CIS to ITB EFI conversion

[dannichols1474]

Today I addressed the vehicle code issue of lights for the license plate on my smoothie rear bumper. Since I wasn't up to fabricating my own solution, I did an internet search for a lighted automobile license plate frame and found Radiantz LED Lighting. Their license plate frame is USA made (Brookings, OR) cast aluminum with a black powder coated finish with 18 white LED's.

I removed the muffler to give me full access to the backside of the rear bumper.

Testing the LED license frame.



This unit uses a simple 2 wire connection: white for +12V and black for ground. I used crimp-on 1/4 inch blade connectors to connect the LED license frame wires to one of the original license plate light wiring pairs (I used the driver's side wiring pair). The original wiring pair is gray for +12V and brown for ground, so my connections are easy to remember: light to light and dark to dark. Of course if I get it wrong, the LED's don't light up as they are polarity specific.

I drilled a 1/2 inch hole in the bumper to pass thru the LED license frame wires.



I bought one of the wiring grommets for the rear bumper (p/n 911-505-347-00) and installed it in the hole I had made to protect the LED license frame wires.



I used black electrical tape and wrapped the LED license frame wires and passed the wires thru the grommet in the bumper. Then I assembled the LED license frame and the license plate to the license plate mounting bracket with M5-0.8 x 10 mm stainless steel counter sunk machine screws that use a 3 mm hex head to make it more challenging to steal.



Next, I made the wiring connections to the car wiring and secured the wiring with adhesive backed mounting squares and cable ties.



I replaced the muffler, then tested the LED license frame.



Now I think I'm done with my DIY projects for a while....

[dannichols1474]

Okay, at the end of an earlier post I said "Now I think I'm done with my DIY projects for a while....", well I guess a short while.

I decided to complete my 75 Carrera front end modifications by replacing the US spec front side marker lights with the Euro front side rubber molding pieces (p/n 911-505-307-00 left and 911-505-308-00 right) as I like their look better.

So I removed the front bumper again and removed the side marker light housings. I also dismounted the front marker light housings so I could pull the front bumper away from the car.



Then I removed the rubber accordion pieces that fit between the bumper and the front fender extensions. I loosely mounted the front side rubber molding pieces. Then I needed to trim off the rubber flap on the bumper end of the rubber accordion pieces with a utility knife and I remounted the rubber accordion pieces to the front bumper. Then I tightened the side rubber molding pieces and the rubber accordion pieces to the bumper.

Then I sat the bumper on the front bumper mounting struts and remounted the front marker light housings to the bumper and finally I remounted the front bumper to the car.

Here is the front of the car in stock US spec trim from April 2023:



Here is the front of the car today with the various modifications done (H4 headlights, bumper tucked 1/2 inch, fog lights removed, and side marker lights replaced with rubber molding pieces plus I reinstalled the right side rear view mirror which I had added in 2007 then later removed).



Good luck with your DIY projects and enjoy!

[al lkosmal]

MoBetta!

[dannichols1474]

When I did the front bumper tuck, I removed the fog lights and filled the mounting and electrical wiring holes with rubber grommets. I knew that was only a temporary fix so this week I got busy with a permanent solution.

I removed the front bumper and the front valance, then removed the chin spoiler from the valance then hammer and dollied the valance to flatten out the area around the center mounting hole. After sanding back to bare metal front and back around the fog light mounting holes, I hand formed two 3 1/2 x 4 inch pieces of .020 inch sheet metal to lay flat against the backside of the valance - one for the each fog light. Then I applied an even coat of JB Weld to each patch piece and placed them on the backside of the valance and held them in place with wood blocks and metal spring clamps.

After a 24 hour cure, I removed the wood blocks and spring clamps and sanded off excess JB Weld from both sides of the valance. Then I masked off the outside lip and applied two light coats of under coating to the backside of the valance and removed the masking tape.

Turning back to the front of the valance, I applied a thin coat of Bondo to each fog light area and waited for the Bondo to harden, then I sanded off the excess Bondo. After cleaning the front side of the valance with degreaser, I applied two coats of primer, then let the valance cure overnight.

The next morning I took the valance to the local body shop to get a sealer coat, base coat in GP White (code 908) and clear coat on both sides.

Two days later, I picked up the painted valance, reassembled the chin spoiler and then installed the valance and the front bumper. The front bumper project is now completed.

Hammer and dolly work



Epoxied patch pieces



Under coated backside



Bondo applied to frontside



Before photo



After photo

[al lkosmal]

Smoooooth, Dan! I like it!

