Calling all RSA experts. Weird one here...

[jcarne]

Okay gents, there I was today on a fuel stop before heading the 20ish miles back home after topping off. Went to start and noticed my Dynon didn't show fuel pressure which is relatively normal once it crosses 150psi after a quick turn where the engine heat causes a pressure rise in the fuel line.

No biggy, just push the mixture in and watch the fuel pressure drop; I do it all the time before a hot start. Welllll today pushing the mixture in resulted in nothing. The pressure did not budge. Tried to start anyways to no success. Knowing that the RSA does slowly leak I went into the FBO for about 15 min and came back out. Turned the Dynon on and fuel pressure finally showed 87psi and dropping. Tried the mixture again, nothing. Waited until 56psi and tried again; finally the system bled down to zero and all was right again. Engine fired right up and after a thorough run-up and cautionary takeoff circling of the field to 12500 I was on my way. Everything operated normally the entire flight.

Okay, now the question; I have an okay knowledge of RSA internals but cannot come up with a reason in my mind as to why putting the manual mixture full rich would not equalize pressure in the two fuel diaphragms like it normally does (while not running). The internal plate should have moved, sent unmetered fuel to the metering jet and equalized. What am I missing here?

Solving the issue in the future is no big deal, just leave the mixture slightly cracked after shutdown until it bleeds down.

P.S. I'm not the only one with this problem. See this thread (post on Jan 5th and 6th 2021). Exact same scenario.

Thanks to anyone who can help me understand a bit more here. Attached is a section for your reference.

P.S. I have an Avstar unit just like in the linked forum.

 

[DanH]

Jereme, if the cockpit mixture control knob or lever moves freely, the only logical explanation (EDIT: nope, wrong) is unmetered fuel pressure (in your case, some value in excess of 150 psi) is holding the metering ball closed. The pilot has no direct mechanical control of that ball. It is connected to an air diaphragm and a fuel diaphragm, and the two deltaP's plus some springs control ball position. Photo of a fuel diaphragm with ball attached below.

I'm not sure if the rotating disc mixture valve is intended to close both the unmetered and metered fuel passages when in ICO, or if it only closes the metered fuel. The Bendix training diagram doesn't tells us. I'll ask Don Rivera. There are interesting ramifications both ways.
.

 

[lr172]

The back of the fuel side diaphragm is pressurized by the inlet pressure to help modulate with variable inlet pressures, I believe, creating a delta for the metered pressure. I suspect that the designers never contemplated inlet pressures of 150 PSI and if they did, they probably assumed that system bleed/leakage would deal with it. My guess is that the mixture control only limits on the metered side but have no idea.

I know that my freshly overhauled RSA 5 would hold 15 PSI overnight. However, after 25 hours of use, would only hold it for minutes not hours. It would seem that the question is why is yours holding that much pressure for so long, especially when it never used to. I would not expect that tolerances would normally allow holding that much pressure for 15-30 minutes. I don't believe that they have that tight of tolerances.

My first guess in varnish build up that has tightened things up internally, allowing it to capture pressures that it normally wouldn't. Do you run mogas by chance? Maybe a can of seafoam in the tank will help to clear it up. I am guessing that this is not much of an issue with 100LL, but definitely is with modern auto gas if you let it sit long enough. While I dont ever look at FP before start up, I have never been unable to start and RSA equipped engine after a heat soak in 1000 hours and that makes me think something is different in your unit, especially considering you haven't either untill yesterday.

Larry

 

[jcarne]

Jereme, if the cockpit mixture control knob or lever moves freely, the only logical explanation is unmetered fuel pressure (in your case, some value in excess of 150 psi) is holding the metering ball closed. The pilot has no direct mechanical control of that ball. It is connected to an air diaphragm and a fuel diaphragm, and the two deltaP's plus some springs control ball position. Photo of a fuel diaphragm with ball attached below.

I'm not sure if the rotating disc mixture valve is intended to close both the unmetered and metered fuel passages when in ICO, or if it only closes the metered fuel. The Bendix training diagram doesn't tells us. I'll ask Don Rivera. There are interesting ramifications both ways.
.

