RCD test oddity

[radiohead]

Testing 2 RCD's in a starbreaker dual board today. MCB's off and testing to the busbar both RCD's @5x, both RCD's gave exactly 4.8ms. Far too quick...and to get the same result on both RCD's didn't seem right. So I tested at two different sockets, both RCD's gave 4.8ms again despite connected loads including a fridge compressor running which would normally dramatically increase the recorded trip time. Suspecting a meter malfunction (ETHOS 8400...dont buy one!) I went back to the office and tried it on our RCD socket which we use to check the meters consistency, 18.6ms bang on the usual reading....so why did the crabtree units produce such clearly inaccurate readings?
Any suggestions?

 

[pc1966]

Any chance of getting the starbreakers tested with another MFT? Or even the office ones tested on another MFT to see if the 18.6ms is trustworthy as well.

 

[radiohead]

Any chance of getting the starbreakers tested with another MFT? Or even the office ones tested on another MFT to see if the 18.6ms is trustworthy as well.

Will try that monday with a borrowed meter.