[dannichols1474]

When I converted the CIS to ITB EFI, then later rebuilt the engine, I decided to leave the evaporative charcoal canister in place and to add a filter to the end of the oil breather hose that originally ran from the filler neck of the oil tank to the post air filter side of the air inlet box on the CIS.



Initially the air filter on the end of the breather hose worked fine with no oil leaks, but after 3k+ miles I started to see oil that had accumulated in the filter dripping onto the engine tin along the peak of the 4,5,6 cam tower and then on down to the lower front corner of the exhaust valve cover and I decided I needed to replace the filter with a breather tank.

After a short internet search, I chose a Moroso unit (#85475) that has an AN-16 breather inlet to match the 1 inch ID of my breather hose. I used a grinder to grind down the AN-16 threads to make the tank inlet a smooth slip fit so I could use the existing hose clamp to secure the breather hose to the tank. i also substituted a 3/8 inch pipe thread plug in place of the Moroso supplied drain valve as the install location does not have room to accommodate the drain valve.

The first step in the breather tank install was to remove the evap canister as I had decided to mount the breather can in the space occupied by the evap canister. I cut the vapor vent hose from the fuel tank and routed it to face down towards the ground in the gap between the rear shocks cross bar and the back of the cabin. I also cut out the canister purge hose that ran from the pipe fitting in the engine fan shroud to the evap canister, and I used a piece of Gorilla tape and a short length of the purge hose to plug the pipe fitting.

Breather tank mounting area:



Removed evap canister:



Plugged fan shroud hose fitting:



Next, I fabricated a mounting bracket between the inner fender well and the left mounting hole of the supplied breather tank mounting clamp; the right mounting hole of the supplied breather tank mounting clamp would be bolted to the existing evap canister mounting bracket. The fabricated mounting bracket was made from 1 1/2 inch wide, 1/8 inch thick aluminum flat bar stock cut to 3 1/2 inches in length. I made about a 20 degree break at the midpoint along the length of the mounting bracket to accommodate the angle of the inner fender. I used 1/4 inch sheet metal screws to attach the bracket to the inner fender and M6 x 20 mm bolts, washers and Nylock nuts to attach the breather tank mounting clamp to the left and right mounting brackets.





I didn't want to leave the pointed ends of the sheet metal screws in the right rear fender well, so I used a grinder to trim back and smooth the ends of the screws and hopefully prevent a catastrophic right rear tire failure.

[dannichols1474]

I have been chasing down and fixing small oil leaks on the 4,5,6 side and have it mostly fixed, but I got to thinking about the orientation of the oil distribution tube in the camshaft housings and decided to pull the intake valve covers and take a look.

Glad I did as I found I had the oil distribution tube on the 4,5,6 cam tower 180 degrees out of alignment - when I was assembling cam towers I got one of them wrong. When you have the oil feed fitting and the oil plug out of the cam tower as I had during engine rebuild, the oil distribution tube can rotate freely by hand.

An oil drip would form at the bottom of the exhaust valve cover at the #6 end:



Incorrect oil tube orientation on the 4,5,6 cam tower - the 2 visible holes should be pointing at the camshaft lobes not at the valve cover:



Correct oil tube orientation on the 1,2,3 cam tower, notice the hole pointing at the rocker arm oil lube hole:



This photo, if you look closely at the scratchy mirror image, shows the oil feed / alignment thru hole visible at both the oil feed and oil plug holes in the cam tower. The oil plug has a pointed end that fits into the thru hole at the end of the oil distribution tube and locks the orientation of the oil feed tube so oil goes where it is supposed to:



I rotated the oil distribution lube to the correct orientation in the 4,5,6 cam tower and re-installed the oil plug at the back end (# 6) of the cam tower to lock the oil distribution tube orientation, then I re-installed the oil feed fitting at the front end (# 4) of the cam tower and re-attached the oil feed line. Then I re-installed the intake valve covers and spark plug wires.



I'll see if not having oil sprayed at the inside of the 4,5,6 valve cover has any effect on my last small oil leak. And clearly having the oil sprayed at the correct locations on the 4,5,6 camshaft lobes and rocker arms will promote longer life.

Living and learning.

[dannichols1474]

After watching a situation develop since I installed ITB EFI and with only 4500 miles on the system, I concluded that the TPS (throttle position sensor) had gone bad and needed to be replaced. By gone bad, I mean the TPS had become increasingly noisy especially at the low % TPS end of the wiper.

With TunerStudio connected to the MegaSquirt2 ECU and the engine running, warmed up (no WUE - warm up enrichment), and at idle (950 rpm), the % TPS would swing from -0.5% to +1.7% and when I SLOWLY opened the throttle just off idle the TPS Acceleration Enrichment would start firing and the idle speed would become erratic and the injector pulse width would start jumping. When I let the throttle return to the idle setting, the system settled back down.

