911 CIS (KJET) to EFI

[bumble]

I bought my 1980 SC in the knowledge that it would need surgery on its injection system which was in pretty ordinary shape.  CIS on Porsche is now 30-40 years old.  Some parts are getting hard to find and can be very expensive - plus there is a certain amount of voodoo involved in getting them to run properly if they go wonky.  Further, CIS doesn't take kindly to aggressive engine mods.

My engine will need a rebuild - it's tired and has at least one broken head stud.  I can't fit a rebuild/upgrade into my schedule for at least 18 months but still want the 911 to be available (and reliable) in the meantime - so what to do about an iffy CIS?  This was discussed a while back here: http://porscheforum.com.au/index.php?/topic/3110-going-efi/#comment-45387

I didn't want to rebuild the CIS and was looking for a solution that was cheap (relatively), proven and available.  I ruled out carburettors because of their horrendous fuel consumption and general ancientness as well as not fitting the 'cheapness' criterion.  ITBs would be nice, but a turnkey solution (e.g. Clewett) is north of $20k.  Anything else has a whiff of guinea-piggery about it.

I decided to use the Bitz racing solution (www.bitzracing.com).  The kit is relatively cheap (about $US1500 when I bought it a year ago).  I comes sans injectors and fuel pressure regulator. The injectors are cheap as chips and I sourced a fuel pressure regulator locally (Japanese manufactured).  Installation is fairly straight forward but removing the CIS with the engine in situ is a bit of a chore.  I also elected to do a heater backdate and clean up the engine bay somewhat.

After the install, the car started OK but had a stumble at about 4000rpm and wouldn't rev any higher.  Merde!  Much head scratching.  I sent a Tuner Studio log to Tony Britzanis (kit supplier) who said everything looks OK but it looked like the ECU missed a trigger event.  The ECU triggers off an adaptor connected to the coil primary. Connecting a scope to the adapter shows the signal to be a complete mess with a lot of ringing.  How the ECU is expected to resolve a signal like this is anyone's guess.  Anyway, I designed a new adapter to run directly from the reluctor to give a nice clean square wave and bingo! - instant 7000rpm.

Here's a quick pic.

[firstone]

Very clever. Well done.  ?

[OBRUT]

Very cool Mark

ps. Long time no see.

[bumble]

Yo Ryan!  Yeah - been a bit quiet lately but intend getting back in the saddle next year...  How about a catch up sometime?

[Coastr]

Reluctor? 

Also, you kept the air box and CIS meter thing?

[AC78]

So does this setup offer any improvements over CIS, such as HP, fuel efficiency, reliability etc?

[bumble]

Hi Coastr - some people want to retain the CIS look (which is what I did due to general laziness).  The fuel distributor is replaced by a metal plate whose mounting screws hold the meter permanently fully open.  Others chop up the whole airbox and install a conical air filter.  On the bitz site you can see various examples of different installations.  About the only active thing left from the CIS is the AAR.

Reluctor - the thingy in the distributor that produces a signal to trigger the ignition unit (replaces points).  In its raw form, the reluctor signal is not suitable to trigger the ECU(I don't have a sample to hand - sorry).  My adaptor takes this signal and produces a waveform (shown below) that the ECU can easily resolve.  Quite why my ECU doesn't like the >4K signal from the Bitz adaptor is a mystery.

[bumble]

Hi AC78 - good question.  I've done this primarily for reliability - not that CIS is inherently unreliable - just that mine's 35 yrs old, probably never been touched and was exhibiting well known symptoms of CISitis (warm/hot start problems).  Full CIS rebuilds can be very expensive and I have future performance plans that preclude the use of CIS.

Before I started the conversion, I took careful records of fuel consumption and had a dyno pull done.  I intend to do a tuning session on the dyno soon and will be able to compare.  Comments from others who have done this indicate better fuel consumption and more power.  We'll see...

[Cheshire Cat]

Whoa! That I suppose is my next move

who's tuning the program etc?

[bumble]

So I find a guy who's got a dyno - in Goulburn about 45 min from where I live - and arrange a time.  Rock up at the appointed hour.  Workshop is full of V8 engines in various states of repair.  Several dirt track sprint cars strewn around.

G'day, I'm Billy sez a guy in overalls.

Mark sez I.  Need to get a base dyno reading before I convert it to EFI.

Right, never seen one of those - how many cylinders did you say?

Six sez I.

Right, where do we put the cooling fan?

Not needed sez I - it carries its own.

Scratches head - absorbs information - raises eyebrow.  Right, well lets run 'er up.

2 full throttle runs and the workshop is full of smoke.

Don't worry sez I, they all do that - they're designed that way - that's why they carry eleven litres of oil.

More head scratching - raises eyebrows - really?

Absolutely sez I.

If you say so.  Here's the print out - 150hp - not bad.  How heavy did you say?

1160 kilos sez I.

