CIS to ITB EFI conversion

[winders]

Originally Posted by mb911

Yup low end will really suffer. Should be 1-1/2”

Yep...for a street car, I wouldn't use 1 3/4" until 3.8L. Even then 1 5/8" would work better for the RPM range you be using it on the street.

[dannichols1474]

Yes, both of the extension tubes come with an O2 sensor bung welded in.

Got down to the tool rental yard just as they opened this morning and rented a knock down engine hoist and a floor jack.

The engine hoist got the engine off the engine stand and onto my floor jack which had the Jack Adapter Tool (PP P/N: PEL-PW-2563) installed.

The second floor jack was helpful getting the transaxle lifted up so I could mate the transaxle to the engine, and once the engine / transaxle was in position under the car, to lift the nose of the transaxle so I could get the shifter shaft into the chassis tunnel.

Thank you to the wife who stood in as a shop assistant and did a great job.









Still have to make all the connections: axle shafts, oil lines, shifter, throttle, electrical (starter & EFI items, ground strap from transaxle to chassis), speedo cable.... plus re-install the rear bumper and side pieces.

[dannichols1474]

Today was a long day in the garage making all the connections to the engine and the transaxle. Everything in the engine bay and under the car is now connected.

I filled the engine with 15W-50 break-in oil and I filled the transaxle with 80W-90 gear oil.

I cranked the engine with the ignition and the fuel systems disabled and no spark plugs to pre-lube the engine prior to start up. The oil pressure warning light went out after 5 seconds of cranking - I cranked the starter 15 seconds, then stopped let it cool down 5 minutes then repeated 15 second cranking 2 more cycles. Each time the oil pressure warning light went out so I know I was building oil pressure, but the oil pressure gauge was reading near zero. I had installed a new "OEM" oil pressure sending unit from Pelican Parts, so I removed the new oil pressure sending unit and re-installed the old sending unit, then re-ran the cranking test, the oil pressure gauge was now reading 15-20 PSI during cranking.

I removed the muffler which I had only loosely installed for the engine / transaxle insertion into the car (I had to get the weight balance correct for the jack adapter I was using). Then I re-installed the rear bumper and rear side pieces and then installed the muffler. Once the muffler was installed, I tightened the v-clamps for the header extensions and installed the wide band O2 sensor and the O2 sensor interface unit.







Next up will be to adjust the clutch, connect the shift lever to the transaxle, install new spark plugs, and add fresh gas to the tank which I pumped dry 4 months ago, then I should be ready for engine start up.

[ToySnakePMC]

You are moving fast -- and it looks stunning! Great photos as always. I'm glad to read that you didn't panic over the initial oil pressure readings. Good on you to have the other sending unit on hand and trying that out as the possible fix. I imagine you were quite relieved to see actual gauge needle movement! Thanks for posting.

[dannichols1474]

Today I got the final connections made: adjusted the clutch, installed the spark plugs, installed the rear anti-sway bar, and rear tires. Finally after nearly 6 months I was able to take the car off jack stands.

Next, I went in the cabin and installed the coupler and tightened down the collar between the coupler and the shifter rod. I had marked the depth and orientation of the shifter rod on the spines of the coupler to make re-assembly easier.



I was ready to make the initial engine start up, so I made 3 trips to the gas station with my CA mandated 2.5 gal gas can to get fresh gas.

Now I started my pre-flight checks which began with a thorough inspection in the engine bay and beneath the car including double checking the engine and transaxle mounting bolts.

Next came switching the ignition on (key on, engine off) with the fuel pump disconnected, then connecting the MegaSquirt ECU to my laptop PC running Tuner Studio with the serial to USB cable I keep attached to the ECU. One, I verified that I was able to communicate with the ECU, and two, I wrote the updated tuner file (2803 cc vs 2683 cc, 7500 rpm stretch to AFR, VE, and timing advance tables) to the ECU.

Finally, I connected the fuel pump and switched the ignition on (key on, engine off) to check for fuel leaks. I heard the fuel pump start running but it sounded odd, then I looked at the pressure gauge on the fuel pressure regulator and it wasn't building pressure (it was set to 38 PSI). Then I realized why, fuel was gushing out of both fuel rails and down each injector. I turned the key off.

I loosened the 6mm screws on each rail to their mounting brackets and pushed down to make sure the injectors were seated in the rails. Then I re-tightened the 6mm screws and re-tested the fuel system - still leaking huge amounts of fuel.

By this time, I had a gasoline lake under my car.



Once the lake evaporated a bit, I removed one of the fuel rails so I could inspect the injectors. I found the upper o-ring on each injector to be very hard and shrunken - there was no sealing / snugness of fit to the pocket in the fuel rail. Meanwhile the lower o-ring was still flexible and supple, the bottom of the injector fit snugly into the pocket of the ITB, as well as, the fuel rail. Clearly, the upper o-ring that is normally always in contact with gasoline hadn't fared well after being dry for 6 months.





Thank you to Al Kosmal for sharing with me his source for new remanufactured fuel injectors: injectorplanet.com. I was able to order a set of fuel injectors. Now I wait for parts.