On the freeway, this behavior translated to bogging / hesitation whenever I was at very light throttle cruising along on level road (for example: at 65 - 70 mph (approx 3000 rpm)). Again, with TunerStudio connected, running and a datalog recording, I could see the TPS Accel Enrichment was triggering and the AFR was very rich at 11:1 with a target of 14.0:1.

The Green Spikes in the bottom graph are from the TPS Accel Enrichment firing:



Since the CIS to ITB EFI conversion kit I installed has ITB's from Race Hardware Developments, I went to the RHD website in search of information regarding the TPS used with my ITB's. I found the TPS part number plus a good discussion on why RHD recommends using Alpha-N (TPS) versus Speed Density (MAP sensor) load sensing with ITB's.



The RHD FAQ page also states that the MAP signal can only be used in conjunction with the Alpha-N. In TunerStudio, that option is on the General Settings page and is called "Multiply MAP (Caution!)". I tried running my engine with this setting set to "multipy" but my vacuum signal is weak and erratic due to the large degree of intake and exhaust valve overlap with the camshaft I am running, and I could not control the engine speed below about 3000 rpm so I run my tune with this setting set to "don't multiply".



So, I replaced the TPS with the BMW unit (13-63-1-721-456) listed on RHD’s website.

When I removed and checked the old TPS with a DMM it looked okay at DC, but I knew it was noisy and triggering from the Accel Enrichment spikes on the datalog I recorded.

Anyway, I ordered a new BMW TPS from Pelican which I received today. After checking all my ground and power wire terminations and wire connectors as okay, I installed the new TPS, performed the TunerStudio TPS calibration, checked for fuel pressure/leaks and started the engine.

TPS Calibration dialog box in TunerStudio:



Now % TPS at idle varies much less - +0.5% to +0.7%, I was able to reduce the TPS dot threshold from 25%/s back to 15%/s, and after running auto tune the car cruises down the freeway with the AFR right around 14.2:1.

I also found that with the new TPS and a stable % TPS signal that I had to add 0.4 ms to the TPSdot Accel Enrichment pulse width adder in order to achieve smooth tip-in response without bogging or intake back firing.



Old TPS (bottom, part number and other markings obscured) versus new TPS (top with BMW logo and part number):

[Showdown]

I had the same issue way back when with my kit- the TPS sensor was trash and causing all sorts of problems. In fact, I ended up replacing every component of my kit other than the ITB's with better, non Amazon-generic crap parts. Makes a big difference, whoda thunk it!

That's a really conservative AE. I'm surprised that you're adding so little fuel!

[dannichols1474]

Now that I have verified that a new TPS has solved my ITB EFI drive-ability problems, I have several take-aways I want to share.

1. If I were starting the CIS to ITB EFI conversion project today, I would have set up my MegaSquirt2 tune file using Alpha-N (TPS) load sensing instead of Speed Density (MAP sensor) load sensing. Even with the relatively benign camshaft in my stock 2.7L CIS motor, I could not get as smooth a tip-in, or throttle transition response, using Speed Density load sensing as I have now with Alpha-N load sensing.

2. I have ordered another BMW TPS (13-63-1-721-456-BOE on Pelican) to keep in stock as a spare (I choose to pay the extra $$ for the BMW branded part instead of the less expense alternatives). I decided it would be a good idea to have a collection of critical spare parts on hand for my now rather bespoke 2.8L hot rod engine. Along with a spare TPS, I have spares for the Pertronix ignitor, the 14Point7 O2 sensor interface, and the wide band O2 sensor.

[al lkosmal]

Dan,
as usual, you go about debugging issues in a data driven, fact based manner. I appreciate the feedback, especially when provided with good data to back up your findings.

I've used quite a few TPS sensors, typically aftermarket,... with good success over the years, but occasionally not.

I also found the OEM BMW sensor recommended by Rama at RHD to work very well, but at a much higher cost then the aftermarket units.

My focus has always been to provide a cost effective EFI kit for the DIY crowd and good price/performance is still my focus....however, if a component provides inadequate performance, the lesser initial cost is misleading.

Thanks for the heads up ....in this case, providing the OEM part makes the performance worth the added price.

regards,
al

[montauk]

Is the stock BMW sensor a standard potentiometer? Is so, can you use a hall effect sensor? Non-contact sensors should last for a very long time.