Dan thanks for posting. That's what I was initially thinking until I just got a call back from a tech at Avstar. I also put an email into Don because I would rather he fix it if it needed any work.

The Avstar tech said that around 200psi the mixture shaft spring will compress as the shaft is pushed out due to the pressure. This in turn causes the roll pin in the shaft to disengage from the mixture plate. After the pressure naturally bled it reseated into the groove. I'm wondering if he is correct because I also noted that usually the mixture control has a "feel" to it like there is some friction. During this time when I couldn't get it to bleed down I do remember the mixture control moving very freely; as in lacking that "feel"/ stiffness. Wish I could have taken off the cowl to see if that shaft really pushed out or not. I'll take it off soon to see if there is any signs but I doubt I will find any.

In the end the tech didn't seem overly concerned and said send it in if I notice any roughness in the controls as he explained how high pressure can cause issues with things rubbing more on the idle parts.

I attached a picture. That roll pin wouldn't have to move very far to disengage. Plausible?

 

[DanH]

Seems plausible, and the Avstar tech should know. I'm sure Don will confirm or deny.

 

[jcarne]

The back of the fuel side diaphragm is pressurized by the inlet pressure to help modulate with variable inlet pressures, I believe, creating a delta for the metered pressure. I suspect that the designers never contemplated inlet pressures of 150 PSI and if they did, they probably assumed that system bleed/leakage would deal with it. My guess is that the mixture control only limits on the metered side but have no idea.

I know that my freshly overhauled RSA 5 would hold 15 PSI overnight. However, after 25 hours of use, would only hold it for minutes not hours. It would seem that the question is why is yours holding that much pressure for so long, especially when it never used to. I would not expect that tolerances would normally allow holding that much pressure for 15-30 minutes. I don't believe that they have that tight of tolerances.

My first guess in varnish build up that has tightened things up internally, allowing it to capture pressures that it normally wouldn't. Do you run mogas by chance? Maybe a can of seafoam in the tank will help to clear it up. I am guessing that this is not much of an issue with 100LL, but definitely is with modern auto gas if you let it sit long enough. While I dont ever look at FP before start up, I have never been unable to start and RSA equipped engine after a heat soak in 1000 hours and that makes me think something is different in your unit, especially considering you haven't either untill yesterday.

Larry

Larry, see post above to see what Avstar told me.

There was one minor change to this flight leading up in that it was probably the longest coldest decent I have ever made. Probably the coldest fuel I have ever had at shutdown while topping off. I'm betting I saw more pressure increase on the fuel in the line than ever before but I can't be certain since Dynon doesn't read past 150psi.

 

[lr172]

Larry, see post above to see what Avstar told me.

There was one minor change to this flight leading up in that it was probably the longest coldest decent I have ever made. Probably the coldest fuel I have ever had at shutdown while topping off. I'm betting I saw more pressure increase on the fuel in the line than ever before but I can't be certain since Dynon doesn't read past 150psi.

That makes some sense. However, if I am understanding the diagrams correctly, there should have been an equal amount of pressure on both sides of the fuel diaphragm (feeds are interconnected before the second restrictor/jet) when in a static condition; I understand the pressure differential once the fuel starts moving, due to the jets. Therefore no imbalance on the fuel diaphragm holding it closed, just the spring. Therefore when you cranked the engine, the vacuum created on the air diaphragm should have opened the outlet, reliving the excessive pressure to the spider circuit. I am guessing this is why we rarely see this complaint. Makes me wonder why it didn't happen in this case. I suspect that at ICO this may not happen and the shaft coming off the mixture plate would explain it, as you were unable to move the plate. But I still suspect that vacuum application should have relieved the pressure even at ICO as the relief occurs as soon as the valve opens regardless of input restriction. The mixture circuit is upstream of the fuel diaphragm.

This can't be that common of a problem as I have never seen it mentioned here and most planes have only a 50 PSI sensor on the fuel system.