Clanking of mental abacus - more eyebrow raising.  Not bad.

 

There are two (that I know of) softwares that can tune the MegaSquirt.  I have no sodding idea of how to tune EFI, but between Billy and me, I'm sure we can figure it out.  The conversion comes with a narrow band O2 sensor which you install in the pre-muffler.  I'm in the process of building a WB controller to work with an LSU-4.9 WB sensor which should aid greatly in tuning.  The sniff test indicates it's running rich most everywhere.

The butt-o-meter tells me it's stronger than it was with CIS.  Or it could be just confirmation bias...

[Coastr]

My butt-o-meter detects an extra 10hp whenever the valves are adjusted.  Notoriously unreliable things, butt-o-meters.

[Cheshire Cat]

That does sound good. 

[Barks]

I think I need to hand in my "Man Card"...I wouldn't even know where to start!! Great job.

[MFX]

Very interesting. I will be looking at doing a similar thing to my 2.7 in the future and it looks like you are local to me. I am in Bowral and would love to have a look some time. 

[bumble]

Hi MFX -I'm in misty Bundanoon.  Normally here Fri thru Sun.  Just PM me whenever...

Mark

[Pfitz911]

Hi bumble

i have a 77 911s and have looked in to this. It looks really good. I want to keep the CIS look with the air box so it passes a rough visual. what I don't understand is how doe this affect the combustion? Don't CIS cars have specific pistons to "swirl" the mix specifically for the CIS system? The tank also has the swirl "thingies (!)" for CIS, is it fine with this? I suppose you don't need the accumulator?, just keep the filter. 

it looks like a good kit. Would you recommend it and was it easy to install (I have minor mechanical experience, but keen to tackle if it's achievable).

hp improvements?

sorry for all the questions!!!

[bumble]

Hi Pfitz,

Sorry for the late reply - only just got back to this.

The combustion is not affected by whether gas is hosed into the port via an electrically operated injector or a mechanically (pressure operated) injector.  I'm not aware of anything in the tank that would cause a problem with EFI.

The accumulator is retained - I simply replaced the filter.

Whether or not this is a good solution depends on your specific circumstances.  If I had a perfectly functioning CIS, I wouldn't bother.  If originality/authenticity is important but I had CIS problems, clearly not - you're better off getting the CIS fixed.  If performance mods are on the cards, CIS limits your options which was a secondary reason (after reliability) for me going down this path - I have a cunning plan...

Of the other options, carburettors are a popular more or less turnkey solution but with horrendous fuel consumption and limited tunability over a range of operating conditions.

ITBs (Jenvey, TWM, PMO etc.) and complete turnkey ITB packages are available (Clewett for example) but are VERY expensive by the time you get it all together.  For a standard or only slightly modified engine, it's difficult to justify that cost considering it won't add much to the engine performance.  Mind you, if bragging rights are important, they look the duck's guts...

Some have retrofitted Carrera 3.2 components.  How this would work on a 2.7 I don't know, but it would hardly be turnkey.

Some have rolled their own from motorcycle components (Triumph TBs) - but here you're getting beyond the capability of most home DIY guys.

IMHO, the Bitz kit is quite good value.  You eliminate the troublesome CIS bits but get to reuse a lot of the original CIS.  You get an ECU that is already set up.  You get fuel rails and mounts fit for purpose.  You get all the correct sensors.  You have to source injectors and a fuel pressure regulator (FPR).  Total cost about $AUD3K.

I changed three things: 1) The kit supplies three in-line blade fuses.  I installed an auxiliary fuse panel with LED indicators.  2) The kit recommends a certain type and mounting for the FPR.  I sourced a SARD unit locally (from EFI Hardware in Melbourne) with an oil-filled pressure gauge.  3) The kit supplies a device to trigger the ECU from the coil.  As explained in a previous post, this didn't work well for me so I hade to design a new trigger mechanism.  It would be even simpler on a points system like the 2.7.  The instructions outline modifications from other installers regarding the position of the Air temp. sensor and the engine temp. sensor both of which I have followed.

Installation is not difficult.  The most difficult (awkward) task was removing the CIS while the engine is in the car and then putting it back in.  Having done it once, I could now do it in a quarter of the time!  Some installers have even removed the mechanical injectors and installed the EFI injectors with the CIS in the car!  Others have done a partial engine drop.  Personally, I think it's better to persevere with removing the CIS.

Tuning is now the next step.  It drives well but feels a bit rich.  For an EFI newbie, this is a challenge.  I'm in the process of building a wide band O2 controller and display (don't ask - others are probably better off going to Tech Edge and buying one!).  Talking to a couple of the Porsche shops in Sydney, they only tune from a WB O2 sensor rather  than on a dyno.  Ultimately, the narrow band sensor that comes with the kit will be reinstalled, but I've designed a small, cheap LED bar display to monitor the AFR from the NB sensor just in case.

