CIS to ITB EFI conversion

[mb911]

So just one point to circle back on this as I specialized in flat six exhaust for the last 25 years. The reason for poor vacuum signal was/is the header size. Back when Richard Parr owned PMO he called me and asked that I stop selling 1.625 headers to folks with engines smaller than 3.0 and it wasn’t once but multiple times. The PMOs had such a poor vacuum signal with larger primaries the cars ran terrible down low and would idle terribly.


Otherwise carry on looks like you’re doing a nice job.

[scarceller]

The very best must have book for tuning theory I've seen is this book written by Greg Banish
https://www.amazon.com/Engine-Management-Advanced-Greg-Banish/dp/1932494421
Must have for anyone wanting to learn the proper way to tune.

His other book is also worth having:
https://www.amazon.com/Designing-Tuning-High-Performance-Injection-Systems/dp/1932494901

Greg was my mentor when I designed and tuned my MAF system for the 84-89 3.2L. Greg is one of the most talented free-lance tuners in the world, often hired by FORD and GM labs for his services. He was instrumental in helping FORD with the eco-boost engine design.

[scarceller]

Here's a hint for AFR target during initial tuning, set your AFR target table to 12.8 across the board. That AFR is safe at WOT and will also work at PT and Idle, it's rich but it allows you to watch for just one single number on the AFR gauge. This means you don't need to think about the target AFR, your target is the same at idle, PT and WOT very easy for you to remember.

Then once the VE table is fixed up to hit your target you can then go into the AFR table and do the final desired settings.

[scarceller]

Feel free to PM me with your email as I have experience in tuning these engines I can give you decent idea of ignition needed at idle, PT and WOT as for fuel just target 12.8 across the board for now, that AFR should run just fine.

But using MAP based tune on ITBs may not work so well, do you have the option to use TPS? or a combo of TPS and MAP?

The very best would be to build a air tight plenum on top of those ITBs then a horn off the plenum with a 3" MAF and use true MAF for metering air very accurately.

Something like this can easily have MAF sensor adapted to it.
https://www.fvd.net/us-en/FVD11096405/rsr-style-individual-throttle-body-kit-964-993-magnesium-for-mass-air-flow-sensor-street.html

Here's example of itbs with MAF, not a 911 engine but to give you the basic idea
https://myndasafn.bmwkraftur.is/d/46292-1/airbox+_Large_.jpg

Originally Posted by dannichols1474

Thank you scarceller for the tip above and also about the tuning books by Banish. You gave me the book recommendation on this thread about a year ago, so I already have "Engine Management Advanced Tuning" on my bookshelf and I will be referring to it shortly...

Today, I finally got the car out on the road for a short, 10 mile test drive. The engine definitely needs tuning - I can tell it wants more fuel and more timing. The character of this engine is definitely different from the previous engine, but I am thankful and relieved for it just to be running. Now, I change the oil, stare at the AFR and timing tables and make changes so I can be ready for the next tuning / test drive....

[scarceller]

Another hint: these engines need very little if any acceleration enrichment. Do not apply any accel enrichment initially, especially with ITBs even less is needed. The reason we need it is to replace the fuel that settles on the walls of the plenum but with ITBs you have no plenum! Greg's book covers this topic in detail.

[scarceller]

OK, so looks like MAP/KPA is a no go, I'm not surprised when using ITBs.
TPS is your best bet with ITBs but you must map out the load rows in the tables to carefully match the various driving conditions. Idle, PT and WOT voltages. No point having a row in the table that you can't get to. For example if the max WOT voltage from the TPS is 4.5vdc then no point having 4.75 and 5.0 vdc rows. Hope that makes sense.

Does the EFI have the ability to have idle dedicated tables? Like could you install a idle switch that the EFI reads? It would allow for things like decel fuel cut, where if the throttle blade goes back to idle stop and down hill coasting the EFI can then shut injectors completely off.

Originally Posted by dannichols1474

Hi Sal (scarceller),

I just sent you a PM with my email.

When I initially converted from CIS to ITB EFI last year, I was running speed density with MAP signal for load sensing and it was working pretty well especially after I got the car dyno tuned and then did additional tweaking to the VE table with auto tune for the conditions the dyno tuner didn't cover.

During the initial engine start up / camshaft break-in, I noticed the MAP signal was in the range of 85-100 kPa between 1500-2000 rpm (and 73 kPa at 1000 rpm idle). So given the apparent weak vacuum signal from the new engine, I converted my tune from speed density to alpha-n so I am now using the TPS signal for load sensing.

