What's new
Van's Air Force

Don't miss anything! Register now for full access to the definitive RV support community.

Pls help: Hot idle problems with new Bendix FI

L

lr172

Well Known Member
Well, I finished my Bendix FI installation. The good news is the high power aspect works great. Need a tweak or two on the restrictors, but pretty close already. Performance is great.

My problem is hot idle. I had been struggling to get the mixture set. I finally adjusted to lowest MAP/highest RPM. this resulted in a 20-30 RPM drop when going to ICO. I did this with the engine hot, but the cowl top off. I was able to have it idle for several minutes without a hiccup. Nice, smooth and stable.

It was 75* OAT today. On initial flight, it idles great and all is good. Once it gets hot under the cowl, the idle is horrific. It will initially go to idle from a higher RPM. After 15-20 seconds it gets a bit rough then it starts to drop out (just about dies and comes back). It will keep doing that until I run it up. If I leave it there too long, I can barely get it up to RPM and it will not clear itself until about 1600 (seem this is the point completely off the idle circuit or FI compliment).

I just tried a whole bunch of different mixture settings (re-cowling each time) and always the same result. I even got it lean enough that it stumbled on acceleration. I suspect this has to be heat related and boiling gas somewhere. However, I don't know how you guys could deal with what I am experiencing. Taxi is quite difficult to keep it running and even had some rough running on short final 1000-1200. Landing roll out is also difficult once I am below 30 mph or so and windmilling effects slows.

The feeling is like it is "loading up." After the idle event, leaning at higher RPMs helps to clear it faster, further supporting that. However, I don't see how I am getting too rich at idle. I have it perfectly set and it purrs with the cover off. I have tried leaning with the mixture knob at idle to get a RPM rise, but I still get the same results.

please help. I messed with this all day and don't know how to address it. I really don't want to pull this all out and put the carb back, but this just doesn't seem safe.

EDIT: BTW, the EGT are all very consistent while this is happening, so I don't believe it is a single cylinder issue. However, I do notice that when trying to clear things out at higher RPM, 3 & 4 come back to higher EGTs much faster than 1 & 2.
Larry
 
Last edited:
Standard issue with fuelinjection no solution for this problem found yet
Fuel is boiling in the injector lines electronic ignition make it worst while trying to ajust timing
 
I assume all your fuel lines under the cowling are fire sleeved? There is a shroud for the fuel pump, but I have never needed one. I have an IO360M1B with same ful injection.
 
vluvelin said:
Standard issue with fuelinjection no solution for this problem found yet
Fuel is boiling in the injector lines electronic ignition make it worst while trying to ajust timing
Click to expand...

Standard fix is to turn on the fuel boost pump. This helps if fuel is boiling in the mechanical pump, sometimes if it's the lines.
 
BobTurner said:
Standard fix is to turn on the fuel boost pump. This helps if fuel is boiling in the mechanical pump, sometimes if it's the lines.
Click to expand...

I did try this on a few occassions and it's impact was marginal but did help some in a couple of cases.

Larry
 
AltonD said:
I assume all your fuel lines under the cowling are fire sleeved? There is a shroud for the fuel pump, but I have never needed one. I have an IO360M1B with same ful injection.
Click to expand...

All lines firesleeved and relatively short. My -4 line to the spider could be a foot, but can't imagine it would help that much.
 
vluvelin said:
electronic ignition make it worst while trying to ajust timing
Click to expand...

Could you elaborate on this? I have an EI and one mag. My timing ramps from 15* at idle to 25* at 1800 (configurable map). Am I better off cranking it up to 25 at idle? I did try turn off the mag and the EI at idle and didn't notice any significant improvement.

Larry
 
Post some photos of the installation.

Lumpy hot idle with MFI is normal. Bad lumpy/won't stay running is not. In my experiments, one inductive EI and fine wire Denso plugs caused a large improvement in idle quality.

I would definitely advance the idle timing to 25 at idle speed, maybe more.
 
lr172 said:
I had been struggling to get the mixture set. I finally adjusted to lowest MAP/highest RPM. this resulted in a 20-30 RPM drop when going to ICO.


