6oz vs 8oz Heavy Copper PCB: What Changes When You Step Up Copper Weight

Stepping up a copper weight is a design change, not a safety margin

Most heavy copper RFQs arrive with the copper weight already chosen, and the choice almost always came from the same place: someone looked at the current, felt that the lower weight was close to the edge, and moved up one step. On the quote sheet that looks like a small decision. In the process it is not.

A copper weight is not an independent line item you can raise without touching anything else. It is the input that sets the window for almost every step that follows. When you go from 6oz to 8oz, the etch has to remove more copper, so the achievable line width and spacing floor rises and the sidewall profile gets more trapezoidal. The solder mask has to climb a taller feature, so coverage behaviour at the trace edge changes. The lamination has to fill a deeper gap between conductors, so resin demand and thickness tolerance change. Drilling and routing see a different material stack. The list of suppliers who will quote the board without a process exception gets shorter.

None of that is a reason to stay at 6oz. It is a reason to know what you are buying before the step gets written into a purchase order, because every one of those consequences lands on your board, not on the quote.

The practical failure is not choosing the wrong weight. It is choosing a weight for one reason — current headroom — and then discovering the other four or five consequences during DFM review, after the schedule is already committed.

Where the step should come from

The copper weight should be derived, not rounded up. Two inputs decide it: the functional current each conductor has to carry, and the thermal budget the board is allowed to operate in. If you have not separated those two, the step decision has nothing to stand on and the discussion defaults to intuition.

We treat the derivation itself as a separate exercise, and it is covered in heavy copper current and thermal budget to copper weight decision. This article picks up where that one ends: you have a derivation that lands somewhere between 6oz and 8oz, or that lands on 6oz with less headroom than you are comfortable with, and you want to know what the step up costs.

Two things are worth being blunt about before going further.

First, if your derivation lands clearly at 6oz, stepping to 8oz for comfort usually buys you a harder board rather than a safer one. The margin you gain in conductor cross-section is real, but you pay for it in the tightest geometry on the board, which is frequently where the actual risk lives.

Second, if your derivation lands clearly at 8oz, do not try to rescue 6oz with wider traces unless the layout genuinely has the room. Widening conductors to compensate for a lower weight is a legitimate move, but it consumes routing area, and on a dense power board the area is often not there.

The question this article answers is the middle case, and the middle case is decided by consequences, not by the current number alone.

The five consequences that move together

1. Achievable line width and spacing floor

This is the consequence that surprises buyers most often. Thicker copper requires more etching to clear the spaces between conductors, and more etching means the minimum line width and the minimum spacing both go up. The relationship is not something you can negotiate per order — it is set by the etch process for that copper weight.

The practical effect: a layout that routes comfortably at 6oz may not be manufacturable at 8oz without changing the layout. If your board has any fine-pitch component, any tight escape routing, or any area where power and signal run close together, the step up can invalidate geometry you already consider settled.

The specific achievable floors depend on the copper weight and the process route the supplier runs, and they have limits that need engineering confirmation for your particular stackup. The structural relationship is documented in heavy copper trace width and spacing limits by copper weight. What matters for the step decision is the direction: you are trading geometry freedom for conductor cross-section, and you need to know whether your layout can afford the trade before you commit.

2. Sidewall profile and effective cross-section

Etched heavy copper does not produce a rectangular conductor. The etchant attacks the top of the conductor for longer than the bottom, so the finished trace has a trapezoidal cross-section with some amount of undercut. The thicker the copper, the more pronounced this becomes.

This matters for two reasons. The cross-section you actually get is smaller than the cross-section you drew, which partially eats into the headroom the step was supposed to buy. And the amount of compensation applied to the artwork to account for this is a process decision — which means two suppliers can both deliver "8oz" and hand you conductors with meaningfully different finished widths.

The buyer-side action is not to specify the compensation. It is to require the supplier to state which dimension their declared line width refers to, and at which point in the cross-section it is measured. If that is not declared, you cannot compare two quotes and you cannot verify the delivered board.

3. Solder mask coverage over taller features

Solder mask has to cover a conductor whose height has increased. At heavy copper weights the mask has to climb the trace sidewall and bridge the gap to the next conductor, and the coverage at the top edge and the base of the conductor is where it gets thin.

The consequences are practical: edge coverage can be incomplete, the mask can pool in the gaps between conductors, and the clearance you assumed around a pad can shift. At the 6oz to 8oz step, mask behaviour is not an afterthought — it is one of the reasons the step has a cost.

The buyer-side action is to make mask expectation explicit rather than inherited. State whether you need coverage over the conductor edge, what you will accept at the base of tall conductors, and whether any area must stay clear. If you leave it to the default, you get whatever the supplier's default is, and defaults differ.

4. Drilling, lamination and routing windows

Thicker copper changes the mechanical picture. Lamination has to fill a deeper gap between conductors with resin, which affects resin content, pressed thickness and thickness tolerance. Drilling goes through more copper per hole, which affects tool wear and hole wall condition. Routing and depanelisation see a different material stack and a different burr behaviour.

None of these is a blocker. All of them consume process window. The reason it matters to a buyer is that consumed process window shows up as more DFM comments, more engineering queries, longer time to a firm quote, and in some cases a request to change your design. If your schedule has no room for a DFM loop, a step up in copper weight is a schedule risk as well as a cost item.

