I. Engine Start with M62 Supercharger

                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                               Send comments to mostlymiata@gmail.com                                                                             


Other Links





Engine Start Up

with

M62 Supercharger




We finished the M62 Supercharger installation on 08/06/2016.  


There were several tasks to be completed when Matt arrived.  Matt took care of a hanger issue on the exhaust.


Meanwhile, I made a bracket to hold the bypass vacuum acutator open so we could run the engine and shake down the supercharger install before we started running boost.  It took about an hour for me to make this little bracket.  Lots of bends and adjustments were required to get it to fit.








This is how the bracket mounts.





Matt had purchased a Nitrous Outlet 5/16" Pushlock Fuel Adapter Fitting from Summit Racing.  He planned to use it to install a sending unit for the fuel pressure gauge that is still in transit.  Since we needed to be sure of the fuel pressure with the new pump, Matt also purchased a mechanical fuel pressure guage which we installed temporarily.




After looking over the engine bay, we primed the fuel pump by inserting a jumper in the fuel pump and ground pins on the Diagnostic connector.  The pump ran for a few moments and the pressure rose to about 60 psi.  Best of all, no fuel leaks on the injectors!




We then connected a laptop to the FIC to make a log while starting the engine.


We were now ready to start the engine.  It took two attempts but the engine started and ran smoothly.  We pulled the car into the driveway to minimize the fumes in the garage and let the car warm up.  We blipped the throttle a couple times to be sure we were staying below boost.


We also made several logs while the engine idled so we could review them later and have a base to start tuning.  The engine was up to operating temperature and running well. We did have some idle droop that would need to be addressed later.


We shut the engine down to eat lunch.  When we returned we decided to make one short run to verify we would stay out of boost for now and to make some more logs.  The engine ran pretty good on the short run.  


Everything worked good.  We made some logs and did not have any boost.  Then when we returned home.  After  the engine had been shut down, it would not restart.  Now we had our first challenge.  We were not sure why it would not restart. 


We were not getting much injector response while trying to start.  We pulled a plug.  No gas.  Listening to the injectors through a long screwdriver revealed clicking like they were firing.  We checked everything we could.


Since there was no gas getting through we decided to remove the temporary fuel gauge. Now we ran into a new problem.  The fuel adapter fitting with the gauge attached to it would not release from the car.  Finally, the clip inside broke enough for the fitting to pull loose, but leaving the clip firmly attached to the steel fuel line.  It had not been attached properly from the start.




I carefully clipped the edge of the clip's ring and was able to tear it off the steel fuel line. The fitting did some damage to the fuel line, so we carefully cleared the burrs with a small file and sandpaper.  It did clean up enough to fit for now without leaking.  I sure hope that lasts.  Matt will get with Nitrous Outlet Monday to see what went wrong.


Frustrated with all the problems, we retreated into the comfort of the house and started looking at the logs.  We studied the start logs again and compared the earlier start log to the no start log.  Matt wanted to increase the fueling on our 13.7 and 0 rows below 700 rpm to see if it would start.  I was not sure that was the problem.  


I put a call in to my friend Roger.  He confirmed Matt's diagnosis.  He had found that the on start up with the engine warm more fuel was needed.  His map had that in it, but I had overlooked it when making our map.


Our map has to reduce the injector pulse by 58.5% to match the 640cc injectors to the original 265cc injectors.  So no change to the OEM ECU output means the FIC reduces the signal by 58.5% by using a -58.5 in the row where zeros used to be.  To make the cells richer, we put a zero back into the first few cells for starting.


The engine would now start again!  Thanks Roger for the help.


We went out to make a few more logs.  The engine did die several times when stopping. It did restart but we did not want to get stranded so we headed home.  When we pulled into the driveway, the engine died again. 


This time we could not get the engine to start.  It was very hot outside too so we just pushed the car back into the garage.  We then spent the rest of the afternoon trying to locate the problem.  Back inside, we started searching why Miatas may not start.  I called Roger again.  He and I had the same experience here.  No start - replace the CKPS and CAS if that does not work.


We lifted the car and spent about an hour replacing the CKPS.  It is difficult to do since the radiator is now a lot closer to the engine.  The engine still did not start.  We did check to see if the tachometer was registering, but it did not.


We replaced the CAS.  Still no start.  This time Matt noticed that not only the tachometer was not reading, we did not have a CEL light on when the key was turned on either.  We knew without the CEL working the engine would not start.  Back inside again, I started reviewing the wiring diagrams for a 2000 Miata.  From that I was certain the issue had to be the 30A fuse serving the main relay, or the relay itself.  It was late now so we would check that out in the morning.


I awoke at 4:30am on 08/07/2016 thinking about the engine issue.  I went out to the garage and started checking the suspected circuit.  On the CAS there is a white wire with red stripe that should have 12V on it with the key on if the relay and fuse is working.  It had nothing.  


The 30A fuse is in the black fusebox in the engine bay on the passenger side.  It checked good.  Next to it is the Main Relay.  It had some tape on it.  I remembered that we had taken that relay apart in the past to clean the contacts and I put the tape on it to keep out moisture and dirt.  It pulled out of the fues box easily since we had broken the retainer tab already.




With the cover pulled off the relay, I pressed on the contact arm.  It would not move.  I put more pressure on it and it clicked hard as it pulled away from the plastic housing. Apparently, the contacts had gotten so hot that when we turned off the key, the relay contact pressed against the plastic housing and welded to it.  The relay magnet could not close the contacts.




I  put it back into the car and turned the key on/  The CEL was now back on.  


When Matt arrived, we connected the laptop and logged the start.  It started right up!