Larry

 

[jcarne]

That makes some sense. However, if I am understanding the diagrams correctly, there should have been an equal amount of pressure on both sides of the fuel diaphragm (feeds are interconnected before the second restrictor/jet) when in a static condition; I understand the pressure differential once the fuel starts moving, due to the jets. Therefore no imbalance on the fuel diaphragm holding it closed, just the spring. Therefore when you cranked the engine, the vacuum created on the air diaphragm should have opened the outlet, reliving the excessive pressure to the spider circuit. I am guessing this is why we rarely see this complaint. Makes me wonder why it didn't happen in this case. I suspect that at ICO this may not happen and the shaft coming off the mixture plate would explain it, as you were unable to move the plate. But I still suspect that vacuum application should have relieved the pressure even at ICO as the relief occurs as soon as the valve opens regardless of input restriction. The mixture circuit is upstream of the fuel diaphragm.

This can't be that common of a problem as I have never seen it mentioned here and most planes have only a 50 PSI sensor on the fuel system.

Larry

Larry I could be missing it on the diagram but it looks to me that the only way for fuel to get to the metered side is through the mixture plate (which would have been completely closed). So with that plate closed I don’t see any way for fuel to get to the metered side to equalize.

 

[lr172]

Larry I could be missing it on the diagram but it looks to me that the only way for fuel to get to the metered side is through the mixture plate (which would have been completely closed). So with that plate closed I don’t see any way for fuel to get to the metered side to equalize.

The first two pics that Dan posted both show the cross over between the metered and unmetered circuits before the restrictor on the metered side. It is near the diaphagm. The second pic has arrows to indicate the pressure from the unmetered side, but that incemental pressure would only exist in a dynamic environment, where the jet is reducing flow and pressure on the metered side. The 2nd pic could be a simplified drawing so may not be applicable. However the first drawing has details on the circuits and shows a clear crossover circuit just before the diaphragm. Even the pic of the diphragm shows a small jet for equalizing the two sides in a static scenario. This makes complete sense as otherwise we would see pressure lock up all the time on these engines. As long as the orifice is small enough it will allow equalizing while static, but still not affect the delta when dynamic.

Larry

 

[jcarne]

However the first drawing has details on the circuits and shows a clear crossover circuit just before the diaphragm. Correct but this is after the fuel has gone through the mixture plate and main jet. Even the pic of the diphragm shows a small jet for equalizing the two sides in a static scenario. It is my understanding that only airflow performance units have this. This makes complete sense as otherwise we would see pressure lock up all the time on these engines. As long as the orifice is small enough it will allow equalizing while static, but still not affect the delta when dynamic.

Larry

Thanks for hanging in with me Larry to help try and understand this. So in the picture I attached the crossover you are referring to is the one I pointed to in green correct?

What I'm wondering is that if that mixture plate blocks both metered and unmetered fuel (which are also connected in that big red chamber after my yellow arrows/mixture plate). In that case, where I drew the yellow arrows all of that pressure would have been blocked from going into either unmetered or metered pathways. I'm thinking that the mixture plate has to block both pathways because if it didn't ICO wouldn't work as there would still be a pathway for fuel to get into the metered and unmetered circuits.

So it would seem that the Avstar tech is suggesting that fuel pressure is building 100% behind the mixture plate and pushing the manual mixture control shaft out and unseating it from mixture plate.

 

[DanH]

The first two pics that Dan posted both show the cross over between the metered and unmetered circuits before the restrictor on the metered side. It is near the diaphagm.

It's actually in the fuel diaphragm, and I forgot about it.

The little button next to the ball is a pinhole passage between the unmetered and metered sides. I seem to recall it being described as a way to get air bubbles out of the unmetered side, and yes, it is open all the time, flowing as a fixed jet.

So, given this constant leak, it does seem unlikely that unmetered fuel pressure can hold the ball valve closed...unless the pinhole is blocked

The second pic has arrows to indicate the pressure from the unmetered side, but that incemental pressure would only exist in a dynamic environment, where the jet is reducing flow and pressure on the metered side. The 2nd pic could be a simplified drawing so may not be applicable.

It is simplified, shows dynamic activity, and it is my drawing, not Bendix, created for an old Kitplanes article.