5. The supplier pool

This is the consequence most often discovered last. Not every shop that quotes 6oz will quote 8oz on the same stackup, and among those that do, not all of them run it as a routine process rather than an exception. As the weight goes up, your pool narrows, and a narrow pool changes your position on price, lead time and change control.

The practical move is to check the pool before you freeze the weight, not after. If stepping to 8oz drops you to a small number of suppliers who will take the board as drawn, that is information you want while you still have the option of widening traces at 6oz instead.

The order to decide in

The step decision goes wrong when it is taken first and justified afterwards. A workable order:

1. Derive the requirement. Functional current per conductor and thermal budget, separately. Do not start from a copper weight.
2. Check what the layout can absorb. If the derivation points up a step, find the tightest geometry on the board and ask whether it survives the higher weight's line and spacing floor. This is usually the step that kills the easy answer.
3. Consider the alternatives to stepping up. Wider conductors at the lower weight. Mixed copper weights so only the power layers carry the heavy copper. Moving current into a different structure entirely. Each has its own cost, but each keeps the rest of the board in an easier window.
4. Price the consequences, not just the copper. Ask for the quote to separate what changes with the weight: achievable geometry, mask treatment, any added process step, any change in lead time.
5. Check the pool. Confirm more than one supplier will take the board as drawn at the chosen weight.
6. Freeze it in a document. Not in an email thread.

What "freeze it" means in practice

A copper weight that lives only in a conversation is not specified. For the step to survive into production, the RFQ and the eventual purchase document need to carry a small set of items that are specific enough to be checked:

• Which copper weight, on which layers. If the board uses more than one weight, say which layer carries which. A single global number on a mixed board is an invitation to a wrong assumption.
• Whether the stated weight is base copper or finished copper. These are not the same thing, and the difference is exactly the kind of gap that surfaces at incoming inspection. The declaration mechanics are covered in finished copper thickness vs base copper in the purchase declaration.
• The minimum line width and spacing you actually need, and an explicit statement that the supplier has confirmed it is achievable at the declared weight on your stackup.
• How the line width will be measured and at which point in the conductor cross-section.
• Solder mask expectation at the conductor edge and in the gaps between tall conductors.
• What happens if the supplier's DFM review says the combination is not achievable — who decides whether the weight changes or the layout changes, and how that decision gets recorded.

That last item is the one most often missing, and it is the one that determines whether a DFM finding costs you a day or a schedule.

Common failure modes

Stepping up for comfort without checking the geometry. The most common one. The board gets a weight it does not need and a line and spacing floor it cannot live with.

Treating the step as a price question only. The copper cost difference is visible on the quote; the consumed process window is not. Buyers who only compare price per panel miss the part that generates DFM loops.

Declaring a weight without declaring base vs finished. Produces two quotes that are not comparable and a delivered board you cannot verify against the document.

Letting the supplier choose the compensation and never asking what it was. You then have no way to know what finished cross-section you received, which means the headroom the step was supposed to buy is unverified.

Applying one weight to the whole board. On many power designs only a few layers need heavy copper. A single global step pushes the whole stackup into a harder window for no functional gain.

Finding out about the supplier pool after the design is frozen. By then the alternatives — widening traces, mixing weights — are no longer available.

Related reading on PCB168

• heavy copper current and thermal budget to copper weight decision — how to derive the weight requirement in the first place.
• heavy copper trace width and spacing limits by copper weight — the geometry relationship behind consequence 1.
• 4oz copper PCB: standard vs heavy copper threshold — where heavy copper starts, if you are below this step.
• heavy copper PCB capability page — what we quote and what inputs we need.
• request a quote — send the derivation and the layout constraint together, not the weight alone.

FAQ

Is 8oz always safer than 6oz?

No. It gives more conductor cross-section, but it raises the minimum line width and spacing, increases sidewall undercut, makes solder mask coverage harder over taller features, and narrows the supplier pool. If the tightest geometry on your board is the actual risk, stepping up can increase total risk rather than reduce it.

Can I keep 6oz and just make the traces wider?

Often yes, and it is frequently the better move because it keeps the rest of the board in an easier process window. The constraint is routing area. On a dense power design the area to widen conductors may not exist, which is exactly when the step up becomes the right answer.

Do I have to use the same copper weight on every layer?

No. Mixed copper weight stackups let only the layers that carry current take the heavy copper. It needs to be declared layer by layer in the RFQ, and it changes the lamination picture, so it is an engineering conversation rather than a line item.

Why do two suppliers quoting the same weight deliver different line widths?

Because the etch compensation applied to the artwork is a process decision, and because "line width" can refer to different points in a trapezoidal cross-section. Require each supplier to state which dimension they are declaring and where it is measured, otherwise the two quotes are not comparable.

Should I specify base copper or finished copper?

Specify which one you mean, explicitly. Both are used in practice and they are not interchangeable. Leaving it implicit is one of the most common sources of a dispute at incoming inspection.

What should I send with the RFQ for a step decision?

The current per conductor and the thermal budget, the tightest line width and spacing in the layout, which layers need the heavy copper, your base vs finished declaration, your solder mask expectation, and whether the layout can be changed if DFM review says the combination is not achievable.

Does stepping up affect lead time?

It can, through two paths: a narrower supplier pool and a higher chance of a DFM loop before the quote is firm. Neither is a fixed quantity, so ask for lead time to be quoted against the specific weight and stackup rather than against the product family.