AI issues suction system

[Pilot boy]

I have an odd AI issue and just wondering others opinions on this.  Started with a vacuum pump failure.  We replaced the pump but then the AI was growling on shut down so we replaced that as well.  The regulator was set too high for the new pump and it broke the vacuums suction gauge and I’m suspicious the new AI may have been “over pressured “The mechanic adjusted it way down now, it’s in the green arc but just barely most of the time..around 4.5 suction. The AI seems like it is always slightly right banked.  The mechanic also leveled it again.  On the first takeoff especially it seems to bank about 10 degrees right during the first 400 ft of climb.  I’m not sure if this is me not using enough rudder....or a higher p factor from prop at 2800 RPM etc.  I don’t recall the Cessnas I trained ifr in ever doing this.  1964 M20E 200 HP.  It also seems like it takes a little bit to level after maneuvering and slow flight but often shows slight right bank when level.  It is so minor it’s barely noticeable for the 98 percent majority of the time...but it subconsciously bugs me somehow.  It also showed up in about 80 degree weather in level flight.  Just a slight 5 to 8 degree bank indication.  Seems like it might be better in colder days so I’m wondering if it’s the suction levels.  We are replacing the AI again this week. All this to say does anyone notice bank indications during takeoff like this?  It’s quite obvious and I wouldn’t want to penetrate a low overcast on takeoff.

Edited November 15, 2021 by Pilot boy

[takair]

I would look for kinked hoses. That might result in high vacuum and low flow. The AI relies on flow, more than vacuum. Kinked hose would explain why you initially got the high vac. Also, why are you seeing 2800RPM?   Red line is typically 2700.

[Guest]

Reversed hoses on the gyro can cause weird indications.

Clarence 

[SSimpson77]

I’m wondering too about the 2800 rpm? I had issues with my vacuum AI, went with G5 and never worry about it again. 

[RobertGary1]

On 11/15/2021 at 6:06 PM, SSimpson77 said:

I’m wondering too about the 2800 rpm? I had issues with my vacuum AI, went with G5 and never worry about it again. 

But you need some diversity backup since dual AHRS failures do exist. The vacuum system is an inexpensive backup. 

[SSimpson77]

On 11/18/2021 at 8:43 AM, RobertGary1 said:

But you need some diversity backup since dual AHRS failures do exist. The vacuum system is an inexpensive backup. 

I guess it’s a personal choice. You don’t “need” diversity backup. It’s nice, but I only have a lot of one thing on my plane… one Johnson bar, one oil pump, one engine, two propellers but those come in pairs. With the MTBF of 13,000 hours for G5 ahrs and separate hsi/ai I figure I’m pretty comfortable flying behind that. But everyone decides what’s more important to themselves. 
If you have a plan for any kind of emergency, you are better off than relying on redundancy. My plan for AI failure(which has happened with gyro AI, but luckily while VMC) is not immediately switch DG to attitude indicator, but to cross check and utilize performance instruments to confirm…alt,airspeed,Vsi, TC. 
Off story- I own some guns and one day while at range one of my handguns “stovepiped” the casing. It was probably my fault, but I will never grab that gun if my life depended on it, even though I’ve sent some 1,000’s of rounds through it. Because that gyro AI went tumbling on me… I don’t trust it but again that’s just me. And no, it’s not for sale but the AI is.

[RobertGary1]

1 minute ago, SSimpson77 said:

I guess it’s a personal choice. You don’t “need” diversity backup. It’s nice, but I only have a lot of one thing on my plane… one Johnson bar, one oil pump, one engine, two propellers but those come in pairs. With the MTBF of 13,000 hours for G5 ahrs and separate hsi/ai I figure I’m pretty comfortable flying behind that. But everyone decides what’s more important to themselves. 
If you have a plan for any kind of emergency, you are better off than relying on redundancy. My plan for AI failure(which has happened with gyro AI, but luckily while VMC) is not immediately switch DG to attitude indicator, but to cross check and utilize performance instruments to confirm…alt,airspeed,Vsi, TC. 
Off story- I own some guns and one day while at range one of my handguns “stovepiped” the casing. It was probably my fault, but I will never grab that gun if my life depended on it, even though I’ve sent some 1,000’s of rounds through it. Because that gyro AI went tumbling on me… I don’t trust it but again that’s just me. And no, it’s not for sale but the AI is.

I guess my perspective is different having had a dual AHRS failure with Garmin. 
All the manufacturers have had software updates to fix dual AHRS failure. Software has bugs 

[SSimpson77]

I respect that for sure. That’s your stovepipe! Really hurts to spend so much on equipment and it fails. 

[SSimpson77]

4 minutes ago, RobertGary1 said:

I guess my perspective is different having had a dual AHRS failure with Garmin. 
All the manufacturers have had software updates to fix dual AHRS failure. Software has bugs 

Does Garmin send a investigation team out or get with you and see what happened? Did you have to file any paperwork with FAA when something like that happens? Just curious, I would hope your incident helps all users in the future.

[RobertGary1]

Just now, SSimpson77 said:

Does Garmin send a investigation team out or get with you and see what happened? Did you have to file any paperwork with FAA when something like that happens? Just curious, I would hope your incident helps all users in the future.

Their director called me and arranged to fly someone out to see me. Then he called back and said they’d arranged a live link at a local shop instead. So I took it there and they were able to access the plane from their office. They kept the plane for a week. Never heard anything back but a SB came out a month later to fix that exact problem. 