EDIT: BTW, the EGT are all very consistent while this is happening, so I don't believe it is a single cylinder issue. However, I do notice that when trying to clear things out at higher RPM, 3 & 4 come back to higher EGTs much faster than 1 & 2.
Larry
Click to expand...

You may want to look for an induction leak. Temporarily replace the sniffle valve with a plug to test that. Replace the intake gaskets and check the connecting sleeves on the two cylinders where EGT goes up faster than the others when leaning at idle.

Carl
 
Not unusual

Same deal with my set up. Air bubbles are forming in the fuel line. Pressure drop around elbows, orifices, etc., can set off cavitation, made more susceptible with heat, or just heat alone. Low fuel flows at idle are easily heated, high power flows can adsorb the heat without vaporizing. If you're curious where the bubbles start, splice a piece of clear tubing in the fuel line various places. My bubbles start at the servo output. I've tried jumpering around aux and engine pumps, and fuel selector to isolate cavitation problems, but no joy. You might try a higher rate spring in the divider; creates a bit more in-line pressure downstream from the servo.

John Siebold
 
Well, Big improvement today! I did some research here and did a few things today. First, I was a little sloppy with two of the SS lines. I re-worked them so that they are directly over the pushrod and as high as possible. This alleviate a couple of small clearance spots with the heads. I also wrapped most of the SS lines with Aluminum tape. It doesn't help with ambient temps, but it does help with radiant heat which I presumed was the biggest problem. I also reset the idle mixture with the cover off to lowest MAP.

The improvement was significant. It wasn't sewing machine smooth, but reliable enough to stay running. If I let it get below 800 RPM it gets rougher and chocked a couple of times, but above that it was reliable, albeit with some RPM fluctuation.

Thanks for your efforts to assist. I am so happy I don't have to rip this out and put the carb back in. With the mix of restrictors that got with my setup, my GAMI spread is only .3 according to my D100 (still need to review logs for details). I'm pretty happy about that, but may tweak a bit and see if I can make it better. I think I was hitting 190 MPH TAS with about 7.5 GPH up at 8000 DA. At least 1GPH better than before FI and one EI.

EDIT: OAT was 80 today.

Larry
 
Last edited:
Still having some issues. I did another flight today and did some touch and go's, I drop to about 1300 RPM on approach and by final I am getting some stumbles. It seems to survive it fine, but a bit unnerving to add the thought of an engine stopping on final. It's worse when I am little high and need to drop to idle. Will windmilling at 85 MPH provide enough energy to re-start if it dies? Does this happen to you guys as well? I am wondering whether I am at the point of everyone else or if there are still things to do. After touchdown and taxi, it is manageable if I goose the throttle once in a while and keep the rpm's around 900. I did let it set without assistance at my spot and it did die after 10 seconds or so.

Please let me know if you think I have areas for improvement or if what I have above is typical. I took some pictures and will post tonight. Need to find my account to post them.

Larry
 
What fuel are your running? 100LL or Mogas?

When you changed out the carburetor and installed the Bendix FI, did you put in a boost pump? What kind? What is your indicated fuel pressure?

+1 what Bob Tuner said. I run my boost pump after hot starts or I get some symptoms like you describe while idling and taxing.
 
Chkaharyer99 said:
What fuel are your running? 100LL or Mogas?

When you changed out the carburetor and installed the Bendix FI, did you put in a boost pump? What kind? What is your indicated fuel pressure?

+1 what Bob Tuner said. I run my boost pump after hot starts or I get some symptoms like you describe while idling and taxing.
Click to expand...

I did. I put in the older style AFP setup. I am getting somewhere in the 25 PSI range. I usually get 2-3 more PSI with the pump on. The issues I saw on Final, posted above, were with the boost pump on. I have seen very slight improvement when running it.

Larry
 
Chkaharyer99 said:
What fuel are your running? 100LL or Mogas?

When you changed out the carburetor and installed the Bendix FI, did you put in a boost pump? What kind? What is your indicated fuel pressure?

+1 what Bob Tuner said. I run my boost pump after hot starts or I get some symptoms like you describe while idling and taxing.
Click to expand...

