View Full Version : Changing bulbs on 747 FMC CDU
blue up
25th Jul 2019, 11:49
Changing bulbs on 747 FMC CDU
Like it says, I have one single bulb out on a Sim CDU and it is the 'Execute' light, so I need to change it. When removing the facia panel I can see there is a circa 32 place board connection and wording on it along the lines of "Don't open this, or the world will end". Before I incur a huge bill for it, has anyone actual experience of changing the bulbs? Real 747-400 unit.
Thanks
Rob
blue up
26th Jul 2019, 17:10
Turned out to be quite easy. Tiny little bulbs about 1/4 the size of a grain of rice. Bulbs arrived with "Midget Flange Bulbs".
Top tip...don't Google "Midget Flange". Not all pics relate to light bulbs.. :eek:
ampclamp
27th Jul 2019, 19:45
For future reference the CDU will have CMM (component maintenance manual) that should tell you how to change the lamps and have its own IPC.
blue up
28th Jul 2019, 17:07
Thanks.
The local MRO didn't have access to one so I had to get 'creative'. Seems to work alright, though.
MurphyWasRight
7th Aug 2019, 23:52
Turned out to be quite easy. Tiny little bulbs about 1/4 the size of a grain of rice. Bulbs arrived with "Midget Flange Bulbs".
Top tip...don't Google "Midget Flange". Not all pics relate to light bulbs.. :eek:
Reminded of a long ago tip from a DEC (Digital Equipment Company) service person back in the day when some computers still had front panels with blinking lights and switches.
When replacing bulb(s) before reinstalling the board give it a few solid raps on a table, this will encourage "on the edge" filaments to let go so they can be replaced before the dozens of tiny screws holding it all together are back in place.
Nothing worse than spending an hour taking the beast apart only to have one or more new failures when it is all back together.
I was an FE on those beasties. One client was a chem lab in a steel plant. High-stakes PDP8. Only got it for one hour once a week. So many of those bulbs went out, we decided to replace them all with LEDs. Got the machine and removed the lamps, soldered in the LEDs, powered up just before the end of the hour and started s****ti*g bricks as they all lit up permanently! They actually used the lamps to read values out! Problem was the lamps all had keep-alive resistors in series so the lamps never got completely cool to stop them being shocked when they turned on. The keep-alive resistor was enough to light the LEDs to full brightness. Got the wire sheers and quickly cut one side of the resistors, problem solved, just as the lab techs walked in to re-start the processing!
MurphyWasRight
8th Aug 2019, 13:10
I was an FE on those beasties. One client was a chem lab in a steel plant. High-stakes PDP8. Only got it for one hour once a week. So many of those bulbs went out, we decided to replace them all with LEDs. Got the machine and removed the lamps, soldered in the LEDs, powered up just before the end of the hour and started s****ti*g bricks as they all lit up permanently! They actually used the lamps to read values out! Problem was the lamps all had keep-alive resistors in series so the lamps never got completely cool to stop them being shocked when they turned on. The keep-alive resistor was enough to light the LEDs to full brightness. Got the wire sheers and quickly cut one side of the resistors, problem solved, just as the lab techs walked in to re-start the processing!
Classic 'what could go wrong' tale.
Through hole PCBs do have their advantages if the keep alive resistors were surface mount, especially in an array, you would have been really screwed.
I think it was the PDP11-20 that had a rom board option with a diode for every bit, factory programming was by selective insertion, field program of a 'blank' board required snipping diodes to flip the bit.