Fuel servo issues

[Jcarne]

Larry have you been flying just fine since discovering this? I had the exact same symptom of mixture control not releasing pressure once pressure builds yesterday at a fuel stop. 
 

I also talked to Avstar and a tech told me that what he thinks is happening is that pressure is building behind the mixture plate to the point where is pushes the mixture shaft out a little and compresses the spring a bit which will disengage the shaft from the mixture plate until pressure subsides. 

[Ragsf15e]

56 minutes ago, Jcarne said:

Larry have you been flying just fine since discovering this? I had the exact same symptom of mixture control not releasing pressure once pressure builds yesterday at a fuel stop. 
 

I also talked to Avstar and a tech told me that what he thinks is happening is that pressure is building behind the mixture plate to the point where is pushes the mixture shaft out a little and compresses the spring a bit which will disengage the shaft from the mixture plate until pressure subsides. 

Just for fun, have you guys tried pushing in both the mixture AND the throttle at the same time to relieve the pressure?  I know much less than @PT20J about the fuel servo, but I have noticed that when I prime, if my throttle is all the way back (like the 800rpm mentioned above), my pressure will build and there’s not much flow even with mixture rich.  Pushing the throttle in  say halfway and pressure drops, flow is good.  
 

Just be advised… if you “relieve the pressure” at a fuel stop and then try to start, there’s a good chance you’ll be flooded as the fuel is going to flow into the cylinders.

[PT20J]

5 hours ago, Jcarne said:

Larry have you been flying just fine since discovering this? I had the exact same symptom of mixture control not releasing pressure once pressure builds yesterday at a fuel stop. 
 

I also talked to Avstar and a tech told me that what he thinks is happening is that pressure is building behind the mixture plate to the point where is pushes the mixture shaft out a little and compresses the spring a bit which will disengage the shaft from the mixture plate until pressure subsides. 

When you shut down a hot engine with the mixture control, fuel is trapped in the line from the fuel pump to the servo, and also the line from the servo to the fuel pressure transducer. I my M20J, that's about four feet of line. With no cooling airflow, the temperature in the engine compartment rises rapidly and it will boil the fuel in these lines forming vapor and increasing the pressure in the lines. My G3X EIS fuel pressure transducer has a upper limit of 70 psi and the pressure will exceed that. 

In the original RSA design, the mixture plate did not form a tight seal when in ICO. There is a leakage spec and some leakage is expected and the only requirement is that the flow is reduced enough to shut down the engine. When AvStar copied the RSA design, it appears that it "improved" the sealing of the mixture plate at ICO. With no leakage, there is no way for the pressure to bleed off and I think that is why the AvStar servos see a high fuel pressure after shutdown. Since this pressure is blocked from the regulator (the part of the servo with the diaphragms) by the mixture plate, I don't see how it can hurt anything. 

The airflow through the servo is too low at idle to meter fuel accurately which would cause the engine to run rich. To solve this, there is an idle valve mechanically connected to the throttle that blocks off most of the main metering jet during closed (or very low) throttle positions. The length of the linkage between the throttle and the idle valve is what you adjust to set the idle mixture. When priming, you should open the throttle somewhat to open the idle valve to allow more fuel to flow to the injectors. 

I don't know what would cause the pressure not to release when moving the mixture out of ICO. I don't see how the AvStar explanation makes sense because the pressure would be on both sides of the spring. A couple of ideas come to mind:

1. The analog gauge is sticking against the stop, or perhaps the transducer takes a bit to recover from the overpressure.

2. The mixture control wasn't kept rich long enough, especially if the throttle was closed.

If the engine starts and runs fine, I wouldn't worry too much about it

Skip

WALLCHART_rsa.pdf

[Jcarne]

4 hours ago, Ragsf15e said:

Just for fun, have you guys tried pushing in both the mixture AND the throttle at the same time to relieve the pressure?  I know much less than @PT20J about the fuel servo, but I have noticed that when I prime, if my throttle is all the way back (like the 800rpm mentioned above), my pressure will build and there’s not much flow even with mixture rich.  Pushing the throttle in  say halfway and pressure drops, flow is good.  
 

