B737 Transformer Rectifier failure and ILS Appraoches

[Blip]

I'm trying to get my head around the concept of isolating the TR units during an ILS approach. It applies to both the classic and the NG B737.

From the B737-800 Operations Manual:

The cross tie relay automatically opens, isolating DC Bus 1 from DC bus 2, under the following conditions:

*At glide slope capture during a flight director or autopilot ILS approach. This isolates the DC busses during approach to prevent a single failure from affecting both navigation receivers and flight control computers.
*Bus transfer switch positioned to OFF.

So before conducting an ILS approach we check that there are no TR failures. If for example we find that TR1 had failed, we would normally choose not to select APP mode if we were wanting to fly an ILS approach because we know that if we did, at Glide Slope capture, the DC Bus 1 would subsequently be isolated to only the TR1 unit and result in DC Bus 1 being unpowered causing all sorts of issues.

So what I don't understand is why isolate? What is wrong with another TR unit powering DC Bus 1 when TR 1 fails? Same thing if the failure happens during the ILS approach. TR 1 fails. TR 2 or TR 3 takes up the slack and the approach continues. What's the big deal?

If the cross tie relay stays closed, how can a "single failure" as referred to above create a problem. It seems to me that it's when the cross tie relay is closed that a single TR failure (TR 1 or TR 2+3 in the case of the B738) causes a problems.

I'm obviously missing something! :confused:

By the way if TR 1 had failed, and we wanted to fly an ILS approach, we'd simply use LOC and either V/S or VNAV anyway.

Also I can not find a list of items powered by the DC BUS 1 and DC BUS 2 for either the B734 or the B738. Does anyone know where I can find a list. It's very frustrating that such information is not included!

Thank you.

[Checkboard]

They're not talking about power supply failures, they're talking about short circuits within the supply or one of the services - with a major problem being able to crash both sides of the electrical system.

[BOAC]

By the way if TR 1 had failed, and we wanted to fly an ILS approach, we'd simply use LOC and either V/S or VNAV anyway. - something wrong with using ILS2?

[rudderrudderrat]

Hi Blip,

I'm not current on 737 any more - but in order to have Fail Active Autopilots during CATIII approaches, it's necessary to have 2 completely independent power supplies and autopilots.

Otherwise the best you can do is Fail Passive CAT III single (even with both APs engaged).

[Blip]

Sorry guys but I still don't really get it. Checkboard sort of makes sense to me but I don't know how that logic is deduced from the information in the Ops manual as I quoted it.

BOAC both ILS receivers continue to work from memory. Captain uses VHF NAV 1. F/O uses VHF NAV 2. It's a question of which pitch mode can be used during the ILS approach. If you try to use APP mode (ie pitch mode tracking the glide slope) the TR units are isolated at G/S capture and the problems begin.

rudderrudderrat, our B737's only have two autopilots so are only fail passive capable. We only just recently started flying to CAT II minimas too.
Is it possible to have only two autopilots and still be Fail Active capable? So if one autopilot fails, the single remaining auto pilot continues to fly the plane and is capable of autoland?? That doesn't sound right to me.

[Denti]

Yup, fail operational 737s only have two autopilots as well. You don't need more than that. From the autopilot side of things you could even fly OEI fail operational approaches, however as the rest of the airplane is out of redundancy it is fail passive only CAT III only (however, still with LAND 3 status).

[STBYRUD]

To my knowledge all standard 737 A/P installations are at most CAT IIIa fail passive. Options for CAT IIIb exist but are rare. But I dont understand the point you are raising, yes, if you just use VOR/LOC and V/S (or VNAV!) you will avoid the bus split - the QRH even tells you not to use the AFDS approach mode in the TR UNIT checklist. The point they are making in the OM is that if you have a catastrophic electrical problem during an approach you retain one half of your DC system (which after all encompasses all critical instruments) instead of frying it all. Does that make any sense?

[Pub User]

To my knowledge all standard 737 A/P installations are at most CAT IIIa fail passive.

No, actually fail-operational A/Ps have been fitted to new 737s for quite a while. Here's a quote from the Rockwell Collins website:

The EDFCS-730 provides Dual Channel autoland in landing conditions down to CATIIIb and rudder controlled rollout. The EDFCS-730 has been the standard autopilot flight director system for all Boeing 737NG series of aircraft since January 2003.

[BOAC]

It's a question of which pitch mode can be used during the ILS approach. - how about the one in your hand or have we forgotten how to fly?

