Mazatrol Bar Puller and Bar Feeder Programming: Where Small Errors Scrap Every Part

Bar work is where Mazatrol programming stops being forgiving. A one-off job with a small error in it makes one bad part and somebody catches it. A bar program with a small error in it makes the same bad part every cycle, unattended, until the bar is gone or the shift ends. I spent 19 years at Mazak Corporation as an Applications Engineer setting up this kind of work, and the years since doing the same thing independently. The bar jobs that scrap parts almost never fail on the machining. They fail on the pull.

A bar puller is not a bar feeder

People use the terms interchangeably, and in programming terms they are different jobs.

A bar feeder pushes. The bar sits in the spindle, fed from behind by the feeder, and the program's job is mostly about managing the cutoff, the facing allowance, and how much stock the feeder advances each cycle. The chuck stays closed and does the holding.

A bar puller pulls. The puller is a tool in the turret — it comes in, grips the bar or the finished part, the chuck opens, and the turret pulls the bar forward to position before the chuck closes again. Now the program is coordinating four things that all have to be true at the same time: the puller's position, its grip, the chuck state, and the spindle state. That coordination is where the programming discipline lives.

If your machine runs a sub-spindle or a parts catcher with the pull, you're coordinating even more of it, and the same logic applies at every handoff.

The pull sequence, in the order it has to happen

Whatever control generation you're on — T-Plus, Fusion 640, Matrix, Smooth — the conceptual sequence is the same:

1. Position the puller at the bar, at a Z position that accounts for where the grip actually lands relative to the part and the cutoff.
2. Grip. The puller closes on the bar. This is not optional sequencing: the grip has to be made before the chuck opens, every time, without exception.
3. Open the chuck only after the grip is made.
4. Pull to the programmed position. The turret draws the bar out to the length the next part needs, plus facing stock. This position is the one that sets your part length consistency, so it deserves the same respect as a work offset.
5. Close the chuck, release the puller, retract clear. Then the machining cycle starts.

Write that order into the program correctly and the machine will do it ten thousand times. Get the order or the positions slightly wrong and it will also do that ten thousand times.

The mistakes I see

Pulling before the grip is real. If the sequence lets the chuck open before the puller has genuinely closed on the bar — a worn puller, a grip diameter that doesn't match the bar, a timing assumption that isn't true — the bar slides, drops, or pulls short. Every symptom downstream looks like a length problem. The cause is upstream, in the handoff.

Z position errors that hide in the math. The pull position has to account for the stock the face cut will remove, the cutoff tool width, and where the gripper sits. Shops often chase part length variation with offsets when the actual error is that the pull position was set once, approximately, and never verified against a measured part. Measure the first parts, then trust the pull. Not the other way around.

Ignoring what part-off does to the setup. Cutoff is the most violent thing that happens in a bar cycle. Conceptually: as the parting tool feeds in and the cross-section thins, the part's behavior changes, and surface speed at the center behaves differently than at the diameter. Controls give you ways to manage how the spindle behaves during part-off, and every experienced bar programmer uses them deliberately instead of accepting whatever the default cycle does. If your cutoff finish is bad, parts are breaking off early, or you're getting a nub, the answer is in how the part-off is programmed, not in a new insert.

Treating the grip as setup instead of program. On a puller job, the grip diameter, the pull position, and the chuck state are program decisions as much as tooling decisions. When the next bar size changes, all of them get reviewed — not just the machining units.

Twin-spindle bar work

On a Multiplex or a Quick Turn with a sub-spindle, bar work often becomes a handoff problem instead of a pull problem: the sub comes over, grips, the main releases, and the part transfers. The discipline is identical — never let both chucks be open on the same part, and never assume a grip is made because the program said so. If your machine does transfers, the sub-spindle transfer guide covers the offset side of that handoff.

Why this matters more than it looks

A bar job that runs right is the best money a lathe shop makes: lights-out cycles, consistent parts, an operator running two machines. A bar job that runs slightly wrong is the worst money it loses, because it loses it unattended and in quantity. The difference is almost never a better machine or a better insert. It's whether the pull sequence, positions, and cutoff were programmed by somebody who had been burned by them before.

When something on a Mazak lathe isn't working

I do this for a living. If you have a bar job that scraps parts overnight, lengths that won't hold, or a puller cycle nobody wants to touch, send me the part, the machine, and the control version. I'll tell you what's possible.

Thank you, Tom Herzog