Just be advised… if you “relieve the pressure” at a fuel stop and then try to start, there’s a good chance you’ll be flooded as the fuel is going to flow into the cylinders.

Yep I tried that to no avail.  No combination of throttle/mixture would bleed it down until about 50-60 psi

[PT20J]

3 minutes ago, Jcarne said:

Yep I tried that to no avail.  No combination of throttle/mixture would bleed it down until about 50-60 psi

If it happens again, I would set mixture rich and WOT and turn on the boost pump and see if fuel starts to run out the sniffle valve at the bottom if the sump. If so, you know that fuel is getting past the mixture plate.

When troubleshooting instrument indications, the first thing to do is determine that it’s not an instrumentation problem. The standard way to do that is to ‘T’ in a known good analog gauge to validate the engine instrumentation.

Skip

[Jcarne]

1 hour ago, PT20J said:

When you shut down a hot engine with the mixture control, fuel is trapped in the line from the fuel pump to the servo, and also the line from the servo to the fuel pressure transducer. I my M20J, that's about four feet of line. With no cooling airflow, the temperature in the engine compartment rises rapidly and it will boil the fuel in these lines forming vapor and increasing the pressure in the lines. My G3X EIS fuel pressure transducer has a upper limit of 70 psi and the pressure will exceed that. 

In the original RSA design, the mixture plate did not form a tight seal when in ICO. There is a leakage spec and some leakage is expected and the only requirement is that the flow is reduced enough to shut down the engine. When AvStar copied the RSA design, it appears that it "improved" the sealing of the mixture plate at ICO. With no leakage, there is no way for the pressure to bleed off and I think that is why the AvStar servos see a high fuel pressure after shutdown. Since this pressure is blocked from the regulator (the part of the servo with the diaphragms) by the mixture plate, I don't see how it can hurt anything. 

The airflow through the servo is too low at idle to meter fuel accurately causing  the engine would run rich. To solve this, there is an idle valve mechanically connected to the throttle that blocks off most of the main metering jet during closed (or very low) throttle positions. The length of the linkage between the throttle and the idle valve is what you adjust to set the idle mixture. When priming, you should open the throttle somewhat to open the idle valve to allow more fuel to flow to the injectors. 

I don't know what would cause the pressure not to release when moving the mixture out of ICO. I don't see how the AvStar explanation makes sense because the pressure would be on both sides of the spring. A couple of ideas come to mind:

1. The analog gauge is sticking against the stop, or perhaps the transducer takes a bit to recover from the overpressure.

2. The mixture control wasn't kept rich long enough, especially if the throttle was closed.

If the engine starts and runs fine, I wouldn't worry too much about it

Skip

WALLCHART_rsa.pdf 432.98 kB · 2 downloads

Thanks for posting the chart so I don't have to find it again.  Haha

While I do agree that there would likely be no pressure inside the diaphragms I do believe the Avstar explanation has some amount of merit.  While there would be pressure on both sides of the spring that spring/shaft is in one chamber of fuel (yellow arrow) so there would be a net force trying to push that shaft out and compressing the spring.  He specifically said this will happen around 200psi.  I have a Dynon HDX and it stops sensing fuel pressure after 150 psi and it crossed that threshold.  

In the end it seems there are only two logical explanations.  Avstars explanation or somehow fuel is getting trapped in the unmetered side of the diaphragm and forcing the ball closed.  I don't see how that last one is a possibility as there is 2 ways for the two chambers to equalize.  Either way it sure is an interesting phenomenon.  

Normally I can open the mixture control and bleed it down with no issues before starting.  In the past it also doesn't matter where my throttle control is, even at idle it will bleed down.  However, on this particular trip I had a long descent in some cold weather (probably coldest fuel in my tanks I have witnessed) so I don't know if that contributed to a higher than ever pressure increase from the fuel getting heat soaked.  