Everyone that's done this has reported more power and better fuel economy.  I have yet to measure these.  There is a Bitz forum on Yahoo if you need further info.

 

Anyway, hope this helps and good luck.

Mark

 

[Pfitz911]

wow great reply

thanks for that. I plan on rebuilding the 2.7 as I was going to put a 3.2 in. I think it will tip the budget over if i purchase a 3.2 outright, not sure yet. anyway Ill check out the forums.You did notice more power though didnt you??

cheers

[Zelrik911]

Hi Pfitz,

.  The most difficult (awkward) task was removing the CIS while the engine is in the car and then putting it back in.  Having done it once, I could now do it in a quarter of the time!  Some installers have even removed the mechanical injectors and installed the EFI injectors with the CIS in the car!  Others have done a partial engine drop.  Personally, I think it's better to persevere with removing the CIS.

 

Anyway, hope this helps and good luck.

Mark

 

Hi Bumble

Can I ask you about "Removing the CIS while the engine is in the car and then putting it back in"?

I am considering doing this to get at oil leaks at back of my motor.

You say that you could do it much quicker now that you have experience.

Do you have any tips &/or traps? For example - did you take out any single inlet runners or pull it all out in one piece?

Thanks, Peter in Melb. 

[bumble]

Hi Peter,

I did mine with zero documentation or factory diagrams, so some things may be obvious if you have this documentation.

Extraction:

1. The nuts on the rear runners are difficult to get to and require a bit of brail.  A bit of experimentation with several socket extensions and universals was required to hit upon the right combination.  From memory, socket, universal, long(ish) extension and ratchet was the right combination.  Figuring this out required quite a bit of cogitation.  And beer...

2. Get a handheld mirror and good flashlight to see what lives behind things.  I had to go down to the chemist shop to buy a mirror.  'Borrowing' the only example in the house turned out not to be such a great idea.  Phrases such as 'If you think MY mirror is going anywhere near your grubby engine, think again!'.  Other words may have been used.

3. There is a mounting bracket behind and to the left of the throttle body which was not initially obvious.  More cogitation.  And beer...

4. Remove the rubber boot from the throttle body to the air filter housing.  This is a tight fit...

5.  Loosen the hose clips that hold the runners to the airbox if you can get to them.  Probably you can't though.

6. Assuming you've removed the hoses and cables and injector lines etc, the whole CIS assembly should be able to be levered off, turned a bit and pulled out to the right side of the fan shroud with the runners intact and a bit of swearing.  And more beer...

Insertion:

1. On the bench, position the runners to where you think they'll fit.

2. Position the hose clips so you can get to them with a screw driver between the airbox spiggots when viewed from the back of the car.  Leave them just loose enough to turn the runners but tight enough so the runners don't fall out.

3. You should be able to insert the CIS from the right of the fan shroud and wiggle the runners over the studs, then tighten the hose clips when it's in position.

I made the mistake of tightening the hose clips 1st, necessitating two goes at this.

The biggest problem is that when the engine is assembled out of the car, no one thinks the CIS will be removed in situ and so the hose clips will probably face the wrong way.

Hope this helps and good luck.

Mark

[Hughesy]

great write up mark,

another example of why this forum works so well 

thanks for the education 

hughesy

[Zelrik911]

Thanks Mark

Thats some great advice!

I sort of thought it could be done but reading other forums ( eg: Pelican ) ends up discouraging you from attempting it. The doom & gloom contributors seem to override any other suggestions.

I will eventually move onto EFI as you have done, but I want to fix a few things first. Please keep posting about your Bitz kit experiences.

Thanks again for the insight.

Peter

 

[Pfitz911]

Hey Mark

Any further with this? Tuning wise, you thought it was running a bit rich...even so did you feel the engine was smoother, more powerful as the write ups suggest on the bitz website. Sounds like a wide band O2 sensor is the way to go.

Thanks for sharing your findings to us that are contemplating this

Cheers

[bumble]

No worries - just hope it helps the cause...

I find excessive forum trawling adds as much confusion as clarity to most things - the BS filter has to be set at a fine resolution.  Opinions seem to outweigh actual experience by about 10:1.

Time has been a bit limited lately but I got a 60km drive in a couple of week ago.  About 30km in car started to miss (oh-oh) but went another 10km and died in front of a gas station - merde!  Luckily Jeff (MFX on this forum) lives around the corner and came to the rescue - beers are owed.  Ran out of gas, which is odd because previously, the gauge has been lower and car still running.  Anyway, a tank of 98 fixed the problem - phew!

Am still debugging the WBO2 controller.  Sensor control is complex.  I think the latest WB sensors are made by an obscure Druidic sect using some sort of voodoo and are more complex than the earlier ones.  In any event, it'll only be used for tuning and the NB sensor will be ultimately reinstated.

Came across this http://www.rasantproducts.com/ the other day for anyone that's interested.

Mark.