Lord mounts J-7402-24

I've noticed that the #2 exhaust pipe has been contacting the lower cowl during flight or under high power settings.   Suspect the engine may be sagging a little.  The mounts have less than 300 hours on them although they've been subjected to turbulence.  I either need to replace them but will check with the group and a local IA when he's available.  These are the recommended J-7402-24 for the O-360.  The 370 weighs a little more than the 360 and is rated at 195 HP @ 2700.  

  -Update:  Lord specs indicate that dimension A should be more than .75" thick and 4 of my insulators were less than that.  At least one measured .68".  Lord tech support recommended J-7402-20 which are stiffer and also used on the Extra 200.  I've installed the new mounts and the engine seems less "wobbly".

   

Top isolator

 

Bottom isolator

Oil cooler exhaust vent.

Saw this on a Beech 17 cowl.  Not sure if it will help clean up the air coming out of the vent before hitting the prop.   I'd like to do some tuft testing and have a before and after comparison.

 

Starter damage, ring gear and broken teeth.

 Gremlins of a particular species like to break off ring gear teeth and kill starters.

Improper clearance between the starter and ring gear.  This is attributed to a known issue with new Superior ring gears and the Skytec 149NL starters having zero clearance. Destroys starters.

Flipped trigger wire on an EI coil.  This caused the wasted spark to occur on the bottom of the intake stroke causing a kick back.

Starting the engine without sufficient fuel pressure/momentarily bumping the starter switch.

Timing not retarding or advanced when starting and light weight props.

Only missing one tooth this time.

The first issue was resolved by replacing the starter.  Oddly, the issue didn't seem to occur until after several initial engine test runs.  I'd started the engine, was wearing a non-ANR headset and could hear a clattering ring ring sound.  I shut the engine down and quickly found the starter had not disengaged.  It shelled a bearing.  Although it had only been used a few times, it was out of a two year warranty.   A call to tech support at Skytec quickly diagnosed the issue was due to the dimensions that Superior and Skytec used to make these components resulted in no clearance between the gears.  A shim would be required to increase the clearance so that the starter would disengage properly.  

Aeroperformance aka Aircraft Spruce near me had starters in stock so I ordered one.  The price had gone up considerably since I'd purchased one several years ago  There is also a refundable core charge that is dependent on Skytec inspecting the core.  

I made a pattern from the starter's mating surface and fabricated a .020" thick shim out of 2024T3.  The starter was mounted and clearance checked based on instructions from some hot rod web sites.   The starter then disengaged properly after that.  

Shim for starter

So, back to first engine runs... The engine would start fine but then one day when starting....a nasty crunch & kickback sound indicated there were still gremlins.  Ultimately, this cause was attributed to one of the EI coil trigger wires being transposed causing a spark at the bottom of the intake stroke.  

All was well for months, but then one day when for some reason (distraction), I'd neglected to turn on the fuel pump(s) and I'd usually also been able to start the engine by just bumping the starter switch. This is my theory, but apparently when the fuel pressure is low and the injectors fire, only a minuscule amount of fuel enters the intake and if the cylinder that's on the bottom of the its stroke, finally accumulates enough fuel to fire, the engine kicks back.   Since that episode I ensure the fuel pumps are operational before starting and I keep the starter engaged until the engine cranks up.  It is odd that the shear pin in the starter didn't let go before breaking a teeth off the ring gear.  Fortunately others with the same EI/EFI and similar engines haven't had this problem.

The flywheel and ring gear have to pulled off the engine.  I asked some local pros if they'd done this before but no luck.  Searching the interwebs for replacing automotive ring gears found several methods.  One was to heat the ring gear with a hot torch to expand it and it would either fall off or with a few pops with a hammer, come off.  I tried this initially but after paint started blackening decide to try the cold chisel method.  Since this method breaks the ring gear, it made sense to help it out by cutting a slice as deep as possible without damaging the surface under the ring gear.  There is a lip that the ring gear is against so a round cut off wheel can only go so deep, but weaken the gear.  The assembly needs a very solid surface so that the chisel strikes are effective.  Place the ring gear retaining lip up so that the ring gear is driven away from it when it is struck from above. The chisel head is placed in the slot that was cut. A 3 lb. hammer, large cold chisel, gloves and eye protection are prudent.  I tried multiple light strikes initially but didn't make any progress.  Experience revealed on the last two ring gears, that it took three solid whacks to break and depart the ring gear. The ring usually moves a little on the first whack, more on the 2nd and breaks on the third. It comes off faster than a scared cat so heavy gloves are needed to protect your hands.  