I asked him suggestions on how to reduce that pressure and he gave me a few.  I asked if cracking the mixture after shutdown would work and he said "it would release the pressure, but we can't endorse that practice".  haha

Edited December 20, 2022 by Jcarne

[Jcarne]

14 minutes ago, PT20J said:

If it happens again, I would set mixture rich and WOT and turn on the boost pump and see if fuel starts to run out the sniffle valve at the bottom if the sump. If so, you know that fuel is getting past the mixture plate.

When troubleshooting instrument indications, the first thing to do is determine that it’s not an instrumentation problem. The standard way to do that is to ‘T’ in a known good analog gauge to validate the engine instrumentation.

Skip

Tried that to no avail as well.  I would be able to hear the fuel flow as well; nothing this time.  I have seen fuel come out my sniffle before after flooding so I got out and looked for it; nothing.  My first thought due to how cold it was was that somehow water got in and iced up right in that fuel jet but quickly realized that didn't make sense.  I could definitely tell that fuel was not getting through that servo though.  

I guess I will just keep the fuel mixture cracked slightly after a shutdown to avoid the problem in the future.  Just thought it was a weird one and surprised not many others have noticed it.  

Edited December 20, 2022 by Jcarne

[Jcarne]

Well after taking the cowl off and inspecting I'm thinking the Avstar explanation may have a bit of logic to it.  You can pull on the mixture control at the servo with your hands and pull the shaft out and compress the internal spring.  I would estimate that it takes about a 15-20 lb force to accomplish this.  If the mixture shaft is 3/8" diameter and you do the math it will take almost exactly 200psi to start pushing that shaft out which is exactly what the Avstar tech said.  I also noted that you can pull that shaft out a fair distance; in other words, I could absolutely see it coming unseated from the mixture plate based on pictures of the parts.  So, I will monitor and let you guys know if I notice anything else.  In the meantime I'll keep the mixture slightly cracked after shutdown to release pressure.  

[larrynimmo]

I am late getting back in this thread….

yes I still experience this issue….I have adopted a bleed down after every shutdown.

I open the throttle a little bit,  then I do 6-7 little twists of the mixture and the pressure very slowly bleeds off…I leave it like that.

occasionally I’ll let it sit and I can always see the pressure climb beyond the gauge…if it gets high enough, you won’t be able to turn in the mixture…if you are Buick, you can remove the top cowling, undo the fitting at the fuel pressure transducer and pressure bleeds off instantaneously.

[A64Pilot]

Why only crack it a little bit? why not full rich? Cracked looks like ICO but isn’t.

If it’s hydro-locked it only takes a tiny bit of fuel to be released to completely drop the pressure.

I’ve never left the battery on to watch for pressure rise after shut down but I will to see now, this high pressure / inability of fuel to flow could explain why some are so difficult to start when hot, when others aren’t.

[Jcarne]

Finally got my cowl back on and did 3 hours today with a few shutdowns.  Bleeding it down seems to have solved it. I found that I don’t need to crack the throttle, just turn the mixture in barely far enough to have it bleed down. 
 

A64pilot, my theory on why just cracked and not full rich is in case anyone turned the prop.  Either way it would sputter to life if a p-lead was broke but I would rather my mixture not be full. 
 

As far as hot starts I didn’t notice much difference.  Still fairly easy, just following the usual mixture ico until she lights trick. 
 

p.s. I also received word that the mixture plate shuts off fuel to both metered and unmetered fuel chambers so the chances of fuel getting “stuck” on the non-metered side is pretty much mute. 

[A64Pilot]

Just got back from flying, shut down from idle by pulling mixture, and sure enough the pressure stayed at 14 and slowly stated climbing, I only watched it for a couple of minutes but the increase was there and it’s not a hot day.

So I left it at idle and just pushed the mix to full rich and pressure pretty quickly went to zero, I pulled the mix back to off and only watched for a min but it didn’t increase any.

Didn’t like what I saw, I think it may be prudent to get into the habit of bleeding the pressure off after shutdown, but of course as thousands of fuel controls have no issue it’s probably not worth worrying about, just confirming what the original poster said exists also on my aircraft. I’ve never seen it before but then master comes off right after placing the keys on the glareshield too.