 

[lr172]

It's actually in the fuel diaphragm, and I forgot about it.

The little button next to the ball is a pinhole passage between the unmetered and metered sides. I seem to recall it being described as a way to get air bubbles out of the unmetered side, and yes, it is open all the time, flowing as a fixed jet.

So, given this constant leak, it does seem unlikely that unmetered fuel pressure can hold the ball valve closed...unless the pinhole is blocked



It is simplified, shows dynamic activity, and it is my drawing, not Bendix, created for an old Kitplanes article.

I dont think it is for air bubbles if it is sitting in the middle. air bubbles would collect at the upper most cavity and a designer would likely put it there for that purpose. I speculate it is specifically to let the two sides equalize to address the specific problem seen by the OP. Just a guess, not a challenge to your idea. I do agree that a blocked passage would allow excessive, unequalized pressure on the unmetered side and the most likely candidate is varnish or whatever the scientific name for that stuff is. I still think it is worth running a can of seafoam through the engine to see if it will clear it.

Larry

 

[lr172]

Thanks for hanging in with me Larry to help try and understand this. So in the picture I attached the crossover you are referring to is the one I pointed to in green correct?

What I'm wondering is that if that mixture plate blocks both metered and unmetered fuel (which are also connected in that big red chamber after my yellow arrows/mixture plate). In that case, where I drew the yellow arrows all of that pressure would have been blocked from going into either unmetered or metered pathways. I'm thinking that the mixture plate has to block both pathways because if it didn't ICO wouldn't work as there would still be a pathway for fuel to get into the metered and unmetered circuits.

So it would seem that the Avstar tech is suggesting that fuel pressure is building 100% behind the mixture plate and pushing the manual mixture control shaft out and unseating it from mixture plate.

I have serious doubt that the mixture plate can hold back 200 PSI of force for a meaningfull amount of time, though I am confident that the ball / seat can. I believe it is just metal plates without a sealing mechanism and is only mostly effective, as that is all it needs to be to meet its design goal. As I mentioned, my freshly overhauled unit would hold 15 PSI for a day, but after only moderate use it would only do that for a short period of time. If that is the case, it shouldn't matter as over time it should equalize on both sides of the diagphragm and should not be an impediment to letting the vacuum force on the air diaphragm from opening the gate and expelling the excess pressure. Meaning that yes, a lot of pressure exists, but after a few revolutions of the engine, it should clear itself. I stick with my recommendation to run seafoam or some other fuel system cleaner to see if you can break up the blockage in that jet. The fact that this problem is relatively unique to your servo tells me that the jet works and in your case it is not and therefore a good chance yous is blocked.

I don't doubt that 200 PSI can build and push the shaft off the plate. But you have to wonder why it is doing it to your unit and preventing a proper start, yet rarely happens to anyone else. It tells me that your unit is somehow unique. LOTS of guys have heat soak on the ramp during a re-fuel, yet almost never hear of this type of problem. Sure it is hard to start, but it eventually starts. Certainly can't be sure that varnish is blocking the jet, but it is a cheap and easy step to throw in a can of cleaner and see if it changes.

Larry

 

[Avanza]

Trapped fuel

To me it seems that the fuel is trapped between the mechanical fuel pump and
the ball valve on the outlet of the fuel servo.
Using rubber hose that can expand a bit and fire sleeve I have no problem.

Good luck

 

[jcarne]

I have serious doubt that the mixture plate can hold back 200 PSI of force for a meaningfull amount of time, though I am confident that the ball / seat can. I believe it is just metal plates without a sealing mechanism and is only mostly effective, as that is all it needs to be to meet its design goal. As I mentioned, my freshly overhauled unit would hold 15 PSI for a day, but after only moderate use it would only do that for a short period of time. If that is the case, it shouldn't matter as over time it should equalize on both sides of the diagphragm and should not be an impediment to letting the vacuum force on the air diaphragm from opening the gate and expelling the excess pressure. Meaning that yes, a lot of pressure exists, but after a few revolutions of the engine, it should clear itself. I stick with my recommendation to run seafoam or some other fuel system cleaner to see if you can break up the blockage in that jet. The fact that this problem is relatively unique to your servo tells me that the jet works and in your case it is not and therefore a good chance yous is blocked.