Cold chisel and slit cut in the ring gear.

  

3 whacks later.

Installing the new gear is easy.  It's important to verify the tooth count is the same.  The U shaped teeth are on the 149 count gears and V shaped are on the 122 count gears.  The ring gears teeth are machined so it's important to mount the gear with the machined side towards the starter oriented away from the retaining lip.  Make sure all of the mating surfaces are clean and free of burs.  Heat the new gear (oven works fine) to 375 - 400F.  Put the other part in the freezer.  When you're ready...and wear gloves to handle the hot ring,  drop the ring gear on.  It is so loose it spins around but in less than a minute, it will have shrunk and secured itself.

The ring gear will fit in a countertop oven.

Fortunately I haven't had any issues for awhile but will keep watch for those gremlins.

Update - I installed a TVS diode across the starter motor relay contacts, suspecting that the starter motor EMF was causing a spike that caused an ignition spark on the bottom of an intake stroke.  The kick back seems to happen when the starter switch is released.

Baffling cracks

 At about 150 hours of flight time, 125 Tach.  After returning from formation flying found a couple of cracked pieces of baffle off the #1 and #4 cylinders.  The metal is .032" 2024.   I removed the broken pieces and will see if there's much effect on CHT's.   These parts were cut per plans.  -Update....I think the broken baffles should have been on the outside, not pinched by the screw.  I made new baffles but this time made a lip so the edge clears the screw where the crack occurred and stiffens the part.  A bead of RTV was applied to reduce fretting. 

The new baffle's edge was folded upward to clear the screw and stiffen where it previously broke.

 

The right cylinder baffle clears the 1/4" screw.  A bead of RTV should help reduce fretting.

SDS EM5 O2 controller

 The SDS programmer hadn't been showing the correct AFR since replacing the controller with the AEM controller and sensor.  I found that the white wire (AFR output) that connects to the SDS ECU had slipped inside the solder sleeve when I heated it, so was not making electrical contact.   After using a new solder sleeve, the programmer now reads 18 with the engine off.  I changed the VPX settings so that the AFR is powered on when the Alt Field switch is turned on.  The AFR controller takes a few seconds to run diagnostics.  It doesn't have a remote status indicator like the 14Point7 did.

I found a small crease in the baffling near #3 cylinder so used a leather punch and lacing tape to get the seal to lay down better.

Reviewing the EFIS data, after making lower cowl ramp changes, the cylinders coldest to hottest are now 1,2,3,4. With a 20-30 degree delta depending on air speed and power setting.   Previously, 1 and 4 were the coolest, 3 was the hottest with up to a 40 degree delta.   If #1 could be brought up a little and #4 down, the delta temp would be great.

Today I'll attempt to use my cheap new laptop to get some log data out of the SDS ECU to dial the map in better.  Based on fuel flow, it looks like with the mixture knob centered, about 10% more fuel is needed during take off to obtain max power.  

Cold Air Sump with SDS butterfly K&N Filter

Superior Cold Air Sump on a Titan 370.  SDS butterfly, Showplanes bracket for the throttle.  SDS air temp sensor is mounted on a bracket under the sump.

Top Engine View

 On a Sonex Podcast, Ross Farnham (SDS) said he had a fellow in Arizona shut his engine down, set for awhile and fired it right back up without vapor locking.  40 .lbs of fuel pressure has its advantages.  Although relocating the injection lines and block under the engine would provide cooler fuel.   I recently taxied to the fuel pumps on a 100+ day, fueled and fired the engine up with no issues.

An IA didn't like the injector hoses being Adel clamped to the push rod tube covers since the tubes can rotate.   So I put nutplates on the (.032") baffling and relocated the Adel clamps.   The IA approved.

Hard to see in this photo, but there's an Adel clamp connected to the baffling between #2 & 4 cylinders securing a fuel injector line.