[larrynimmo]

4 hours ago, A64Pilot said:

Just got back from flying, shut down from idle by pulling mixture, and sure enough the pressure stayed at 14 and slowly stated climbing, I only watched it for a couple of minutes but the increase was there and it’s not a hot day.

So I left it at idle and just pushed the mix to full rich and pressure pretty quickly went to zero, I pulled the mix back to off and only watched for a min but it didn’t increase any.

Didn’t like what I saw, I think it may be prudent to get into the habit of bleeding the pressure off after shutdown, but of course as thousands of fuel controls have no issue it’s probably not worth worrying about, just confirming what the original poster said exists also on my aircraft. I’ve never seen it before but then master comes off right after placing the keys on the glareshield too.

To be very honest, I wouldn’t be bothered in the very least other than the fact that I have ruined two fuel injectors…one under Lycomming warranty, one that I had to reach in my pocket for…I also don’t want to get stranded some where…

it shouldn’t be this way…but I need to bleed down after every flight to ensure my continued fuel injector health

[Will.iam]

Just to confirm this is a lycoming issue not a continental issue correct? Or is it a certain fuel system manufacturer issue?

[EricJ]

1 minute ago, Will.iam said:

Just to confirm this is a lycoming issue not a continental issue correct? Or is it a certain fuel system manufacturer issue?

Yes, and perhaps only a particular vendor of the Bendix/Precision Airmotive/Avstar fuel servos for Lycoming engines.

[Will.iam]

1 minute ago, EricJ said:

Yes, and perhaps only a particular vendor of the Bendix/Precision Airmotive/Avstar fuel servos for Lycoming engines.

Wheew one less thing to worry about thanks. 

[A64Pilot]

13 hours ago, larrynimmo said:

To be very honest, I wouldn’t be bothered in the very least other than the fact that I have ruined two fuel injectors…one under Lycomming warranty, one that I had to reach in my pocket for…I also don’t want to get stranded some where…

it shouldn’t be this way…but I need to bleed down after every flight to ensure my continued fuel injector health

I concur and don’t have an answer, but I think it prudent to bleed the pressure off, then pull back to off. I don’t think you need to leave mixture open, but have not checked I think that will take hot wx to validate, funny that say going to lunch and getting back in I have never noticed high pressure, I would have thought I would, I guess maybe I should add check static indication of every gauge before engine start? 

Mine has been flying for years and I think about 600 hours with no issue, but I don’t like the idea of something with a redline, regularly way exceeding the redline

[PT20J]

1 hour ago, A64Pilot said:

I concur and don’t have an answer, but I think it prudent to bleed the pressure off, then pull back to off. I don’t think you need to leave mixture open, but have not checked I think that will take hot wx to validate, funny that say going to lunch and getting back in I have never noticed high pressure, I would have thought I would, I guess maybe I should add check static indication of every gauge before engine start? 

Mine has been flying for years and I think about 600 hours with no issue, but I don’t like the idea of something with a redline, regularly way exceeding the redline

Do you have a Precision Airmotive (Bendix) RSA or an AvStar servo? 

If you bleed it off and then go back to ICO the pressure will start to rise again shortly. The heat in the engine compartment is boiling off trapped fuel in the lines causing the pressure to increase and this will continue until the engine cools. Try it.

[Fly Boomer]

14 hours ago, larrynimmo said:

To be very honest, I wouldn’t be bothered in the very least other than the fact that I have ruined two fuel injectors…one under Lycomming warranty, one that I had to reach in my pocket for…I also don’t want to get stranded some where…

it shouldn’t be this way…but I need to bleed down after every flight to ensure my continued fuel injector health

I don't have a Lycoming, but I'm trying to understand the issue anyway.  I thought what was at risk was the servo -- how can this issue ruin an injector?

[A64Pilot]

15 minutes ago, PT20J said:

Do you have a Precision Airmotive (Bendix) RSA or an AvStar servo? 

If you bleed it off and then go back to ICO the pressure will start to rise again shortly. The heat in the engine compartment is boiling off trapped fuel in the lines causing the pressure to increase and this will continue until the engine cools. Try it.

