Field note · Sourcing

Why Chinese mmWave Radar Is Worth a Second Look

Most integrators arrive at Chinese radar with one of two assumptions: that it is cheap and therefore fragile, or that it is the same module as the brand-name part with a different label. Both are wrong, and both cost money.

Where the modules actually come from

Open a mid-range presence sensor or a perimeter radar and there is a good chance the board inside was designed and built in China — including units sold under European and American brand names. The silicon is often Western: TI, Infineon, NXP, or increasingly Chinese fabless vendors such as Calterah. The antenna stack-up, RF layout, firmware, enclosure and production test are where the Chinese supply chain has moved in.

This matters because it changes what you are comparing. When you put a €90 European sensor next to a €25 Chinese one, you are rarely comparing two different technologies. You are comparing two different cost structures.

The price gap is mostly not a quality gap

A 60 GHz presence sensor that leaves a Chinese factory at one price and reaches a European buyer at three to four times that has not become better on the way. The difference is distribution margin, certification overhead spread over small volumes, local technical support, stock held in region, and brand.

Those are real costs, and sometimes you genuinely need what they buy: a phone number that answers in your timezone, a replacement unit in two days, a name your end client recognises. But you should pay for them deliberately, not by default.

What Chinese radar is genuinely good at

• Volume manufacturing at consistent spec. The 24 GHz and 77 GHz designs shipping today are not prototypes. They are in traffic systems, perimeter installations and building automation across Asia, the Middle East and Latin America.
• Iteration speed. A firmware change, a different connector, a narrower beam, a custom enclosure — a Chinese vendor will typically quote weeks. A European brand with a global product line will quote a quarter, if they quote at all.
• Willingness to do small custom runs. If you need 200 units with your own label and a modified mounting bracket, that conversation is possible in Shenzhen and difficult in Munich.
• Cost headroom for the rest of the system. Saving 60% on the sensor is what lets a project afford the better controller, the proper enclosure, or the field test that nobody budgets for.

Where Chinese radar still loses

This is the part that most sourcing pages leave out, and it is the part that decides whether a project succeeds.

• Long-term field reliability data is thin. A European vendor can show you five years of failure rates from installed base. Most Chinese vendors cannot, not because the product is worse but because the data was never collected. You are buying on design margin and test reports, not on a population study.
• Documentation quality varies enormously. The same company that ships a solid board may send you a datasheet with three parameters undefined and an English translation done by machine. This is where projects quietly go wrong.
• Certification support depends on the relationship. FCC, CE, and band approvals are the vendor's responsibility in principle, but getting the actual test reports and the honest answer about which bands are covered requires someone who can push internally.
• Brand backing is absent. If your end client's procurement department insists on a recognised name, a Chinese module makes your bid harder, regardless of performance.

When the answer is yes

Chinese radar tends to win when most of these are true:

• You have engineering capability to validate the part yourself.
• Volume is meaningful — roughly 100 units and up, or a recurring programme.
• Cost pressure is real and the sensor is a large share of the BOM.
• You need a variant, a custom label, or a fast change.
• Your end client buys on performance, not on brand.

When the answer is no

Walk away if any of these is decisive:

• You need a replacement unit tomorrow and cannot hold stock.
• The project has no test budget and no engineer who can characterise a sensor.
• Certification is on the critical path and you cannot absorb a delay.
• You are buying ten units. The freight, duty and validation cost will eat the saving.
• The buyer's spec sheet names a brand and the spec sheet is not negotiable.

Next

If you want the technical side of the decision, start with the range calculator and the band compliance lookup. If you want to know how I work and what I charge, that is on the next page.