First engine test

I'll edit this post as more data is gathered and issues are resolved.

The tires were aired up to help make it easier to roll the plane.

I put a gallon of fuel in each tank and checked for leaks.  The sump drain plug needed tightened on the copilot side.  I got disrupted before finding the leak and have a nice blue stain already.    A plug in the fuel pump assembly needed sealant and tightened.

8 quarts of mineral oil were added.   Spark plugs removed.  Fogging spray (lubricant) was shot into the cylinders.   The engine was easily spun by hand initially.    Then the starter was used to bring the oil pressure up and it peaked at 50 psi.   No leaks were detected.

The EI/EFI system was turned on/off several times to cycle the injectors while the fuel system was pressurized in hopes of purging any air out of the lines.   The injector line was temporarily removed from #1 with a quick cycle on/off of the fuel pumps and a shot of fuel came out of the line.  Additional fuel was added, to the left tank for a total of about 5 gallons in the left and only one in the right.

Several neighbors and friends assisted and observed, although I probably would have wanted to do this alone, some friends and neighbors wanted to be on site for the event. Some of the folks were really helpful with observation and analysis and for that I'm very grateful.

Wheel chocks and traffic cones were put in place when the plane was pushed out on 08/30/2021 6:15 PM.

Several people looked things over, loose items were moved to avoid anything going through the prop.  Some were wearing ear protection (this thing is loud).

Both master/buss relays were energized along with the A fuel pump, A & B ECU's and coils.   The throttle was set to idle, mixture centered.  Prop was clear, starter engaged.  Several combustion cycles occurred but the engine died.   A quick inspection revealed fuel was leaking at the connection to the fuel pressure sensor and the connection was tightened.  Some of the spark plug wires didn't seem fully engaged so were all rechecked with a few of them going on a bit further.

Another attempt to start was made resulting in a rather rough running engine.   The #3 cylinder appeared not to be getting fuel or spark.   We couldn't hear the #3 injector clicking when cycling the ECU off/on so I disconnected the injectors electrical plug and hit it a few times with 12 vdc.  The first time, it didn't click but each consecutive time it did.   We also noticed the #4 CHT probe was giving anomalous readings (rather cold) despite the head was hot to the touch.  The spade connections appeared good to the probe.

The engine was restarted with the throttle opened slightly.  Adjusting the mixture control didn't seem to have much effect.  The Dynon panel was showing 2x the RPM that the SDS programmer was showing and the MAP pressure on Skyview and SDS was suspiciously high (20 inches) with RPM now at about 1200.

A small amount of oil had begun to drip from the back of the motor and was determined to be at one of the AN-8 hose connections.  It was tightened.

We discussed the probable cause of the erroneous MAP pressure and suspect the port being used at the butterfly was to close to ambient pressure.  Two of the more mechanically experienced helpers advised that the port should be closer to a cylinder intake port, although they're unfamiliar with the SDS system, so I committed to researching this further and will test for leaks.

Update:  The Borla fuel pressure regulator itself was leaking vacuum.  I used some Loctite 680 on the nut/nipple/screw threads.  Tightened the oil hose near the oil filter.  Swapped 2 & 4 CHT connections, found another leaking plug in the fuel pump assembly.  Set the Dynon RPM tach settings to 2 pulses per revolution.    Ran the engine a second time, started up and idled decently at 800 rpm.   The Dynon tach more closely matches the SDS programmer but isn't exact.  The pad alternator doesn't seem to be charging.  Switching coils off/on, found Coil A (top plugs) didn't seem to be working.  Found it had been switched off in the VPX.  It normally is activated when VPX switch #1 is on and a panel switch is on. The CHT's seemed to be operating properly now so swapped probes back.

Next update:  Found a problem where the #1 coil was not firing.   Traced the issue to transposed white/white blue wires between the firewall and DB25 on the primary ECI to pins 9 and 11.  

Update 9/11/21:  Per instructions from Ross, updated the magnet positions, increased static fuel pressure to 45 psi.  Set the lower RPM advance to 20 degrees.  Tested redundancy by only turning coils on/off and switching the injectors.  Adjusted the mixture knob to -24% for a smooth idle at 700 RPM.  Main alternator is putting out but the pad alternator didn't appear to be putting out at idle.  Increased RPM to 1500 and it started charging.