I learned today that my boost pump may not be performing adequately. I have ordered a new pump and we'll see if that helps at all.

Larry
 
Pics of my setup after I modified the injector lines, but before aluminum tape.

n2ecxs.jpg
el5rg9.jpg
vpxq54.jpg

zlsr6g.jpg
 
lr172 said:
Could you elaborate on this? I have an EI and one mag. My timing ramps from 15* at idle to 25* at 1800 (configurable map). Am I better off cranking it up to 25 at idle? I did try turn off the mag and the EI at idle and didn't notice any significant improvement.

Larry
Click to expand...

If you will connect vacume gage and see MAP you will see it is fluctuating
Imagine what is doing your EI to compensate for pressure changes

Tri to set EI to 25 btdc without compensation while working on this issue

I dont have issues while flying or on final after landing takes about 1 minut of taxi time to get hickups. Make sure no induction leaks
 
vluvelin said:
If you will connect vacume gage and see MAP you will see it is fluctuating
Imagine what is doing your EI to compensate for pressure changes

Tri to set EI to 25 btdc without compensation while working on this issue

I dont have issues while flying or on final after landing takes about 1 minut of taxi time to get hickups. Make sure no induction leaks
Click to expand...

I have no MAP compensation in my timing map until about 2000 RPM. It is a custom map, so I know what is there. It is now 25* from 500 to 2000 RPM.

I am pretty sure I do not have an intake leak, An intake leak in a port injection system would only create one cylinder running leaner than the others, but it would be consistent and not intermittant. I have checked my EGT spreads and they are the same at WOT (30") and 23" of MAP. An induction leak would cause the leaking intake to be lean at lower MAP and same as others when MAP is 30 or ambient pressure. I also see my EGT's very close to each other in the 700-1300 range where the intake leak would cause the most leanness.

Larry
 
Last edited:
I haven't seen dozens of engines, but the ones I have seen and owned all had the FI nozzles at the top of the cylinders. My -4 had an IO-320 but never had any significant idle problems other than a little bit lumpy when hot (kinda like a race cam in a car).

Looks a bit odd to me to see the SS lines from the spider running to the bottom of the cylinders, however maybe this is a standard setup on some engines??
 
Injector lines

SMO said:
I haven't seen dozens of engines, but the ones I have seen and owned all had the FI nozzles at the top of the cylinders. My -4 had an IO-320 but never had any significant idle problems other than a little bit lumpy when hot (kinda like a race cam in a car).

Looks a bit odd to me to see the SS lines from the spider running to the bottom of the cylinders, however maybe this is a standard setup on some engines??
Click to expand...

plus 1 on that. The lines are much longer than normal and significant portions are outside the baffles where there is little if any cooling airflow. That is probably a significant part of your problem.
 
jrs14855 said:
The lines are much longer than normal and significant portions are outside the baffles where there is little if any cooling airflow. That is probably a significant part of your problem.
Click to expand...

Agree. Installing the injector nozzles in the primer ports is allowed, but I wouldn't want to do it.

Second issue; you've complied with the instructions to install your red cube in the line between the fuel control and the flow divider. Although it will no doubt meter faultlessly there, it appears your line runs rearward from the fuel control, up the firewall, forward through the rear baffle wall, across the top of two cylinders, then loops 180 degrees to turn back to the divider. Only the last 12-16 inches has fuzzy firesleeve, so that's at least 4 feet of line serving as an excellent fuel heater.

There's a third issue too. Nozzle bodies have an air bleed hole. When installed upside down they tend to drip fuel after shutdown, so it's good practice to use turbo nozzles, and install a turbo bleed air rail as a drain. It gets the fuel away from the hot pipes.
 
DanH said:
Second issue; you've complied with the instructions to install your red cube in the line between the fuel control and the flow divider. Although it will no doubt meter faultlessly there, it appears your line runs rearward from the fuel control, up the firewall, forward through the rear baffle wall, across the top of two cylinders, then loops 180 degrees to turn back to the divider. Only the last 12-16 inches has fuzzy firesleeve, so that's at least 4 feet of line serving as an excellent fuel heater.
Click to expand...