The main relay powers the injection system, some of the ECU and most importantly, the fuel pump.  The new pump draws more current than the old one.  Probably about 10-12A (up to possibly 18A deadheaded) instead of around 8A.  So, we started making plans to add a separate fuel pump relay with a direct feed from the battery.  This would relieve the load on the OEM relay and ensure good voltage at the pump.


Our plan was to install a 30A relay in the trunk, feed the new relay contacts directly from the battery with a 15A fuse, 12 gauge wire and use the OEM power feed to the pump to turn on the new 30A relay.


We already had some 12ga wire, a 30A relay,  and relay socket from Matt's !@#$! VW project (there is no such thing as just a VW).  We only needed some eyelet connectors and a fuse holder.


The battery was disconnected for safety.




We took the rear deck cover off to allow us to run wires to intercept the OEM fuel pump feed.





The OEM fuel pump feed is the red wire with lavender stripe going to the fuel pump connector.  We intercepted it outside the tank cover using crimp on connectors.




All the connectors were covered with heat-shrink and then the wires were covered with split-loom.






We made a run to Advance Auto parts and got some connectors and a fuse holder.




The relay was mounted and grounded to the trunk hinge bolts.




The fuse holder was attached right at the battery.




Now we wanted to make a few more logs and go get some fuel.  We drove about 5 miles to town and filled up the tank, making logs as we drove.  When we were ready to head home, the engine would not start...AGAIN!


We pulled the relay and checked it.  It seemed OK.  We had adjusted the cells we put zeros in to a lesser value.  So we moved them back to zero.  The engine started.


When we returned home I pulled the Main relay from my 99 AE and we installed it in Matt's car.


One note.  It takes the proper tool to remove a main relay without breaking the locking tab.  This is the tool I made out of a piece of steel wire to remove the relay.  It is inserted between the relay and the fuse block on the fender side.  Once it is down far enough, it can be twisted to push the locking tab on the fuse block back, releasing the relay. Otherwise, the tab will be broken getting the relay out.




We now wanted to run the car in boost before the day was out so we headed to Matt's house 20 miles away to get the hardtop.  It is much easier to listen for knock with the hardtop on the car.


We had not yet installed the hood.  On the way back, we ran into several rain showers. The last one caused the engine to start sputtering.  Luckily, we were almost home and made it...barely.


We blew off the ignition with air to dry out the connections and spark plug holes.  That cured the rough running.  The hood was reinstalled.  We should have done that earlier.


Now the our boost blocking bracket was removed from the bypass valve.  The boost alarm was set to go off at 20.0 psia (just over 5 psi) and to shut off at 19.7 psia (5 psi).  It was time to try to run some boost!


Also, Matt ordered a new main relay and a 46" drive belt for the supercharger.  The 46.5" belt is just a little long.  They should be here during the week.


Off to do some tuning!  See the tuning entry for more details.



Update 08/20/2016


On 08/08/2016 Matt called Nitrous Outlet to tell them about the fuel line fitting and see if they knew what went wrong.  Without hesitation, they pointed out that the earlier version of the fitting had that issue.  They thought all of them had been taken out of stores, but Summit Racing apparently still had some.  They are sending a new, improved version to replace the fitting.  


Well, during the week, a new fitting arrived from Nitrous Outlet.  It was not the new version they had promised to send.  And it had the same problem fitting as the one we had break before.  This time when Matt called Nitrous Outlet, they were not very helpful.  They said they would not send any new improved versions out until all the old versions were gone. 


We test fitted the one they sent and it would not fit.  One of the ears on the clip was about to break anyway.  So, Matt returned it to Summit since it was useless to us.


In order to get the new gauge to work, we had to make a new keeper for the original fitting that had failed.  It took almost half a day to get it working.  We would not buy a fitting like this again!


To see the fuel pressure gauge install, go to E. Supercharger Gauge Set Intall 


The new kevlar belt arrived and I installed it on the supercharger.


  


This is the clearance with the old belt.



And this is the clearance with the new kevlar belt.




It does give a little more clearance for the autotensioner to work with.  Maybe it will not stretch like the other belt did.


Again, off to do some tuning!




Update 08/27/2016


We were able to work on the engine and tune again today.  Matt had been driving the car without running in boost for the past week.  He had noticed some problems that needed to be addressed.   Matt also had been corresponding with Ken Hill about the logs we had previously made.  Changes were made to the fuel map and new logs were needed to see if the changes had the expected effect.


First we addressed the misfire code for cylinder number 3.  We pulled the COP and the spark plug and found no issues.  To test the COP we swapped it with the one on cylinder 2.  While making new logs, we did not get the code again, so we belive it was just a bad connection.


The engine had too much idle droop, especially. when stopping from speed.  I adjusted the flow control vavle on the IAC block out about 2 turns.  Now the engine rpm drops slowly from 1250 rpm to idle speed.  The engine has not died on stopping since the adjustment.  We hope that fixed it!


We made a new log while holding boost at 5 psi from 2000 rpm up to about 6500 rpm to see if the lean spot we were targeting had been addressed.  That area may be a little too rich now, but that can be addressed with another change to the map.  The good news there is that it did have the desired effect.


Matt had noticed that when the engine was running about 3000 to 4500 rpm in closed loop, without boost and under light loads it had a slight, continuous stumble.  So, he tried making that area of the map richer.  It did help, but we are not sure if that fixed it. Hopefully, Ken will be able to help us there too.  So we also made a log of that issue to assist us with changes.


We also noted on the new fuel pressure gauge that we consitently are running about 66 psi of fuel pressure instead of 60 psi.  This may mean we need to adjust the injector dead time down a little to 1.2 ms instead of 1.296 ms.  That may also have an effect on the stumbling we are having.   We may just have to try that and see.


More to come as it happens!





   

Vendor Links







Comments