Optoma videoprojector, no light, how ballast works

[Vicus]

I have to repair an optoma ust projector that not light up, at all. It's try 3 time that lamp error turn on. I'm focusing to the ballast, it's a philips but no schematic, nothing eveng close. On the input I have 380V, no short on mosfet or vast majority of the power silicon, There is a stepdown that produce 12V and it seems to word, there is a step down that provide voltage to the output bridge. It's an odd one because it high side driven with a nmos, it use the high side part of an fan7380 do drive it. This part seems to work because after this stepdown voltare rise breafly when it try to start. Output mosfet get warm during the 3 try. I think that the ignite part not work but I can't figure it out how it produce the voltage spyke needed. Thanks.
PS: i did try 2 lamp, the one that did work and a new compatile (it use a philips bulb).

[Vicus]

I did try a smaller ballast, for a 132W lamp, and the lamp lights up  so I'm shure it's the ballast. By the way the small 5 pin connector use a very simple interface. One optoisolator is the lamp on, another one is the eco/full power and third is on output that tell if the lamp is on or not. In my case I can't find the faulty component, I did check every passive or simple component.

[Vicus]

It turn out that things are way more complicated that what I did think. I ordered a genuine lamp ballast, I replaced and lamp did turn on correclty. Than I didn't use the projector for some month and now lamps works fine until a signal is applied. I try VGA and HDMI and as soon as the signal in sync the lamps turn off and the blue led start to blink. So I have to do more research and I found this:
https://monitor.espec.ws/files/5-pin-interface_-version-2.9_-01.02.2010_194.pdf
https://www.eevblog.com/forum/beginners/video-projector-ballast-bypass-help/?action=dlattach;attach=138036
It think that the pin SCI can be used to force the ballast to turn on the lamp but is also use to SYNC. In my case projector use a Philips ballast but I think that this protocol is an industry standard. I did verify that when you turn the projector on the DLPC4422 via a npn bjt activate the SCI input, then after some seconds the signal disappear and the lamp stay on until some siignal il apply. Menu works fine and also the ECO mode (turn down the light) works fine. I did try to force the SCI signal and the lamp stay on with a signal in but is worth mentionig that forcing the SCI signal disable the uart communication. Next step will be to try to read the serial communication and try to spot if it is the ballast that tell the projetor something that turn off the lamp or what.

[Vicus]

There is communication between the 4422 and the ballast so I think there is no hardware fault. But how can it be a software fault?

[Vicus]

I try with my DS1054z and sigrok software to decode the serial comunication. I did get some resul but it's clear that to do that a logic analazer is mandatory. With the scope you get some picture and not the entire film so some small message con be lost.

[Vicus]

Some advice for a logic analyzer or a real time capture device with decoding capability?

[Vicus]

So, at the end I did solve but still a lot of doubt. With the old ballast model EUC 260g D/V27 and the new one (same model) side to side I did measure the output coil and shure my suspicious were right. The faulty did measure 10uH the new 100uH (maybe more, I don't remember) so I did swap the coil and the old one come back to life even with a input signal to the projector. This mean that the spare part that I did buy from the official service center was not programmed. Or maybe the projhector need to be programmed but in either case it's a shame that it's not possible to swap oem parts. WHY??