For my TPS, I sourced the Variohm sensor from Digikey—it features a standard D connector and fit my BMW throttle motor. It's also dual sensing which is needed for DBW.

https://www.variohm.com/products/motorsports-sensors/rotary-position-sensors-for-motorsport/euro-xpd-d-shaft

https://www.digikey.com/en/products/detail/variohm/EURO-XPD-2832-812-214-911-00/22286984

[dannichols1474]

Originally Posted by montauk

Is the stock BMW sensor a standard potentiometer? Is so, can you use a hall effect sensor? Non-contact sensors should last for a very long time.

For my TPS, I sourced the Variohm sensor from Digikey—it features a standard D connector and fit my BMW throttle motor. It's also dual sensing which is needed for DBW.

Yes, the RHD recommended TPS is a potentiometer. Yes, you can replace the potentiometer TPS with a Hall effect TPS. You may need to fabricate a mounting plate to mount it to the RHD ITB. The RHD ITB's in my set up are not DBW, so dual sensing is not required.

[montauk]

Originally Posted by dannichols1474

Yes, the RHD recommended TPS is a potentiometer. Yes, you can replace the potentiometer TPS with a Hall effect TPS. You may need to fabricate a mounting plate to mount it to the RHD ITB. The RHD ITB's in my set up are not DBW, so dual sensing is not required.

I don't think the mount would be a problem. Mine is fitted to a BMW throttle motor. In fact, I removed the BMW sensor from the throttle motor. I imagine I tossed it. Too bad, I would have given it to you. I've learned a lot from your thread.

[mikesarge]

I also had problems with the non-oem TPS (two broken aftermarket units, two ruined track days). The BMW unit was much more stable. I think the potentiometer is more ‘curve’ than linear though, and it still gave me problems with tuning (persistent backfires at low-rpm tip-in) so I ditched it for a hall effect sensor which allowed me to use Alpha-N with a lot of success.

[rwest]

Dan,

Can I trouble you to post a picture of your TPS VE table?

I’m going to switch to TPS myself soon.

Thanks,
Rutager

[dannichols1474]

Originally Posted by rwest

Dan,

Can I trouble you to post a picture of your TPS VE table?

I’m going to switch to TPS myself soon.

Thanks,
Rutager

Hi Rutager,

Here are screenshots of my current Alpha-N tune with the new TPS - I still need to do some more Auto Tuning in the cruise region of the VE table.

[dannichols1474]

Hi Rutager,

Here is a close up of the VE table, the other screenshot was too small.

[rwest]

Thanks so much Dan!

I haven’t done a ton of comparisons on my MAP versus TPS to get a basic tune, but I wonder if making a few of the sections in normal cruising range closer together just like you have done near idle would make a smoother running car?

[dannichols1474]

Originally Posted by rwest

I haven’t done a ton of comparisons on my MAP versus TPS to get a basic tune, but I wonder if making a few of the sections in normal cruising range closer together just like you have done near idle would make a smoother running car?

So, yes, making the vertical scale (% TPS Load) smaller steps near 0% TPS gives you finer control for tip-in as the engine moves off idle and I did that with both the VE and the Ignition Advance tables - that was a recommendation from Al and it really works.

You can probably stretch out the horizontal scale a bit as you are probably rev'ing to 6500 rpm max?

Before you make the switch be sure you know what the kPa is for idle and for cruising on the highway, level and steady, at say 65 mph. Let's say it's 50 kPa at idle and 70 kPa at cruise. Those two load levels in Alpha-N will correspond to say 0-1% TPS for idle and 10% TPS for highway cruise.

Now you can take the cell values from the 50 kPA row of the Speed Density VE table and put them in the 0% TPS row of the Alpha-N VE table.

Next take the cell values from the 70 kPa row of the Speed Density VE table and put them in the 10% TPS row of the Alpha-N VE table.

Since 100 kPa is wide open throttle on a naturally aspirated engine, then take the cell values from the 100 kPa row of the Speed Density VE table and put them in the 100 % TPS row of the Alpha-N VE table.

Then take the rest of the cell values from the Speed Density VE table and fill in the rest of the Alpha-N VE table. You won't be using any of the cell values for rows with less than 50 kPa (or your actual idle kPa value) on the Speed Density VE table in the Alpha-N VE table as these rows represent % TPS values less than 0 % TPS.

Use the same procedure above for the AFR table and the Ignition table.

Then I used "Tune Analyze Live!" (auto tune) while driving the car over a range of conditions to tweak in the VE table.

Then if you choose, you can take the car to a dyno shop and have a calibration technician optimize the VE and Ignition table values for max torque and hp.

This is the procedure I followed when I converted my tune from Speed Density (kPa load) to Alpha-N (% TPS load).