B737 APU Start Fault

[ImbracableCrunk]

For the first time last night, I had an APU fail to start on a B73NG. We had a FAULT light on the overhead. This was after it had spun up and the EGT rolled back. It started after a second attempt.

When we chocked in, the CA spoke with the engineer/mechanic and he said that you should hold the APU switch to ON for 3 seconds before releasing.

I smiled and said, "Oh, I see."

That's the second time I've heard this, the first time being before entering ETOPS at cruise and the CA told me to move the switch slowly since we were so high.

Now, before I add any more emoticons, could there be any truth to this theory? To me, it's not like you're priming a recip engine. You're just sending a signal to a computer. I suppose there could be a solenoid/RCCB somewhere that needs just a second to latch, but I doubt it. If the APU tried to start, then it has nothing to do with the switch.

[Bobcanfixit]

The fault light and reason would need to be determined via a bite check. Possible overspeed condition, oil pressure etc (if it even did light off?). If in flt common for cold soak, no light off and/or overtemp at alt, on gnd less common, but does happen.

The switch has nothing to do with it (if it in fact started the sequence?). The reason you hold it in "ön" for a couple of seconds is to allow the FADEC to power up (it needs to be powered to allow the start sequence, if not going to start will do nothing).

Select "on" and watch the EGT gauge, when it flicks, you have APU FADEC power.

[BOAC]

I was always taught '1000 - 2000' for the APU from -200 days, ground or air. For ETOPS, why not get the thing lit on the ground and leave it running? At least you'd have a man with a 'Manchester Screwdriver' on hand to fix it.

[Sciolistes]

It is definitely the case on the Classic that if you are too quick in allowing the start switch to return from "Start" to "On" it can sometimes fail to start.

[CaptainSandL]

I was also taught “1000-2000” and, whilst I don’t have a 100% start rate, my unscientific gut feeling is that it gives a slightly higher success rate than those who don’t pause.

I have no doubt that the APU start switch is simply a switch and the pause can’t be shown to help on paper, in much the same way that the engine start levers are switches and should not be used like a mixture control. My theory for its success is that you sometimes see people move the APU start switch very quickly, by jabbing an index finger at it. The end of the switch rotates and although it has a rubber end your finger can deflect off meaning that you may not fully go to the start position. This may mean either that the start position was never achieved or that it was so fast it confused the electrics. Moral of the story is slow down and make any switch selections slowly and carefully.

The 737NG APU is a far more reliable starter than earlier generations but they have also offered some Mods which your aircraft may not have fitted. Most notably these were a new start converter unit and start power unit from 2001 and also ECU software is now at version 4.

S&L

[ImbracableCrunk]

BOAC wrote:

For ETOPS, why not get the thing lit on the ground and leave it running? At least you'd have a man with a 'Manchester Screwdriver' on hand to fix it.

That works well if you're entering ETOPS not long after take-off. I have a sector in western China where it's a few hours until ETOPS and policy is to leave APU on only if ETOPS EEP is within 1 hour of departure.

[BOAC]

Hell, its only a few kg an hour as against an abandoned trip? You will also avoid any cold soak problems.