FYI the fuel line from the servo to the spider on my -4 was routed up between cylinders 1 and 3, made it quite a bit shorter run and was in the cooler air flow. It was a brown flexible line with integrated fire sleeve - looks like that is what yours is as well.
 
I've never seen a builder go to so much trouble to setup FI lines so wrong. Step one, get the zip ties and sleeves off the injector tubes. Step two, re-do them correctly. Get a new set of tubes from Don at AFP, Install the nozzzles and tubes per the instructions. I wouldn't pass that setup on a condition inspection and it shouldn't have passed airworthiness inspection. By the time you get it done right you will be like "what were they thinkin?".
 
My cylinders do not have injector ports on the top, like newer cylinders. In fact, they don't even have the bosses cast in, so I couldn't drill and tap them. Using the primer ports is NOT un-airworthy and I have learned several helicopter applications have them installed this way. I have to deal with this, as I am not prepared to spend $4K for new cylinders at this point.

dripping from the injectors is an an issue at shut-down and I chose to use the rubber tubes around the injectors to capture the drips on shut-down. They are loose fitting to allow air in (injectors are air bleed type) and the zip ties just hold the rubber tubing so it doesn't shift. I will see how this works before investing in the turbo nozzles. So far, It is has contained all dripping.

The hose from the servo to the red cube is integral firesleeve. However, I think that I am going to re-route it between the cylinders and add a 90 fitting on the spider to shorten the distance. I will try to mount the red cube below the engine and firesleeve it for insulation. I think I can eliminate at least a couple feet of hose.this way. That hose came off a well running engine, so I assumed length wasn't an issue. I'm learning.

I spoke with Don this morning and he suggested .022 restrictors, as I am running this on a 320. This should get my pressure up at idle and help the problem. He also suggested insulating the red cube with firesleeve, which I will also do.
 
Last edited:
DanH said:
Only the last 12-16 inches has fuzzy firesleeve, so that's at least 4 feet of line serving as an excellent fuel heater.
Click to expand...

Does the fuzzy firesleeve insulate better than the integral firsesleeve hose? If so, I can put a layer of additional firesleeve on that hose if it will help.

Larry
 
lr172 said:
I spoke with Don this morning and he suggested .022 restrictors, as I am running this on a 320. This should get my pressure up at idle and help the problem. He also suggested insulating the red cube with firesleeve, which I will also do.
Click to expand...

Changing to .022 nozzles helped my "hot idle" issue quite a bit.
 
lr172 said:
dripping from the injectors is an an issue at shut-down and I chose to use the rubber tubes around the injectors to capture the drips on shut-down. They are loose fitting to allow air in (injectors are air bleed type) and the zip ties just hold the rubber tubing so it doesn't shift. I will see how this works before investing in the turbo nozzles. So far, It is has contained all dripping.
Click to expand...

Ahh, so that's what the black rubber sleeves are for. I thought you were shooting for some kind or radiant heat shielding.

Any chance those sleeves are catching and holding fuel, like a sink trap? It would really screw up the air bleed function at idle.

lr172 said:
Does the fuzzy firesleeve insulate better than the integral firsesleeve hose? If so, I can put a layer of additional firesleeve on that hose if it will help.
Click to expand...

It will help. The integral silicone rubber layer is a pretty good insulator, but not nearly as good as adding a second layer of silicone enclosing a trapped air space.
 
DanH said:
Ahh, so that's what the black rubber sleeves are for. I thought you were shooting for some kind or radiant heat shielding.

Any chance those sleeves are catching and holding fuel, like a sink trap? It would really screw up the air bleed function at idle.
Click to expand...

It seems very unlikely. I watched it drip once or twice and there is not much volume. Due to the angle, the tubes are such that fuel would spill out on the head side before fully covering the injector. In theory this could still cause blockage, depending upon where the bleed hole is oriented. However, My problems are quite consistent and will happen after on hour's flight. If the tube had collected gas from the prior shut down, I have all confidence that an hour's running would have forced it to evaporate. Fuel cannot leak or dribble when the engine is running.

I have done quite a bit of testing and it has never not purred like a kitten when cool or with the cowl off, even with numerous shut downs in a row that, in theory, should have loaded up the tube with fuel.