I don't doubt that 200 PSI can build and push the shaft off the plate. But you have to wonder why it is doing it to your unit and preventing a proper start, yet rarely happens to anyone else. It tells me that your unit is somehow unique. LOTS of guys have heat soak on the ramp during a re-fuel, yet almost never hear of this type of problem. Sure it is hard to start, but it eventually starts. Certainly can't be sure that varnish is blocking the jet, but it is a cheap and easy step to throw in a can of cleaner and see if it changes.

Larry

Larry I think we may be underestimating how well these Avstar units are sealing on that mixture plate as I'm not the only one that has seen it (albeit a low number for sure). I will say that this pressure rise and holding has been happening since day 1 on my engine. This is just the first time that pressure appears to have built enough to push that shaft out. As mentioned above this is definitely the coldest fuel/ conditions I have had to date which may explain a higher temp/ pressure rise in the line after shutdown. The pressure building and holding for a long time is not by any means a new development with my engine; I have just been always able to bleed it down by pushing the mixture control in; but I have always seen it rise and hold for a long time.

Yesterday I removed the cowl and now think that the Avstar tech may be correct. I tried pulling on the manual mixture shaft with my hands and sure enough you can pull it out a bit and it will spring load back in. I would estimate the spring pressure is about 15-20 lbs. If that shaft is 3/8" diameter 200 psi would be the perfect amount to push it out.

Your theory may still have some merit and running a cleaner through may not hurt anything but I would have to research that before I put anything in my tank that isn't 100LL. It also doesn't make sense in my mind that less pressure would result in a blockage completely clearing with less pressure coincidentally at the same time my mixture control started working again. I will of course research the Seafoam though, so thanks for the recommendation.

 

[jcarne]

To me it seems that the fuel is trapped between the mechanical fuel pump and
the ball valve on the outlet of the fuel servo.
Using rubber hose that can expand a bit and fire sleeve I have no problem.

Good luck

Rubber hose on a high pressure fuel line?

 

[lr172]

Rubber hose on a high pressure fuel line?

Find me a car that doesn't use rubber hose in the high pressure fuel circuit. The engine moves relative to the frame and therefore THERE MUST be rubber somewhere in the fuel circuit between the frame and the engine. You will also find them in front end loaders using 2-3000 PSI of fluid pressure.

Larry

 

[lr172]

Larry I think we may be underestimating how well these Avstar units are sealing on that mixture plate as I'm not the only one that has seen it (albeit a low number for sure). I will say that this pressure rise and holding has been happening since day 1 on my engine. This is just the first time that pressure appears to have built enough to push that shaft out. As mentioned above this is definitely the coldest fuel/ conditions I have had to date which may explain a higher temp/ pressure rise in the line after shutdown. The pressure building and holding for a long time is not by any means a new development with my engine; I have just been always able to bleed it down by pushing the mixture control in; but I have always seen it rise and hold for a long time.

Yesterday I removed the cowl and now think that the Avstar tech may be correct. I tried pulling on the manual mixture shaft with my hands and sure enough you can pull it out a bit and it will spring load back in. I would estimate the spring pressure is about 15-20 lbs. If that shaft is 3/8" diameter 200 psi would be the perfect amount to push it out.

Your theory may still have some merit and running a cleaner through may not hurt anything but I would have to research that before I put anything in my tank that isn't 100LL. It also doesn't make sense in my mind that less pressure would result in a blockage completely clearing with less pressure coincidentally at the same time my mixture control started working again. I will of course research the Seafoam though, so thanks for the recommendation.

You may be right. However, you cannot be certain whether or not it is the mixture plate holding the pressure back or the ball/seat at the fuel exit doing that. The fact that opening of hte mixture valve doesn't relieve the pressure tells me that it is more likely that the mixture plates are not holding it back, but instead the exit ball is. With the ball holding the pressure in, all of the fuel in the whole unit is under 200 PSI of force and it is this force pushing out the stem. The doesn't necessarily mean that the mixture plates are creating the pressure. Just a different opinion, as I struggle to see how those plates can hold 200 PSI. Then again I have never had one apart so really have no idea.