[SSimpson77]

1 minute ago, RobertGary1 said:

Their director called me and arranged to fly someone out to see me. Then he called back and said they’d arranged a live link at a local shop instead. So I took it there and they were able to access the plane from their office. They kept the plane for a week. Never heard anything back but a SB came out a month later to fix that exact problem. 

And they left you with a big “ sorry for your trouble, here’s a refund check” right haha

[Pilot boy]

On 11/15/2021 at 5:29 AM, takair said:

I would look for kinked hoses. That might result in high vacuum and low flow. The AI relies on flow, more than vacuum. Kinked hose would explain why you initially got the high vac. Also, why are you seeing 2800RPM?   Red line is typically 2700.

I’ll have to recheck that RPM.  I think in this E the red line is 2800 or 2750.

[Pilot boy]

On 11/15/2021 at 5:29 AM, takair said:

I would look for kinked hoses. That might result in high vacuum and low flow. The AI relies on flow, more than vacuum. Kinked hose would explain why you initially got the high vac. Also, why are you seeing 2800RPM?   Red line is typically 2700.

Mechanic replaced the AI with a new one this week as we just asssumed manufacturer defect as a first step...I will test it and post back soon

[takair]

1 hour ago, Pilot boy said:

I’ll have to recheck that RPM.  I think in this E the red line is 2800 or 2750.

Mine is 2700 which is per the TCDS.  I didn’t think other years were different.  Might be worth a check…

[Warren Dunes]

On 11/20/2021 at 1:47 PM, SSimpson77 said:

One day while at range one of my handguns “stovepiped” the casing. It was probably my fault, but I will never grab that gun if my life depended on it, even though I’ve sent some 1,000’s of rounds through it...

If you have sent 1000 rounds downrange through the same firearm without incident. And semi-clean it stovepipes once, I'd tend to trust the firearm more than the ammo.

Edited December 3, 2021 by Warren Dunes

[Smiles201]

Lesson I learned the hard way. After vacuum pump failure you have to clean the lines out of material from the broken pump vanes, vacuum out the debris with shop vac. Otherwise new AI will be damaged by the carbon bits and dust.

[Pilot boy]

On 12/4/2021 at 1:18 PM, Smiles201 said:

Lesson I learned the hard way. After vacuum pump failure you have to clean the lines out of material from the broken pump vanes, vacuum out the debris with shop vac. Otherwise new AI will be damaged by the carbon bits and dust.

Well it seems to be working now.  The mechanic replaced the AI a 2nd time and adjusted the regulator upwards a bit more.  I think he finally tested it in the ground and could see it wasn’t working quite right.  He just didn’t want to readjust it cuz I guess it’s a pain to get to.  Seems resolved....I will have to mention this to him to see if he cleaned it that first time.  For now, it works so hopefully the first AI took all the damage.  Though I think the pump had not quite fully failed when I brought it in.  He said another hour or two of operation and it would have broken apart so maybe it was just not adjusting the regulator right the first time that caused issues.

art

[Pilot boy]

On 11/15/2021 at 6:29 AM, takair said:

I would look for kinked hoses. That might result in high vacuum and low flow. The AI relies on flow, more than vacuum. Kinked hose would explain why you initially got the high vac. Also, why are you seeing 2800RPM?   Red line is typically 2700.

I finally got back up in the E once the AI is now repaired and you were right.  It is red lined at 2700 RPM and that’s what I’m seeing in my E.  I just had recalled it wrong.  I think the MP will go a little higher than 27 in, like 28, but RPM caps at 2700.

[carusoam]

Summary…

1) Stuff fails…

2) Describing the failure takes extra effort…

3) Diversity for AI equipment is incredibly important for people that fly in IMC… not so much in VMC….

4) We have a lot of single failure points on our planes… that is a different argument to having a single point of failure for an AI in IMC…

 

5) Ask for help when having difficulty describing your failure…

6) The more details you supply, the better the responses become….

7) More details… more things to look out for when getting it fixed….

8) Unfortunately… all AI manufacturers, including the back up systems have had terrible failures…. Found in flight…

9) Nothing worse than believing one system is so good it won’t fail… then finding it failing in flight….

10) Another failure found… two identical looking displays from one manufacturer…. The previous owner didn’t have the second display to be a back up to the first…. The second owner thought it was normal that one would back up the other…. (Near suicidal error) later finding there was no back-up AI hardware to rely on….

11) Many of the issues originally found… are normal for vac instruments that are failing…. Not normal for instruments leaving an avionics shop…

12) If you don’t know your vac level… get a vac gauge instrument… Some Mooneys were mistakenly sent out without one… having one installed is a low price solution to keep from being surprised one day…

13) broken vac pumps are normal challenges for instrument shops… getting them replaced isn’t really difficult…. But not knowing if carbon was possibly sucked back into the instruments…. Consider your instrument shop carefully…. The system was designed to prevent this from happening…. Did something simple get skipped?

14) It really helps to know your vac system… the drawings are available, often in the POH….

 

Know what you have… test what you have in flight…. Be sure it works as expected….

Even the best equipment has had bad data, or bad software updates rendering them dangerous at one point or another…

 

Critical thinking for flying in IMC often….

There used to be a time when a TC was believed to be a great back-up device…. Until they failed over and over again….

The internet has provided lots of factual info regarding device failures…. Very little trust for any one device is left for flying in IMC….

 

Go MS!

Best regards,

-a-