Your point here is quite valid and I have thought about it as well. If my problems persist, it is my plan to re-test without the tubes.

Larry
 
Last edited:
DanH said:
It will help. The integral silicone rubber layer is a pretty good insulator, but not nearly as good as adding a second layer of silicone enclosing a trapped air space.
Click to expand...

I will add this if the next remediation steps don't fully work.

Larry
 
My current remediation plan:

- Add aluminum tape to remainder of SS lines to help reject radiant heat.
- install .022 restrictors
- change 1' piece of -6 line with shorter -4 line between spider and Red cube
- add aluminum tape to red cube to reject radiant heat
- Re-route -4 line from servo to cube between #2 & #4 cylinders. This should eliminate at least 3' of fuel line or more.
- Add aluminum tape to -4 brown fuel line to help reject radiant heat from exhaust and cylinders in proximity

I will report results once I can get installed and tested. Thanks for all of the assistance and ideas here.

Larry
 
Last edited:
Ok. The cylinders looked new style and I couldn't imagine they didn't have top ports. So, that's what you have carry on. I still wouldn't have any zips for chafe makers long term.
 
aerhed said:
Ok. The cylinders looked new style and I couldn't imagine they didn't have top ports. So, that's what you have carry on. I still wouldn't have any zips for chafe makers long term.
Click to expand...

The zips are not permanent. I am still working through the best long-term approach to keep the injectors from dribbling on the exhaust at shut down. Currently they only contact the rubber, not the SS line; so no chafing concerns.
 
aerhed said:
Side question: Do new 320ND cylinders all lack the top port?
Click to expand...

My research indicates that they either have the port or a boss in which to drill the port. My cylinders are from the 60's and back then, Lycoming was happy with two different castings and therefore no boss.

Larry
 
lr172 said:
- Re-route -4 line from servo to cube between #2 & #4 cylinders. This should eliminate at least 3' of fuel line or more.
Click to expand...

An interesting change to quantify. Removing 3 feet of 5/8 OD hose eliminates about 70 square inches of heat absorbing surface area.
 
DanH said:
An interesting change to quantify. Removing 3 feet of 5/8 OD hose eliminates about 70 square inches of heat absorbing surface area.
Click to expand...

Most of the footage will come from the area of the cowl that is probably in the 180* range, so I am thinking this should help. I am sure this is all cumulative and will make some difference, at least in the severity. Just didn't to think about this stuff with carb, as it is immune from all but severe boiling.

Larry
 
I completed all of the steps that I outlined in my last post and things have definitely improved. I no longer have problems in the air, thank goodness! When cold, it idles great. I start to get a little roughness on roll out and taxi after landing is a bit rough. I am getting a bit more than what you guys are referring to as a loppy idle with 50 RPM variations. It still drops out occasionally, but now just drops to a lower RPM (200 RPM drop) and doesn't try to do die. If I run it up to around 1000 RPM, I have no issues. At 800, I get the roughness and drop outs, but it doesn't feel like it will die if I don't catch it. I thnk I can live with the higher idle during taxi. I just may have to change brake pads a little more often. Also, having the boost pump on seems to help a little.

Does this sound like I am in the neighborhood of what you guys are experiencing? Ambient today was 80*

Surprisingly, the .022 restrictors left me almost perfect distribution and T/O flow was the same or higher as with the .028's. I may need to up one of them, but they are really close.

EDIT: I ended up getting rid of about 1.5-2" of fuel hose betweeen the servo and the spider. I can get rid of another foot if I route the hose between #1&3 cylinders. However, I have to run the hose within 1-2'" of the exhaust cross-over pipe, with some difficulty in holding the tubing away from the pipe. I was afraid of picking up too much heat from the exh pipe to make it worthwhile.