 

[jcarne]

Find me a car that doesn't use rubber hose in the high pressure fuel circuit. The engine moves relative to the frame and therefore THERE MUST be rubber somewhere in the fuel circuit between the frame and the engine. You will also find them in front end loaders using 2-3000 PSI of fluid pressure.

Larry

I may have interpreted his response incorrectly for a different type of fuel line. When I hear rubber line I think barbs and clamps.

You may be right. However, you cannot be certain whether or not it is the mixture plate holding the pressure back or the ball/seat at the fuel exit doing that. The fact that opening of hte mixture valve doesn't relieve the pressure tells me that it is more likely that the mixture plates are not holding it back, but instead the exit ball is. With the ball holding the pressure in, all of the fuel in the whole unit is under 200 PSI of force and it is this force pushing out the stem. The doesn't necessarily mean that the mixture plates are creating the pressure. Just a different opinion, as I struggle to see how those plates can hold 200 PSI. Then again I have never had one apart so really have no idea.

Ya I'm kind of surprised at the lack of pictures/videos of the internals on these servos. Would be awesome to take one of them bad boys apart and make a video of it. At any rate I'll just leave the mixture cracked a bit after shutdown. If that works consistently I suppose that will answer the question. Interestingly the tech at Avstar said this has become a somewhat bigger problem since Lycoming started using pumps that don't let any backwards bleed down of the pressure. He said the tempest pumps seem to let it bleed down a bit. I don't know how true this is, just passing on what he said when I asked him how various other aircraft manufacturers deal with this pressure rise on shutdown. He also brought up Airflow Performance bypass valves too.

 

[DanH]

The fact that opening of the mixture valve doesn't relieve the pressure tells me that it is more likely that the mixture plates are not holding it back, but instead the exit ball is.

The AvStar tech described a scenario in which the mixture lever is temporarily disconnected from the mixture plate.

Would be awesome to take one of them bad boys apart and make a video of it.

I volunteer, if someone has an RSA-5 available for the task.

He also brought up Airflow Performance bypass valves too.

An AFP purge valve would not change this particular issue. It vents pressure after the servo, at the divider inlet.

The drum-type mixture valve on earlier AFP throttle bodies had a high leak rate. The disc-style mixture valve has a far lower leak rate. And the way Don has the parts arranged in the AFP disc valve (essentially opposite the Bendix approach), it does not appear it can disconnect itself from the shaft due to a pressure rise.
.

 

[skylor]

Larry I think we may be underestimating how well these Avstar units are sealing on that mixture plate as I'm not the only one that has seen it (albeit a low number for sure). I will say that this pressure rise and holding has been happening since day 1 on my engine. This is just the first time that pressure appears to have built enough to push that shaft out. As mentioned above this is definitely the coldest fuel/ conditions I have had to date which may explain a higher temp/ pressure rise in the line after shutdown. The pressure building and holding for a long time is not by any means a new development with my engine; I have just been always able to bleed it down by pushing the mixture control in; but I have always seen it rise and hold for a long time.

Yesterday I removed the cowl and now think that the Avstar tech may be correct. I tried pulling on the manual mixture shaft with my hands and sure enough you can pull it out a bit and it will spring load back in. I would estimate the spring pressure is about 15-20 lbs. If that shaft is 3/8" diameter 200 psi would be the perfect amount to push it out.

Your theory may still have some merit and running a cleaner through may not hurt anything but I would have to research that before I put anything in my tank that isn't 100LL. It also doesn't make sense in my mind that less pressure would result in a blockage completely clearing with less pressure coincidentally at the same time my mixture control started working again. I will of course research the Seafoam though, so thanks for the recommendation.

After looking at the posted servo diagrams carefully, I believe the only possible scenario for what you experienced is the pushed out/disconnected mixture lever described by the Avstar rep.