Both receivers will NOT still be working - ILS 1 will be inop, surely?

[STBYRUD]

Pub User: Oh I'm sure that the Rockwell kit is capable, I suppose its just determined by a pin, DIP switch or a software bit whether they act as fail operational or passive - our new ships, delivered in the past months, are still fail passive.

[Denti]

You need different pieces of equipment to be fail operational capable. The new FCCs which include the thrust computer into the FCCs, the ISFD ist required as well as the autopilot rudder actuator to accomodate rudder control and automatic rollout. And of course you have an additional C/R (Cancel/Recall) button on your MFD control panel (the small one below the SPD REF knob) to recall or cancel autoland advisory messages. The MCP looks slightly different too, it has larger buttons, however that is not in itself an indicator of fail operational capability.

[STBYRUD]

Interesting stuff! Thanks denti :ok: Erm... sorry about the thread creep, Blip, do you feel like your question has been answered sufficiently?

[BillyHSAT]

Reviving an old thread—and the question I have has probably already been answered a million times but I can’t find it. The org I’m with doesn’t do auto land, but we are authorized for Cat III. The question doesn’t have anything to do with CAT III though. Back to the original question (kinda)—why can’t I accomplish an AFDS in Approach mode on the side with the good TR(s) (assuming I’ve determined whether it’s TR 1 or TRs 2 & 3 that are bad)?

[Goldenrivett]

I don’t understand how you can plan to do a manual landing from a Cat III approach. We can only plan to do an automatic landing from a Cat III approach. See Getting to grips with Cat II / III Ops (https://www.skybrary.aero/bookshelf/books/1480.pdf)

“2.2.4 CAT III OBJECTIVE The main objective of CAT Ill operations is to provide a level of safety equivalent to other operations but in the most adverse weather conditions and associated visibility. In contrast to other operations, CAT III weather minima do not provide sufficient visual references to allow a manual landing to be made. The minima only permit the pilot to decide if the aircraft will land in the touchdown zone (basically CAT III A) and to ensure safety during rollout (basically CAT III B). Therefore an automatic landing system is mandatory to perform Category III operations. Its reliability must be sufficient to control the aircraft to touchdown in CAT III A operations and through rollout to a safe taxi speed in CAT III B (and CAT III C when authorized).”

Hence the 2 autopilots must have separate power supplies.

[Denti]

A HUD can replace a fail passive autopilot system in certain installations. With that a 737 and other planes, like CRJs can be CAT IIIa approved. Only ever tried the HUD thing in the simulator, as we had fail operational 737s anyway and therefore did not have a HUD as that would be just wasted money. But apparently it works quite well.

And yes, Airbus never really bothered with that to begin with. They do offer optional HUD installations, but seeing how few airlines opt for them i guess there is no real added value in them.

[EEngr]

Checkboard

Almost exactly this scenario. But not just subsequent TR unit failures. Any short circuit in any device connected to a bus will pull the voltage on that bus low and disrupt the operation of other unfaulted loads until its (the faulted load's) circuit breaker pops.

With the DC tie closed, both DC buses are operating as a single bus, susceptible to any single fault. While cruising, it's not a big problem to wait a few seconds for a breaker to open and other connected devices to reboot and recover. Not. something you want happening while landing. Particularly when you really need that ILS working (low visibility, etc.)

[LME (GOD)]

It's a CAT 2 requirement to have certain units with independent power supplies. For normal ops the DC busses are tied together, for a CAT 2 approach the busses are isolated from each other. there is no impact to the aircraft operation as everything is still powered.

[Alpine Flyer]

Denti

Airlines probably don’t opt for them to save money….just the way they don’t buy other stuff that would enhance safety but is not mandatory. On the A220 most buyers seem to opt for them, though.

A HUD is useful during all phases of flight as it will allow you too look outside and see flight parameters at the same time. As the depiction is “conformal” you will see where in the real world a heading leads, where the aircraft is moving, whether something is above or below you, etc. They allow for very precise flying as the depiction has a much higher resolution than a typical PFD, so even when you think you’re all over the place, an observer watching the PFD will admire your rock-steady approach. Their only real drawback is a bit less visibility in dusk/night conditions which can be overcome by looking around them. You could also argue that they spoil you for HUD-less aircraft as you get used to the simplicity of having all essential parameters in a dime-sized area. In real ops opinions are divided but I’d say less pilots hate than love them. IFALPA recommends them as a safety feature for all aircraft.