Production hipot tester makes offline LED lamps malfunction

Hello,

Please assist us with our Production test equipment, because it seems to interfere with our product on the production line..

We have a HAL104 production tester for our offline, DALI dimmable, non-isolated 40W LED lamps.

During production, we connect our lamps to the mains via this HAL tester. The HAL tester then does a high voltage test between live/neutral inputs and earth. Then, with our lamps still connected to the mains via the HAL tester (there seems no reason to disconnect them), we connect a comms cable to the lamp’s DALI input (optocoupler isolated from lamp's microcontroller) and configure the lamp with operation parameters such as…
..Maximum dimming level.
..Shutdown temperature.
..etc, etc

Anyway, the configuration program in the lamps’s microcontroller receives the configuration input as a stream of bytes…..it then acknowledges by replying back with what bytes it “thinks” it just got sent. If the “sent” bytes, and “reply back” bytes, don’t correspond, then the Configuration App in the laptop reports “Fail to Configure Lamp”.

Anyway, when the lamps are connected to the mains via the HAL tester, about 20% of the time they “Fail to Configure Lamp”. Though whenever the lamps are just connected to the mains directly (not via the HAL tester), they always configure correctly.

Lamps that “Fail to Configure Lamp” on the HAL tester, always configure correctly if subsequently connected to the mains directly.

Do you have any ideas why the lamps often “Fail to Configure Lamp” when connected to the mains via the HAL tester?

(incidentally, these LED lamps are linear regulator based and don’t have a Common mode or differential mode mains input filter. ..Well, there is a small amount of capacitance to give some diff mode filtering)

Seaward HAL tester (HAL104)
https://www.tester.co.uk/seaward-cl...MIwO-qopGH2wIVxhbTCh0HmQAUEAAYAyAAEgKNJfD_BwE

DALI optocouplers used are TCLT1009..
TCLT1009 optocoupler
https://www.vishay.com/docs/83515/tclt1000.pdf

lamps are connected to the mains via the HAL tester

Click to expand...

Please clarify the test circuit.

Thanks ill find out exact details and get back
By the way, the lamps PCB doesnt have an earth connection , but the lamp pcb sits on an earthed heatsink (on a thermal pad)

Just a guess, if the tester is switching AC line voltage during a data transfer, there may be capacitive charging/discharging along the data link that corrupts the serial transfer. As FvM requests, we need a more detailed view of the test bed to assess likely failure mechanisms. The point that they work after test suggests it is in the DALI interface rather than the unit itself.

Brian.