Larry
 
Last edited:
FI woes

Wish I would have caught this thread earlier but it sounds like you well down the solution path. Just another perspective about fuel injector location. I was helping with a Reno Racer project and we used the primer ports for the fuel injectors as well. Some Dyno testing showed some minor HP gains with that location. Also we didn't use a flow divider on top of the engine but a distribution block underneath the engine keeping the FI lines running uphill. Don at AFP provided the distribution block. As a further refinement several racing seasons latter switched to the turbo nozzles, mainly to get an upper deck reference at our Ram Air inlet because we were getting a good bit more MP having the Ram Inlet pick up air off the trailing edge of the prop. Poor mans supercharger. Several Nationsl records latter still works fine and all because we were running old narrow deck heads with out the provision for injectors.
 
theduff said:
Wish I would have caught this thread earlier but it sounds like you well down the solution path. Just another perspective about fuel injector location. I was helping with a Reno Racer project and we used the primer ports for the fuel injectors as well. Some Dyno testing showed some minor HP gains with that location. Also we didn't use a flow divider on top of the engine but a distribution block underneath the engine keeping the FI lines running uphill. Don at AFP provided the distribution block. As a further refinement several racing seasons latter switched to the turbo nozzles, mainly to get an upper deck reference at our Ram Air inlet because we were getting a good bit more MP having the Ram Inlet pick up air off the trailing edge of the prop. Poor mans supercharger. Several Nationsl records latter still works fine and all because we were running old narrow deck heads with out the provision for injectors.
Click to expand...

Some interesting ideas here. By moving to a distribution block below, I could cut at least 6" from each of my injector lines, maybe more. However, they would now be outside of the cool airflow up top (wondering of that would bring problems outside of idle). I could also eliminate 3-4 feet of -4 hose, which I am sure would do a lot to keep my fuel temps down. One issue would be how to secure the lines for stress reduction, though I could probably get at least one clamp on the intake tubes.

I know that distribution blocks were used for many years before the flow divider was introduced or at least not used universally early on. I am trying to remember what I read about it's benefits. I remember it being mostly idle and shut-down related, but can't recall he exact benefits. I thought it was related to equally dividing flow at lower PSI and positive shut-off.

Do you remember how that engine did in hot idle conditions?

I am not surprised that the primer port location produced more power. That port is in a much more favorable position to allow complete mixture of the fuel with the air. Coming from the top, too much air can slip by the bottom without mixing (relying more heavily on chamber swirl for mixture), especially with the low pressure and resulting poor atomization associated with the low pressures used in this system.
Larry
 
Last edited:
Thought I would post my progress. Idle was better, but not much. 90* days were still causing problems. I tapped off a return line from the servo. Ran it through a .8mm orifce and back to tank with an on/off valve. Figured that would insure cool fuel at least to the servo. That helped some, but still roughness. Today I installed a 4lb spring in the flow divider. Much better. Not perfect, but was able to do a 4 minute taxi after flight (90* ambient) with no significant events with the engine falling off 200 RPM. Pretty large improvement. Will need some more testing, but definitely better and can finely say it's not behaving badly.

Larry
 
Last edited:
Mixture

Leaning the mixture aggressively for all ground operations will normally result in much better low r/m operation.
 
jrs14855 said:
Leaning the mixture aggressively for all ground operations will normally result in much better low r/m operation.
Click to expand...

Thanks. My idle mixture is adjusted just a hair on the lean side of optimum (20 RPM drop at 800 - just lean of lowest MAP).

Larry
 
airguy said:
Could you expand on this just a bit? This is a standard Bendix FI, correct?
Click to expand...

Yes, Bendix RSA. The spider has a 2lb spring which develops about 1-2 PSI head pressure, pre-spider, at idle. The 4lb (might actually be a 4 PSI, not lbs, but not sure) spring delivers 4 PSI of head pressure. The increased pressure increase the boiling point of the fuel between servo and spider. It was created due to the high heat environment of the rear engine on a Cessna 206 (an SI was issued).

Don encouraged me see if I could get it running well without it, but I didn't quiz him on why. Given all of my struggles, I thought I would try.

I am still considering running the -4 line up between #1-#3, but I will need to pull the inter-cylinder baffle to mod it for a hole (the bracing pieces prevent me from just drilling it). It is some work to get it out and was hoping to do it when I pull the cylinders next winter.

I am also considering putting the spider under the engine, as Duff suggested. But that is a lot of work and won't do it unless I have no other choice.

Larry
 
You must log in or register to reply here.
Back
Top