We just got a new helium leak detector, a Varian 938, about 15 years old, shown at right. Our old detector, a Veeco, had not worked well for years, and then we could no longer get it fixed. For a while we outsourced our leak checking, and on rush jobs it didn’t always get done. Now we can check everything in-house at much better sensitivity than before.
It used to be that helium leak checking services were common locally, in Detroit, but we lost a lot of manufacturing and our quality checkers left too, for Ohio, then abroad. Companies that used to do He leak checking switched to fluorescent die checking, which isn’t as good, and then stopped checking altogether. But small cracks at welds are not uncommon, particularly with laser welding of stainless steel. They form as the metal cools, and are too small to see. They show up in helium leakage through the micro-cracks, and if not detected and corrected, these cracks will grow and destroy a pipe or system over time. For my products, hydrogen purifiers, the micro cracks affect hydrogen purity. Any crack that helium can get through will allow impurities to diffuse into our output hydrogen. Because of the problems of micro-leaks and durability, it used to be obligatory in the nuclear navy to helium leak check every part.
The picture above was taken while calibrating our new helium leak detector. I’d bought some calibrated sources, in this case, one certified to bleed at 3.6×10-8 cc/sec, which is equivalent to 2.1 x 10-9 slpm. If a leak of that rate would appear in one of my 1 slpm purifiers, that would imply a delivered hydrogen purity of 99.999998%, and that’s generally the purity limit I certify. This is also a common leak standard in the industry, but my new detector is better than that standard. We can test down to 0.5×10-9cc/s . A 1 slpm purifier with that leak rate will deliver hydrogen at 99.9999999+ purity, more than nine-nines, as they say. On the calibration day, above, I tested three new purifiers, and found they were all perfect to the higher standard, ≤ 0.5×10-9cc/s. I’m so happy with this, I thinking of going into free-lance, helium leak checking. There’s no one local who does it.
In case you want to know, I bought this detector on eBay but found it didn’t work. I tried cleaning key parts, replaced the pump oil, nothing helped. Then bought another leak detector, with a spare filament, but without a pump or a mass-spec unit. It included the standard leaks that I now use to calibrate. That one didn’t work in the same way that the original one didn’t; we never reached pressure, and the filament never came on. Eventually, I came to believe that the vacuum pump was a fault. I sent it to a great pump repair shop in Greentown, PA, and discovered they also repair helium leak detectors. I hope to visit them. They replaced the pump and spiffed up my detector at a decent prices ($1500 as I recall). Now, it works like a charm.
Robert Buxbaum, July 1, 2026
