin rather than the good-part bin. A timing that is slightly early shows up in scrap before it shows up in a length measurement.
Only then take the press to full rate, and re-measure after an hour of running. Thermal growth in the press and a warming gearbox both move the timing slightly.
Where the Timing Calculation Stops Being Useful
The arithmetic above assumes the window can be moved. On several press types it cannot, and that changes the decision.
A mechanical press has a fixed cam. The angles that release the pilot are machined into the press.
If the window is too short at production speed, the feed length or the strokes per minute have to change, and no feeder setting will help.
The window shrinks in milliseconds as speed rises. A 200 degree window is 167 milliseconds at 200 SPM and 83 milliseconds at 400 SPM. Doubling the rate halves the time, while the acceleration demand quadruples.
Acceleration figures in a catalogue are measured without strip. A quoted figure describes the drive and the rolls, not the friction pair between the rolls and your material at your oil film thickness.
Long feeds and short windows are incompatible. Above roughly 400 mm of advance, a high-speed mechanical press rarely offers a window wide enough.
The honest answer there is a slower line, a shorter pitch or a different process route.
What Timing Cannot Recover
Timing is a control-layer fix. Four problems on a press line sit below it and will survive any amount of retuning.
Slip from the wrong roll surface. If the coating or the oil film cannot transmit the required force, a longer window reduces the demand but does not remove the mismatch.
Mechanical backlash. Gearbox backlash and coupling wear add an error that no signal timing corrects. It shows up as a length error that changes direction with the load.
An undersized feeder for the strip mass. A heavy, wide strip carries inertia that has to be accelerated and stopped. Beyond a point, the answer is a larger machine rather than a cleverer profile.
A die that cannot release its pilots. If the tooling holds the strip through the whole cycle, there is no window to time against, and the job needs a die modification first.
FANTY runs more than 80 engineers across design and commissioning, and the timing conversation is usually short once the window and the feed length are both on the table.
Recording the Setting So It Survives a Shift Change
A timing that is proved on one shift and lost on the next is a timing that will be rediscovered in scrap.
Record the feed length, the window in degrees, the profile, the settling time and the strokes per minute as one recipe. A recipe that stores the feed length alone cannot be reproduced.
Note the material and the oil film alongside it. Both change the acceleration the rolls can transmit, so the same recipe on a different coil can slip.
Keep the pilot witness check as a routine inspection rather than a commissioning step. It is the cheapest early warning that timing has drifted.
FANTY machines are in service across more than 60 countries, and the sites that hold feed length across years are the ones that treat the recipe as a controlled document.
Questions that come up when a feed window is too short.
How many milliseconds is one press revolution at 200 SPM?
300 milliseconds. At 400 SPM it is 150 milliseconds, which is why the same feed becomes far harder at double the rate.
How much window does a 250 mm feed need at 200 SPM?
Around 200 degrees or more. Below that the required peak acceleration approaches 3.6 g, which is near the grip limit on oily 1.2 mm steel.
How much settling time should the strip have before the pilot enters?
Typically 10 to 20 milliseconds. Less than that shows up as pilot witness marks, and more than that steals window from the move.
Can the cam timing on a mechanical press be changed?
Rarely. The pilot release angles are built into the press cam, so the feed length or the strokes per minute have to change instead.
What speed should the first proving run use?
80 per cent of production rate, for three coils, with the feed length measured on the strip rather than read from the screen.
Does an S-curve profile reduce the required acceleration?
It reduces peak demand and jerk by roughly 15 to 25 per cent, and it adds a few milliseconds of its own. It changes the shape, not the average.
Where to go next
Timing decides whether the strip arrives in time. How much of the remaining error the line can absorb, and where it comes from, is worked through in the feed accuracy guide.
the coil line guide library
