I just bought an induction retrofit lamp with E40 base and elliptical shape it came with it's RF generator but when I connected and turned power on there was a low vibrating noise coming from the coil in the bulb but no light. I tried it again but this time no noise and also no light. The brand is from China(explains it) LC Lighting with 5 years warranty but I'm not sure who died the generator or the coil. What do you guys think?

Try to zap the tube with a piezo element - does it even light ? If no then vacuum loss

Could be I din't think of it. But the generator is dead since it's not sending any vibes to the coil since the first switch on but I did not connect the earth wire of the generator because I only use L and N on my test bench could that be the cause?

I dont think so. But maybe it died from powering up without load ?

The leaflet did mention in horrible english that the supply has to be disconnected before removing the lamp. But so fast? The noise lasted just for a few seconds.

Decent ballast would be either protected and/or immune against both failed/missing lamp (when it looses vacuum,...) and coupler disconnection.

The coupler (coil coupling the HF current into the discharge) could be damaged only by heat. It could take only few seconds, but it would be accompanied by a cloud of smoke...
But if the ballast is made very cheap (selfoscillating, no protection,...), it could fail within seconds of operation without the lamp.
But I doubt it would do anything when the coupler is disconnected - it's inductance is usually an essential part of the oscillator circuit...


Generally I would rather guess faulty electronic - a fault, caused by e.g. mechanical stress (shock,... - so loose joint,...)
So open the ballast box and check the internals, mainly all the soldering. And the same for all the accessible connections on the coupler...

It's under warranty so I can't do much except hope the guy at the shop accepts my claim that I din't fix it wrong. I will post video of it when I get a new one hopefully.

Then claim the warranty.

It a good practice to connect power only after everything is connected as it should be - lamps to ballasts, lamps in place,...
The only exception are controlled troubleshooting experiments, where you know, what the disconnection and/or removal shall cause.

But then the things are usually way more sensitive for any departures from the originally planned procedures (Well, a good example is the Chernobyl disaster - how ugly could turn the otherwise quite common test, when the test procedures are not followed)...

Much of the rules are actually health/safety rather than for keeping the lamp integrity. Take for example screwing lamp in live socket - It is ok for HID and incandescent, but still not advised and you can see it on any lamp package. And its good practice to screw in the new lamp with the box on it and then take off the box, it s both safer and prevents you from leaving stains on the lamp glass

Actually its CFLs which should not be replaced under voltage, yet i did not see many that say this on the pacage.... The issue is the possibility that the ballast enclosure of the old one will shatter and expose live parts, and when screwing in the new it makes arc (charging the capacitor) which can be pulled to between the contacts in the socket and make a bang

Ash: When you replace HPS and MH lamps on a circuit with an ignitor when the power is on, your can get zapped by 3-5KV from the ignitor: much dangerous than with incandescent/MV/self-starting HPS/probestart MH lamps.
A situation when the ballast enclosure of a CFL shatters on your hands, and you reach live parts is very rare and an exception usually.
Every relamping (Regardless of lamp type) when the power is on, can be dangerous with naked hands.

Holding the lamp by its glass, away from the socket prevents this possibility. Glass lamps dont shatter from material degradation like plastic does

And i think among HID Probe Start MH is the most dangerous in this sense - Ignitor's 3-5kv pulses are relatively low energy, and are superimposed on plain 240v. but the CWA ballasts for Probe Start lamps really have OCV higher than 240v so the continuous shock current wih them would be higher as well

In real life maintenance of lighting fixtures the supply will not be off because either the switch is too far away or that there are people working under the lights so it has to be replaced with the power on. When I worked in maintenance for GLS lamps we would use paper napkins to unscrew the bulb and replacement of electronic or emergency ballast was also done on live supply. And in shopping malls HID lighting is replaced live also since they cannot afford to turn all lights off being computer controlled. The rules do not apply in real life.

Went to the shop today and found out the problem is the coil. It is shorted and damaged the new ballast as well during testing.

Is the coil non-separatable from the discharge baloon ?

The coil is detachable from the bulb vessel but I guess there was a dead short because the ballast emitted a buzz before it went pop. But the guy also checked with a meter the coil there was some reading. I think even if the vessel lost vacuum it still shouldn't pop like that should it?