Thank God I didn't start the engine....

[al lkosmal]

Dan,
you were smart to do the mandatory fuel pressure/leak test, prior to trying to start the engine......I do that, as well as having two charged fire extinguishers at the ready....every time I do the initial start-up.

regards,
al

[dannichols1474]

Today at 8AM, I received a set of new fuel injectors - I decided to order new injectors and have them shipped overnight first delivery in addition to the remanufactured units already on order from injectorplanet.com (I'll keep the reman units on hand as back-ups).

I lightly lubricated the upper and lower o-rings with vasoline, inserted the injectors into the fuel rail, inserted the fuel rail / injector assembly into the ITB's, inserted the 6mm screws thru the fuel rail hold down brackets and into the fuel rail and tightened, then reconnected and tightened the AN-6 fuel supply line fitting to the fuel rail and repeated the procedure for the other fuel rail.

I reconnected the electrical plug to the fuel pump and turned the ignition switch to "On". This time, the fuel pump started and pressure in the fuel rails built up to 40 PSI on the gauge at the fuel pressure regulator and there were no fuel leaks. Fuel leak problem solved.



Later today I will be receiving new copper 12 x 16 x 1.5 mm sealing washers to replace the weak aluminum units that were supplied in the Pelican Parts sourced engine gasket kit and I will replace the aluminum washers for the copper washers on the banjo bolts and oil restriction adapters for the cam tower oil supply lines. I don't like the way the aluminum washers felt when I tightened the adapter and the banjo bolts and I don't want oil leaks there from the get go.

Following Al's recommendation, I bought a second A-B-C fire extinguisher, so I can have two fully charged fire extinguishers at the ready when I do the initial engine start up hopefully later today.

[montauk]

Originally Posted by dannichols1474

Following Al's recommendation, I bought a second A-B-C fire extinguisher, so I can have two fully charged fire extinguishers at the ready when I do the initial engine start up hopefully later today.

Great idea!!! BTW, I've been following along on this thread. Great work and good luck with the start.
Dave

[dannichols1474]

Yesterday, I received and installed the new copper sealing washers for the cam tower oil lines, ran one 15 second cranking cycle to check oil pressure, made a final visual inspection before start up, ran a final fuel pressure / leak check, then opened the hand throttle a bit and cranked over the engine. It started almost immediately and I adjusted the hand throttle to give about 2000 rpm. The engine ran rather rough, and had a hard time holding 2000 rpm, it either wanted to run at 1500 rpm or rev up to 3000+ rpm. Also, I noted the intake vacuum was reading 85 - 100 kPa between 1500 - 2000 rpm, I visually checked for sources of vacuum leaks but found none. I ran the engine for 20 minutes, then closed the hand throttle and let the engine drop down to idle and shut it off. First engine run to break-in the camshafts was completed.

I immediately jacked up the car and drained the engine oil - it had started out darkish amber and was now a grayish amber. I change the oil filter and added fresh 15W-50 break-in oil.

Next steps are to check and adjust the timing advance offset to make sure the timing commanded by the ECU matches the physical timing of the engine as measured with a timing light, and then balance the airflow thru the ITB's.

Since the vacuum signal with the S-cams appears to be weak, I have created a new project in Tuner Studio which uses alpha-n (TPS signal) for loading sensing rather than speed density (MAP signal) I was using with the engine before the rebuild.

[rwest]

Congratulations Dan!

I will be curious to how the TPS based tune works for you. I’m thinking about trying the TPS myself so I can free up the ports on the throttle bodies for idle air.

I am able to get around 50 kpa at idle on mine though.

[dannichols1474]

Thank you. Before the rebuild with stock CIS cams and the ITB's, I got 50-55 kPa at idle.

[Showdown]

Not for nothing but balancing the ITBs should be the first order of business otherwise everything else will be tail chasing.

[dannichols1474]

Originally Posted by dannichols1474

Next steps are to check and adjust the timing advance offset to make sure the timing commanded by the ECU matches the physical timing of the engine as measured with a timing light, and then balance the airflow thru the ITB's.

Since the vacuum signal with the S-cams appears to be weak, I have created a new project in Tuner Studio which uses alpha-n (TPS signal) for loading sensing rather than speed density (MAP signal) I was using with the engine before the rebuild.

After checking ignition timing, balancing the ITB's is the next step - I am pretty sure the left and right banks are out of sync, so I expect the rough running at 2000 rpm to settle down after I balance the ITB's as it did when I originally installed and set up the ITB EFI last year.

Only then can I get a real measure of the vacuum signal quality - thank you Showdown for your input.

[Showdown]

Curious to see what you do with your tune and TPS mode vs SD….

Idle at 85kpa is wild!

[Jeff Alton]

Those ITB's are sensitive to the alignment of all the brackets holding the fuel rail. They are FAR from precision. Had the same problem the first (and likely only time) we used that product.... was not the o-rings as injectors were brand new.....

Cheers

[dannichols1474]

Originally Posted by dannichols1474

Finally, I connected the fuel pump and switched the ignition on (key on, engine off) to check for fuel leaks. I heard the fuel pump start running but it sounded odd, then I looked at the pressure gauge on the fuel pressure regulator and it wasn't building pressure (it was set to 38 PSI). Then I realized why, fuel was gushing out of both fuel rails and down each injector. I turned the key off.

