PCB Lead Time Planning Guide

Lead time is the sum of the time a board spends being worked on and the time it spends waiting, and in most board shops the waiting is the larger part. A panel that takes four hours of actual processing can easily occupy ten working days between the order confirmation and the shipment. Understanding where those days go is what makes production planning useful, because the levers that shorten a queue are different from the levers that shorten a process, and only one of them responds to more equipment.

What Lead Time Is Made Of

A lead time contains four parts: the engineering and data preparation time before the order is released, the queue time at each process step, the process time itself, and the final inspection, test, packing and shipping time. Engineering and data preparation are short but they are a gate, because nothing moves until the data is verified and the tooling is planned. Process time is largely fixed by the technology of the board, and it is the part that most people picture when they think about lead time.

Queues are the variable part and the part that grows without anyone deciding to let it grow. A step that is loaded to eighty percent of its capacity will hold a short queue, while the same step loaded to ninety five percent will hold a queue several times longer, and the difference is arithmetic rather than a matter of effort. This is the single most useful fact in lead time planning, and it explains why shops that are only slightly busier quote lead times that are much longer.

Why the Queue Dominates

Queue time is a consequence of variability. If every panel arrived at exactly the right moment and every process took exactly the same time, a line could run at one hundred percent utilisation with no waiting at all. Real panels arrive in batches, real processes stop for maintenance and rework, and the resulting variability has to be absorbed somewhere, which is what a queue does.

The practical implication is that lead time is controlled by how much of the variability the shop is willing to absorb and where. Batching work into large lots reduces setup frequency and increases the waiting time for everything behind it, while splitting a lot increases the number of setups and shortens the queue. The trade is between setup capacity and lead time, and it should be made deliberately with the process flow in front of the people making the decision.

Panel and Tooling Preparation

The front end of the order is where a lead time can be lost cheaply or expensively. Data verification, drill program generation, artwork preparation and tooling selection are all short tasks, but a missing layer, an ambiguous note or a stackup that cannot be built will send the order back to the customer and cost days. A checklist applied to every package at this stage is the cheapest lead time protection available.

Material availability also sits at the front end. A special laminate or a non standard thickness may have to be ordered, and that single fact can add a week to the delivery schedule whatever the shop does afterwards. Fabrication notes that record the material decisions make this visible at the quotation stage, which is early enough for the customer to choose a stock material or to accept the longer date knowingly.

Building a Delivery Schedule

A delivery schedule should be built from the capacity of the bottleneck step rather than from the nominal capacity of the shop. The bottleneck sets the rate at which orders can leave, and planning that ignores it produces schedules that are met on paper and missed in the yard every week. Working backwards from the promised date through the bottleneck is more reliable than working forwards from the order date.

The schedule also needs a small amount of protective time in front of the customer promise. That buffer is not padding; it is the mechanism that absorbs the ordinary variation of the shop without a late delivery, and it should be sized from the observed variability rather than from a habit or a negotiation. A schedule with no buffer will meet its date most of the time and fail badly in exactly the weeks when the customer can least afford it.

Reducing Cycle Time

Cycle time can be reduced by shortening the process, by removing the queue, or by reducing the variability that creates the queue. The first is usually the most expensive because it means new equipment, and the last is usually the cheapest because it means better data and better discipline. Reducing rework, stabilising the process and holding to a standard setup are all cycle time projects even though they are usually filed as quality projects.

Lot sizing is the most available lever in the middle of the flow. Smaller lots at the bottleneck reduce the waiting for every order behind them and often release capacity at the same time, because the work in progress that used to be sitting between steps is no longer occupying the line. The change costs setup time, so it should be measured against the queue it removes rather than judged by the setup figure alone.

Order Scheduling in Practice

Order scheduling in a board shop is mostly about deciding which orders may be split, which may be batched with others and which must run alone. Orders that share a material and a stackup can often travel together through the front end and separate later, which reduces setup without delaying either of them. Orders with different finishes or different tolerances usually cannot, and trying to force them together creates rework instead of savings.

The rules should be written down and applied consistently, because an informal priority system is the fastest way to destroy a schedule. Everyone agrees that a genuine emergency deserves priority, but if the definition of an emergency is negotiable then the schedule becomes a queue ordered by whoever telephones most often, and the lead time quoted to everyone else grows as a result.

Practical Rules

Quote a lead time that reflects the bottleneck and the current load, keep a small buffer in front of the promise, and measure the actual delivery performance against it with the quality data for each shipment. Review the load once a week rather than once a quarter, because a queue that is visible early can still be managed.

Attack variability before capacity, since it is cheaper and it improves the delivery schedule for every order at once. A shop that reduces rework and stabilises its setups will find that its lead time falls without any investment, and it will be able to promise dates that cost reduction work and customer relationships both benefit from keeping.

FAQ

Why is my lead time longer than the process time? Because most of a lead time is queue time. Work waits between process steps, and the wait grows quickly when a step is loaded close to its capacity.

Does splitting lots help? Often yes. Smaller lots reduce waiting for other orders and free work in progress, at the cost of more setups, so the trade should be measured.

How much buffer should a schedule carry? Enough to absorb the normal variation of the shop, sized from observed performance rather than from habit or from a negotiation with the customer.