Can a Bendix servo introduce air into the fuel line?

I need a tutorial from someone who understand the workings of a Bendix RS-5 fuel servo.

I'm getting air bubbles in the feed line from servo to divider and consequently horrible engine performance and idle. Substituting clear bing fuel hoses along the pipeline shows clear fuel into the servo, plenty of bubbles in the hose downstream of the servo. Seemingly the servo therefore is allowing air into the feed line. How is this possible?

There are no fuel leaks evident anywhere.

The inlet fitting joining hose to servo is not cracked (and the line holds pressure after shutdown so the O-ring seal is good).

Each line segment is evaluated using a clear hose substituting for the high pressure AN hose, not just introduced in series, to eliminate the possibility of the hose itself leaking.

It's not a heat issue boiling the fuel (aka vapor lock) since testing is done uncowled, hoses and components are fairly cool.

I'm wondering if somehow an internal gasket or seal or a crack in a metal component within the servo can allow air into the fuel line. Ideas from the brain trust?

John Siebold

John,
I don't know if my experience is the same but I did a fuel flow test before the servo and I got air. I spent a great deal of time looking for air leaks that could have caused a problem since the lines would be under a negative pressure. I even called Precision and they had no idea. I decided to add a ball valve to add pressure to the system and the bubbles went away. I appears that at high flow and no pressure I got vapor not air. Under pressure the vapor is not generated. You might try adding a restrictor and see if you still get good flow and no bubbles.
I am surprised you get much flow after the servo as I thought fuel flow was regulated to airflow.
I just reread your post. Are you checking this while the engine is running?

Better call Saul! I mean a Don!

I'm no expert, but I just can't see anywhere in the system down stream of the pump(s) where the fuel pressure could be less than atmospheric. Seems much more likely your getting it from the suction side of a pump and your just not seeing it in your tubing yet. Perhaps the bubbles are very small going into the servo and they are consolidating in there into visible air. Maybe build an air trap "T" in your test line, see if it fills
Good luck,
Tim

A little more info

Yes, the engine is running, various rpms and long times; the bubbles are regenerative, never clearing out.

With or without the boast pump running there are no bubbles in the lines leading to the servo. The AC pump to servo hose is always under positive pressure. From the boast pump to the AC pump it's likely negative pressure 'cause the pump has to suck from the tanks without the boost pump. Yet there are no bubbles leading to the servo.

The feed line to the divider is very low positive pressure - 2-7psi. The mystery is how air gets into positive pressure lines.

Interestingly, the engine runs well from a cold start through climb out. It's after everything's warmed up trouble is obvious. Again, the bubbles don't go away with high power settings as vapor lock would. But tomorrow I'm going to observe the lines from a cold start to see what shows up.

I've have a notion there may be some material flaw that when warmed opens an air gap. There is mucho negative pressure in the venturi, especially at idle. I'm wondering if the leak is somehow associated with the venture or other innards of the servo. Anyone know RS-5s well enough to say if such a leak path is possible?

John Siebold

Guessing an air leak past the mixture shaft might be happening.

I had this same exact experience and you are probably seeing vapor bubbles from the fuel which has been sufficiently heated to vaporize under low pressure that exists after the servo outlet at low flow rates. The servo inlet pressure is 15 psi+ and the outlet pressure can be below 2 psi at idle. The flow divider keeps a fixed backpressure on the outlet side and can be setup low. In my case, the pressure set in the flow divider was low. Increasing the pressure setting to max ( I forget the value ) stopped the problem, but I had done everything I could to keep the fuel cool prior to the engine driven pump first.

The symptoms were identical. At high flow rates the pressure is higher and the bubbles go away, engine runs fine. When hot, it idled poorly, sort of surging.

Hope that helps.