I loosened the 6mm screws on each rail to their mounting brackets and pushed down to make sure the injectors were seated in the rails. Then I re-tightened the 6mm screws and re-tested the fuel system - still leaking huge amounts of fuel.

By this time, I had a gasoline lake under my car.



Once the lake evaporated a bit, I removed one of the fuel rails so I could inspect the injectors. I found the upper o-ring on each injector to be very hard and shrunken - there was no sealing / snugness of fit to the pocket in the fuel rail. Meanwhile the lower o-ring was still flexible and supple, the bottom of the injector fit snugly into the pocket of the ITB, as well as, the fuel rail. Clearly, the upper o-ring that is normally always in contact with gasoline hadn't fared well after being dry for 6 months.

In my case the issue was not the design of the fuel rail nor their mounting brackets. It was due to the upper o-ring on the injector becoming hard and shrunken after being in service and in constant contact with fuel then being dried out for six months while I rebuild the engine.

When I installed new injectors with fresh o-rings, the injectors fit snugly into the fuel rails and the ITB's and the fuel leak problem was solved.

[rwest]

Hi Dan,

Nice work, if you don’t mind, could you show your VE table? Curious how you graduated the sensor position and also how you made the conversion from MAP to TPS. Did you use a data log to try and see what MAP value related to throttle position?

Thanks,
Rutager

[Showdown]

Something to consider;

I have been playing with using secondary table (or table switching, both do the same thing bus through different mechanisms) with speed density effectively creating a VE table with 16x32 bins of resolution.

My primary table goes from 35 kpa to 80kpa and my secondary table from 80-101kpa.

The amount of resolution this affords means that I can have individual rows for 85-101kpa giving me all the control I want and need. With my engine basically everything happens from about 84-100kpa and being able to precisely control the fueling for each step in kpa has meant that I don’t need to use ITB mode or Alpha-N.

Just food for thought.

[rwest]

Originally Posted by Showdown

Something to consider;

I have been playing with using secondary table (or table switching, both do the same thing bus through different mechanisms) with speed density effectively creating a VE table with 16x32 bins of resolution.

My primary table goes from 35 kpa to 80kpa and my secondary table from 80-101kpa.

The amount of resolution this affords means that I can have individual rows for 85-101kpa giving me all the control I want and need. With my engine basically everything happens from about 84-100kpa and being able to precisely control the fueling for each step in kpa has meant that I don’t need to use ITB mode or Alpha-N.

Just food for thought.

Hi Julian,

How do you accomplish having two tables. Do they change automatically, or do you need to have a switch or sensor?

Thanks

[Showdown]

Hey Rutager, et al…

Not to hijack thread but it’s topical. I can create a separate thread later with pics if there’s interest. I’m out of town for a few days though so it may be a week…

There are two ways; table switching and secondary table.

I wasn’t having great luck with table switching (jamie and I were comparing approaches in real time and he had luck but he has a PC and I have a Mac and we suspect there may be a software difference as we set them up the same.) but I had success with secondary tables.

Secondary tables and table switching are ostensibly used for boost applications but we can trick TunerStudio and Megasquirt into using them for load in non boost applications.

I’ll have to reference my TunerStudio setup and I’m away but IIRC, the setting is in “Advanced/Boost” and you turn secondary table on and set what y axis is uses (alpha-n, SD, %baro…) and it MUST be the same as your primary VE table 1. You then set the switch point (80kpa, 90% baro, etc) and whether the formula is additive or multiplicative. I use additive which adds the bottom row of VE table 3 (your secondary table) to the top row of VE table 1 (your primary table) this is only for secondary table not table switching.

If you’re using table switching that gets setup in general settings and it creates VE 1 and VE 2.

Then you tune. You have to run two VEAL windows, one on each table at the same time (it works) in order to autotune. But really, autotune should only be used to get a basic basic running setup and manual tuning should be the primary way (I firmly believe this for any efi system: autotune is just to get you in the stadium). Plus, you now have a lot more resolution so you really don’t want to leave it up to the computer which is kinda dumb.

It really helps to start off with a good single table tune and use those numbers in the bins on the two tables. Just transfer over the applicable values and manually interpolate (or you can use the auto interpolate) between the known points.

On my setup, the only odd part that took me a minute to figure out was the transition from VE 1 to VE 3. With additive I had to set the values in VE 3 as if I were adding them to the last, top row value of VE 1. So some of the bins on VE 3 were 2 & 3 because they were added to the previous VE bins of 50 & 51. If I used 52 & 54 the car was so rich it died because the actual bin values were 50 & 51 (from VE 1) PLUS 52 & 54 (from VE 3) making 102 & 105 which was way too rich. But by using bin value 2 & 3 on VE 3 they were added to 50 & 51 from VE 1 giving me 52 & 44… make sense?

It’s a bit more convoluted as you have two tables to manage now but the resolution is amazing.

Not super easy, but worth a try on a second, experimental tune.