Not sure, it looked like a Bendix but other than adjusting idle mixture I’ve not paid it much attention.

I understand that when you go to ICO your blocking the fuel off and pressure will build, but after zeroing it out once I wonder how high?

This is sort of similar to setting the parking brake on hot brakes, many do it without harm, but once in a blue moon usually the O-ring in the caliper let’s go. Build enough pressure and something has to give.

Many are uncomfortable leaving the mixture in any position except off, it doesn’t bother me cause my C-85 you can’t kill with mixture, I just don’t turn the prop on a motor unless the plugs are out, I’m used to a motor that can start.

Yesterday when I flew at 23 squared and 2000 ft it was tough to keep the cyl head temp in the green with cowl flaps closed, so it wouldn’t have been a good day to see how high pressure would go if it was released once, but I’m sure it wouldn’t go as high as if it was never released.

From what the original poster said if it gets high enough you can’t release the pressure with mixture, meaning no fuel flow, if nothing else I can see how that may make restarts harder, sorry I just don’t like the term hot start used casually  

[PT20J]

1 hour ago, A64Pilot said:

Not sure, it looked like a Bendix but other than adjusting idle mixture I’ve not paid it much attention.

I think the AvStars are mostly on engines from the Lycoming factory (new, rebuilt, overhauled) since Lycoming  switched vendors a few years back. Since the AvStar is a copy of the RSA design, they look pretty similar. The easiest way to tell is that the RSA body is cast aluminum whereas the AvStar is machined aluminum.

I don't think anyone ever noticed this until someone posted here about noticing it on an AvStar. The RSA probably does it also -- it depends on how well the mixture plate seals in the ICO position. It's just a flat plate. Precision used to hand lap them and they always had some leakage in ICO.. The specification is 5cc/min. max. AvStars may seal better and exacerbate the "problem."

1 hour ago, A64Pilot said:

I understand that when you go to ICO your blocking the fuel off and pressure will build, but after zeroing it out once I wonder how high?

If you release the pressure and then go back to ICO immediately after shutdown, you will have let some of the trapped fuel out of the line and reduced the pressure to zero, but the heat source (hot, cowled engine with no airflow) remains. I have tried this and the pressure does increase again shortly. I don't know how high it goes but it will peg my G3X EIS gauge which has a transducer that pegs at 70 psi. However, I would imagine that it peaks lower than not having relieved the pressure and if the pressure can get high enough to prevent movement of the mixture plate, then doing this might keep it from getting that high.

It would be easy to determine if the pressure is really forcing the mixture shaft out of the servo and disconnecting it from the mixture plate by uncowling the engine after the pressure has had some time to build and checking to see if the shaft has moved (or noticing that the spring tension seems less than when cold).

Skip

EDIT: added excerpt from RSA troubleshooting manual.

[A64Pilot]

99% sure mine isn’t machined from billet, but cast. Machined would have caught my eye.

I’ve never had any issue with mine perhaps it leaks some. Just never noticed the pressure build up is all.

[Bob Mackey]

I am installing an IO-360-A3B6 into a Lancair. The M20J came with this engine so I am hoping that the Mooney fuel mixture cable bracket could be used for my installation.
What does that bracket look like? Where could I find one?

Is that just two Adel clamps in the photo?

Edited January 29, 2023 by Bob Mackey
add photo

[PT20J]

34 minutes ago, Bob Mackey said:

I am installing an IO-360-A3B6 into a Lancair. The M20J came with this engine so I am hoping that the Mooney fuel mixture cable bracket could be used for my installation.
What does that bracket look like? Where could I find one?

Is that just two Adel clamps in the photo?

Yes. Here is the detail from the IPC. Your picture shows the old style ball and socket attachment to the mixture arm which sometimes separated. You should use a rod end.

Skip

[Bob Mackey]

Thank you P20J!

 

I also just found this in MAN205 part manual (see below).

It looks like two Adel clamps (AN742-5, AN742-3) attached to the sump with AN3-12A bolts, plus washers and nuts.

Couldn't be easier. Will have it mounted in a few minutes.