Mass PCB Production: Quality or Price? How to Decide

It is one of those questions that keeps coming up in engineering forums, and it always ends in an argument: when you move from prototype to building thousands of boards, do you pay for quality or squeeze the price?

There are two camps, and both have solid arguments. But deep down, this is not an engineering question. It is an accounting one.

The dilemma: from cheap prototypes to mass production

The usual framing goes like this. Online prototype services offer incredibly cheap PCBs, and for small test runs they are honestly great. But once you scale up to production, the requirements change: batch-to-batch consistency may fall short of what a demanding product needs, and some complex processes are either unavailable or get much more expensive.

At the other end are the premium manufacturers: tighter tolerances, cleaner vias, more stable surface finishes. In exchange, each board costs significantly more.

And in between there is a huge range. Some large, well-equipped factories have advanced process control and turn out extremely reliable boards. Others focus purely on price, and it shows: warped boards, misaligned solder mask, uneven plating.

Some people question the premise

Not everyone buys the idea that "cheap manufacturers get worse at volume". The objection is fair: why would a manufacturer change its process just because you order more boards?

And it is partly right. What changes when you move to mass production is usually not the manufacturer's process, but what you get to see. A defect that affects 0.5% of boards will probably not show up in an order of 50 prototypes. In an order of 50,000, that is 250 boards.

The case for quality: the PCB is the cheapest part of the board

The argument you hear most from those who would pay more is simple: the bare PCB is almost always a small part of the total cost. On top of it go the components and the assembly. If a bad board forces you to rework it, reflow it again or scrap the unit, what you lose is not the PCB: it is everything soldered onto it.

That is what you see on the factory floor. What you do not see is worse: the board that passes inspection and fails at the customer's site. At that point the cost is no longer measured in cents, but in warranty claims, in reputation and, in extreme cases, in lives.

The quality camp puts all of this on the scale:

• Rework and scrap: technician hours and units you cannot sell.
• Latent failures: defects that do not show up in final test and appear months later.
• Field and application risk: a failed remote control is not the same as a failed industrial controller.
• Warranty, support and reputation: every return costs money, and every bad review costs sales.

That is why, for critical, high-end or long-life products, most people are clear: quality. Chargers and battery packs, cameras, drones, inspection and surveillance equipment, automation controls. Products where a failure is expensive or dangerous, or where the brand name is everything. And if the product has a battery, regulation tightens even further: we cover it in CE marking when your product has a battery.

The case for price: sometimes cheap really is cheaper

The other camp is not defending sloppy work. It is arguing that quality has a price too, and paying it does not always make sense.

Its strongest argument: for cheap, short-lived products, you can absorb a 20% reject rate and still come out ahead. Selfie sticks, remote controls, toys. Throwaway products that nobody expects to last ten years, where every cent of unit cost counts.

The condition is not to trust blindly. If you go cheap, make up for it with your own control: test every board on your line, or a sample from every batch, and rework whatever can be recovered. What you can do in-house decides how much risk you can take on.

In crowded markets the reasoning goes one step further. If you compete on price against many others and are unlikely to win a large share, it can make sense to accept an expected failure rate with cheap boards, as long as you can detect and fix the failures. It costs you brand image, yes. But how much that matters depends on what you are making.

The three questions the price camp asks:

• How much margin does the product have?
• Can I test and repair in-house, or do I depend on the board arriving perfect?
• If it fails, is it a return worth a few euros or a serious problem?

One caveat: making the PCB cheaper is not the same as making the design cheaper. Removing components to save money is a different conversation, and it usually ends worse. We saw it when we compared the schematic of an original charger with a fake one.

The numbers that settle the argument

Both camps end up in the same place: the number that matters is not the price of the PCB, but the cost of each good unit.

Two illustrative examples, with round figures:

A product with expensive electronics

• Cheap PCB: €1.00 + €24.00 in components and assembly = €25.00. At 95% yield, each good unit costs €26.32.
• Quality PCB: €1.80 + €24.00 = €25.80. At 99.5% yield, €25.93.

The board that costs 80% more ends up almost 40 cents cheaper per unit. And that is without counting a single field failure.

A low-cost toy

• Cheap PCB: €0.20 + €1.30 = €1.50. At 90% yield, €1.67 per good unit.
• Quality PCB: €0.45 + €1.30 = €1.75. At 99% yield, €1.77.

Here the cheap board wins, even while scrapping one in ten. Over a hundred thousand units, that is around €10,000.

That is why the question has no universal answer. It is more a management and accounting decision than a purely engineering one.

What everyone agrees on: yield is everything

Among people who have actually taken a product from prototype to mass production, there is a consensus: yield is everything.

When you go into production you have yield and reliability targets. They may be more or less demanding depending on the product, but they are written down, and you pick the supplier that meets them for the specific complexity of your board. Not the cheapest or the most expensive: the one that meets them.

And how do you know which one does? With data, not catalogues:

• Pre-production runs: the runs before full production tell you the real yield you get with that manufacturer and that design.
• Reliability testing: done properly, it lets you estimate how the boards will behave over the long term.
• A supplier you trust: a good manufacturer tells you when your design will not come out consistently in their process, and flags the problem areas before production.

So, quality or price?

• Price, if the product is cheap and short-lived, a failure has minor consequences, and you can test and rework in-house.
• Quality, if the product is critical, expensive or built to last for years, if it carries expensive components, or if your brand lives on its reliability.
• The numbers, for everything in between, which is almost everything. Work out the cost per good unit with data from a pre-production run, and decide with figures.

And one trap worth avoiding: choosing a manufacturer based only on the prototype price. What the test board cost you says very little about what each good unit will cost when you are building fifty thousand.

How we approach it at RobotUNO

When a product moves from prototype to mass production, the manufacturer is chosen from the product's numbers, not from the PCB price list:

• We review the design for manufacturing before requesting quotes, so the board yields well in any serious factory.
• We define what has to be specified in writing: acceptance class, tolerances, finish and testing.
• We validate with a pre-production run before committing to volume, and measure the real yield.
• We work out the cost per good unit from that data, not from catalogue prices.

It is part of our industrialization and mass manufacturing work and of the cost study for taking a prototype to market, and it starts with the PCB design itself.

Frequently asked questions

Can a cheap prototype manufacturer handle mass production?

It can. It depends on your board's complexity and on the yield and reliability you need. What you should not do is assume it because the prototypes came out fine: check it with a pre-production run.

How much of an electronic product's cost is the PCB?

In most products, little compared with components and assembly. That is why saving on the PCB can backfire: a defective board takes everything soldered onto it down with it.

What is manufacturing yield?

The percentage of units that come off the line good. At 95% yield, 5 out of every 100 units built have to be reworked or scrapped, and the 95 good ones carry their cost.

How do I ask a PCB manufacturer for quality?

With specific requirements: the IPC acceptance class (usually Class 2, or Class 3 for high reliability), the critical tolerances, the surface finish and electrical testing of the bare boards. "Good quality" cannot be verified; an IPC class can.

Should I switch manufacturers when moving to mass production?

Not by default. If your prototype manufacturer meets your yield targets in the pre-production run, switching only adds risk. If it does not, then yes, but the new one also has to pass its own pre-production run.