Chapter 23 Engine Installation - SDS PC Datalogging info

Information regarding data logging from the EM5 ECU:

SDS PC Datalogging V6.3 links

Hi
Thanks for purchasing the SDS datalogging option. 

Two downloads are required for SDS PC datalogging to function on your PC. 
Please download and install each. 

1. Link for the USB to serial cable: 
WinXP:
https://www.silabs.com/documents/public/software/CP210x_VCP_Windows.zip
Win 7/8/8.1:
https://www.silabs.com/documents/public/software/CP210x_Windows_Drivers.zip
Win10:
https://www.silabs.com/documents/public/software/CP210x_Universal_Windows_Driver.zip

  

2. Click the link below to access the file *SDSDash_installer.exe*
https://cp.sync.com/dl/7244689e0#rjb23zqt-vciddca7-gpnr7a97-fhhximh6

V6.3 Required for V31 software to properly display RPM Fuel values which now have maximum of 350 instead of the older 255. Also fixes 1 missing manifold pressure fuel value when printing. Missing value was midway in the values set, also same with Manifold Pressure RET-ADV. V6.3 will work for older SDS software versions back to V25.

Version 6.2 is needed when using SDS version 30.0. V30 is used for the new 6.25" wide programmer released in Jan 2020. Also V6.2 will properly display the new LOP windows that appear in V30. SDSDash6.2 will work with older SDS ecu software version also.


Help:
Users of Windows 7,8,10 a popup box will say "Windows protected your PC" click on "More info" then click on "Run anyway" to continue the installation. 

SDSdash app scans for comports and typically the comport shown in the dropdown box is the correct one to talk to the cable, so you just have to click on "Open Com Port" and the PC and SDS will connect in less than 1 second. The app will download all your values from the Em-5 when first connected, and then begin to drive all the gauges on your PC screen. 

Click on the Log Chart tab and then click on the Go radiobutton about midway down left side then the data lines will begin scrolling across the Logchart display. 

You can save the Logchart data at any point in time. A Title/filename is created based on current date/time, but can also be edited to your liking. Two files get created, one to store logging, and a second file saving the downloaded SDS fuel and ignition values from the EM-5. Both Files will have the same filename but file extensions differ. 
File extensions are: 
.SDL is the data logging file. 
.SDV are Downloaded EM-5 programmed values. 
If you are sending files to us for viewing send both the SDL and SDV file, they go together as a set.

When opening a previously saved file, open the SDL file and the SDV file of the same name will automatically open simultaneously. 

Windows10 users: If you have trouble connecting, adjust the Data Interval(top middle of screen) number up to 200, then try clicking Open Com Port to attempt to connect to SDS. Windows10 appears to be very busy and may have trouble with short data intervals. Once connected you may be able to decrease the interval down into the low 100's, or even lower. Too low and data transfer may freeze, if this happens increase the Data Interval. Doesn't seem to be a problem with XP or Win8.1. We do not test with Win10. 

If you get an error requiring a 64bit operation please let me know. 
Please email me if you have questions! 


Troubleshooting Comport:
In windows Device manager you need to check and see if the SiliconLabs CP210x Device shows up under com ports. If it does not show up then you must install the Driver software. If the cable is showing up in your device list, then note the com port number, which will vary from one PC to the next. SDS can accept Com1 up to Com39.

Here's how to check for the SDS com cable in Win XP: Scroll down further for Win8.1+.

For Win8.1 when on your desktop, move mouse to upper right corner of desktop and a sidebar should appear on the right side, click on settings(looks like a mechanical gear), then click on control panel.

How to install SDS EM-5 data logger board.

Below plugging sub board into 4 pin header.



Below Fully plugged in connector up is showing.



Below connector board must have 1/16nylon washer underneath it. Use supplied 1/4" long screw in place of the original shorter screw which won't reach.




You may need to loosen the screw on the black transistor that hangs off the side. This can help the connector board drop into place. Remember to tighten the transistor screw.




Barry, iPhone 


-- Cheers, Barry.

Barry.