Brazed Plate Heat Exchanger Helium Leak Testing Machine 1.0E-9 Mbar.L/S Two Chambers For HVAC Chiller Production

If you braze plate heat exchangers for chillers or heat pumps, this is the machine that decides whether a weeping core leaves your plant. Two chambers, a gross-leak pressure stage in front of the sensitive stage, and a recovery loop that keeps the running cost honest — sized to your core range, not to a catalogue envelope.

The failure you are actually paying for

A brazed core has no gasket and no bolts. Once it leaves the furnace, the only thing standing between your customer and a refrigerant leak is the braze itself. That is why the failure mode is so expensive: nobody can repair it in the field. The chiller comes apart, the core is replaced, the refrigerant is recovered and recharged, and the labour is yours whether the warranty says so or not.

We have been asked to look at enough of these claims to know where they come from. Four places, almost every time:

• Port collars. The braze has the furthest to flow around the four ports. A void there survives the factory test, opens under pressure cycling, and shows up in the second heating season.
• Edge bars. A joint that looks continuous but has a channel two or three plates in. You cannot see it, and a pressure test at 1 bar will not find it.
• Plate erosion at the inlet. High velocity thins the first plate. The leak starts as a weep measured in 10^-5 mbar·L/s, which is exactly the band a cheap test cannot resolve.
• Handling cracks. A dropped stack often looks fine and fails at the collar.

Every one of those sits below what an air decay test can resolve in a cycle time your line will accept. That is the whole argument for helium on this component — not sensitivity for its own sake, but sensitivity where the failures actually are.

Design choices we made, and the reasons

A gross-leak stage in front of the fine stage

Helium enters at up to 3.0 MPa and the pressure is held for a few seconds before the analyser is even asked to look. A core with an open void is rejected there, in seconds, and the sensitive measurement only ever sees cores that have a chance of passing. It sounds trivial. It is the difference between a 45-second cycle and a 3-minute one.

Two chambers out of phase

One chamber measures while the other is loaded. The operator never stands idle waiting for a pump-down, which is what makes 100% inspection affordable with one person. A single chamber system with the same sensitivity would need two operators to hit the same output.

The core is never evacuated from the inside

The vacuum is drawn in the chamber, around the core. That matters on a thin plate stack: pulling vacuum through the channels can deform plates and create the very leak you are testing for. Chamber-side vacuum, helium inside the core.

Recovery built in, not bolted on

Helium leaves the core and the chamber at the end of every cycle. Without recovery you are venting the most expensive part of the test to atmosphere several hundred times a day. The loop is integrated, with the purification stage after the buffer tank, so the gas that goes back into the core is within specification.

Port connectors that survive production

The part of a leak test bench that fails first is almost never the analyser — it is the connector that clamps onto the port four hundred times a shift. Ours are built around your port geometry with replaceable seals, and the seal part numbers are on the drawing so you can stock them.

How a cycle runs

• Load and clamp — core into the chamber, two pneumatic connectors onto the refrigerant-side ports, barcode scan for traceability.
• Evacuate — roots and rotary pump set pull the chamber down; the recipe waits for a stable base pressure, not a fixed timer.
• Gross leak — helium charge to the recipe pressure, hold, watch for a drop. Fail here and the cycle ends.
• Fine leak — analyser integrates the chamber signal over a fixed window; the result is converted to a leak rate and compared with the limit.
• Recover and release — helium pulled back through the recovery loop, chamber vented, core released with a verdict and a stored record.

Technical parameters

What it finds

Setting the acceptance limit — and proving it

Most arguments between a machine supplier and a customer happen here, so this part is worth getting right before the machine arrives.

• Start from the consequence. Take the refrigerant charge, the acceptable annual loss and the service life, and work backwards to a rate. For R32 and R410A chiller cores, most of our customers land between 1×10^-6 and 1×10^-7 mbar·L/s.
• Then prove repeatability. Run one known-good core ten times and one core with a calibrated leak ten times. If the spread overlaps the limit, the limit is wrong or the test time is too short — this is a ten-minute exercise that saves months of argument.
• Set the reject line with margin. We normally set the reject threshold at roughly one third of the specified rate, so a core that is genuinely marginal is caught before your customer catches it.
• Write the method into the drawing. Helium, chamber method, charge pressure, dwell time, ambient band. A criterion without a method is not a criterion.

Compared with the alternatives

A brazed core cannot be immersion tested and then dried. Water sits in the channels and destroys the thing you were protecting, which is why this industry moved to helium in the first place. Once you accept that, the question is only whether the machine can keep up with your line — and that is a question about cycle time and helium cost, not about sensitivity.

Delivery, installation and the first month

• Factory acceptance test on your own cores before shipment, with video and the calibration certificate.
• Port connector tooling built to your drawings, drawings handed over with the machine.
• On-site commissioning: connector prove-out, leak-rate correlation against a calibrated orifice, operator training.
• Ten-run repeatability study on a good core and a calibrated-leak core, signed off with your quality engineer.
• Recipe set up for every core in your range, selected by barcode at the station.
• 24-month warranty on mechanics, pneumatics and control; vacuum pumps per manufacturer terms.
• Spares held in Guangzhou: chamber seals, connector seals, helium valves, pump service kits — 48-hour dispatch.
• Annual calibration with a traceable certificate, plus remote support over the network for recipe tuning.

Frequently asked questions

About Guangzhou Hejin Automation Control Technology

We have been building leak detection and electrical test equipment in Guangzhou since 2003, with 50–100 employees and roughly half of our output exported to Europe, North America and Southeast Asia. Our engineers commission vacuum chamber systems, tracer gas hoods and air decay testers, so the method recommendation you get comes from production data rather than from a catalogue.

Every system is assembled, pump-down tested against a calibrated leak and recorded under its serial number before shipment. Years later, when you want to know whether a new result is comparable to the original acceptance data, we can still answer.