Since the unmetered fuel pressure side of the flow control diaphragm feeds from the downstream side of the mixture/ICO valve and comes from the same cavity that supplies the main metering jet of the servo, I don’t believe that the pressure on the diaphragm could ever build fast enough in a heat soak situation for the diaphragm to completely shut off the the ball valve since fuel will be bled through the main jet/idle valve side as well. The main metering jet/idle valve never closes completely because it has to supply fuel during idle.

Skylor

 

[DanH]

Got an informative note back from Don at Airflow Performance. He says Bendix and AFP disk valves close both metered and unmetered fuel in the ICO position. The bleed between the metered and unmetered sides of the fuel diaphragm would equalize pressure, and the spring would open the ball valve. So, a closed ball seems unlikely. He has not run pressure numbers to push out the mixture shaft, or checked the necessary clearance to unseat the roll pin from the disk valve in an Avstar.

Until shown otherwise, I'll go with the Avstar explanation, and Jereme, apparently you have a mixture disk which seals well.

 

[jcarne]

The AvStar tech described a scenario in which the mixture lever is temporarily disconnected from the mixture plate.



I volunteer, if someone has an RSA-5 available for the task.



An AFP purge valve would not change this particular issue. It vents pressure after the servo, at the divider inlet.

The drum-type mixture valve on earlier AFP throttle bodies had a high leak rate. The disc-style mixture valve has a far lower leak rate. And the way Don has the parts arranged in the AFP disc valve (essentially opposite the Bendix approach), it does not appear it can disconnect itself from the shaft due to a pressure rise.
.

That AFP unit sure does look different inside. Also very nice!

Dan I will also ask around to some mechanics and see if we can get a hold of an old RSA to take apart. Would be fun!

After looking at the posted servo diagrams carefully, I believe the only possible scenario for what you experienced is the pushed out/disconnected mixture lever described by the Avstar rep.

Since the unmetered fuel pressure side of the flow control diaphragm feeds from the downstream side of the mixture/ICO valve and comes from the same cavity that supplies the main metering jet of the servo, I don’t believe that the pressure on the diaphragm could ever build fast enough in a heat soak situation for the diaphragm to completely shut off the the ball valve since fuel will be bled through the main jet/idle valve side as well. The main metering jet/idle valve never closes completely because it has to supply fuel during idle.

Skylor

Thanks Skylor, agree with that myself; especially after being able to pull it out by hand with no pressure on the system.

Got an informative note back from Don at Airflow Performance. He says Bendix and AFP disk valves close both metered and unmetered fuel in the ICO position. The bleed between the metered and unmetered sides of the fuel diaphragm would equalize pressure, and the spring would open the ball valve. So, a closed ball seems unlikely. He has not run pressure numbers to push out the mixture shaft, or checked the necessary clearance to unseat the roll pin from the disk valve in an Avstar.

Until shown otherwise, I'll go with the Avstar explanation, and Jereme, apparently you have a mixture disk which seals well.

Dan thanks for reaching out to Don and posting what you got back. Appreciate it, and the added knowledge from the thread.

 

[lr172]

Got an informative note back from Don at Airflow Performance. He says Bendix and AFP disk valves close both metered and unmetered fuel in the ICO position. The bleed between the metered and unmetered sides of the fuel diaphragm would equalize pressure, and the spring would open the ball valve. So, a closed ball seems unlikely. He has not run pressure numbers to push out the mixture shaft, or checked the necessary clearance to unseat the roll pin from the disk valve in an Avstar.

Until shown otherwise, I'll go with the Avstar explanation, and Jereme, apparently you have a mixture disk which seals well.

Interesting and informative. Thanks for getting that. I would have guessed that the spring was pushing the ball against the seat, but apparently that was not correct.

 

[Jpm757]

With that crazy high fuel pressure, I hope you are using the updated pressure transmitter which was the subject of a Dynon Service Letter.

 

[jcarne]

With that crazy high fuel pressure, I hope you are using the updated pressure transmitter which was the subject of a Dynon Service Letter.

Yeppers, still part of my inspection before I cowl up again though.