Regarding a plenum on top of the ITB's: I know direct air flow mass measurement with a MAF is most accurate, but I'm going after a retro hot rod look and the plenum would spoil the look for me, so no plenum.

[LukasM]

Really nice work you are doing on your car Dan, and very thorough documentation that will help other people following in your footsteps.

Also some great pointers on things you should watch out for from people with a lot of experience building and tuning these cars and their engines.

Not sure if you missed it but did you address the (too much) amount of Loctite on the cam boxes that Ian and Mike have pointed out?

I'd listen to Ben on exhaust and header selection advice as well, not many people know as much about them as he does....

Originally Posted by mb911

So just one point to circle back on this as I specialized in flat six exhaust for the last 25 years. The reason for poor vacuum signal was/is the header size. Back when Richard Parr owned PMO he called me and asked that I stop selling 1.625 headers to folks with engines smaller than 3.0 and it wasn’t once but multiple times. The PMOs had such a poor vacuum signal with larger primaries the cars ran terrible down low and would idle terribly.


Otherwise carry on looks like you’re doing a nice job.

Cheers,
Lukas

[rwest]

Dan, Julian and Sal(hope I didn’t miss anyone),

Thanks to all of you for sharing tables and tips on tuning. I have so much to learn and everything helps!

Dan,

I have seen similar results when checking my TPS on the data log and I wonder if you would(me) would be best served with a bunch of tiny increments in the zero to maybe 5% and then larger jumps up the scale? My take is that for normal cruising, you aren’t really giving it much gas.

Rutager

[rwest]

Dan,

Sorry if I missed it, but your spark advance seems very high in some cells. Can you explain the reasoning and how it works?

Good luck at the dyno. I called every dyno place I could find in the metro area of MN and couldn’t find anyone would could/would tune my car. Would have loved to have had a pro get there hands on mine. Looking forward to reading about your results.

Thanks,
Rutager

[winders]

Originally Posted by dannichols1474

I finally got it thru my thick skull that I need to adopt at least some of what scarceller has been recommending for the control of timing around the idle set point. So, I broke the 1000 rpm column into two columns: 800 rpm and 1300 rpm - both columns have the same timing set points so when the engine is idling at 950-1000 rpm the timing stays rock solid at 14 degrees.
.
.
.
I also came to the same conclusion with the new engine as I did last year with the CSI to ITB EFI conversion: stop beating my head against the damn wall! So I got in contact with Gary Bains at Bains Tuning and put down a deposit to get on his tuning schedule for the first week of August. I told Gary what I have going on and that I need a tuning job just up to the break-in rpm limit of 5000 rpm.

From experience, I know he will get the tune dialed in and then I will only need to polish the tune with the auto tune program for the driving conditions he doesn't cover in a 2 hour dyno tuning session. To me, it's money well spent.

Then after I complete the break-in period, I will make another dyno appt and let Gary wring out the engine to 7300 rpm.

Like Clint Eastwood in Dirty Harry said: "A man's got to know his limitations..." and I know tuning EFI is one of mine.

Why don't you have him break in the engine on the dyno and do the whole tune? He can break the engine in in less than an hour...

See here:

https://www.hpacademy.com/previous-webinars/229-engine-break-in-myths-dispelled/

I hate seeing people waste their time...

[rwest]

Originally Posted by dannichols1474

rwest,

Check out Banish's book, specifically Chp 7 Ignition. At no load / light load the throttle valve(s) is(are) mostly closed so the cylinders are only partially filling with fresh air / fuel charge that is being diluted with exhaust gas which decreases the air / fuel burn rate. The slower burn rate requires more timing advance to ensure peak cylinder pressure occurs just after top dead center.

For my car, the OEM timing spec of 35 +/- 3 degrees at 6000 rpm is with vacuum disconnected. The 47 degrees of advance in the Ignition table represents 37 degrees of advance to account for peak engine speed plus 10 degrees of additional advance to take into account dilution effects of light load under cruise conditions. As load increases with increasing % TPS, the dilution effects go away, the burn rate increases so the additional 10 degrees of timing advance is not required and total timing goes back to just the amount needed relative to engine speed.

Dan,

Thanks, great explanation. I just ordered the book as well.

Rutager

[pampadori]

How did you end up at 14° for your idle?

[Tom_in_NH]

The twin plugged 964 pushes timing up to nearly 45 degrees at low load. Map courtesy of Steve Wong's 911 Chips site:

[Tony V]

Damn Dan!
You know I’m going to be calling on you when I finally get my ‘74 body back from paint and place Al’s little gem in the back. Transmission rebuild started today. Took 6 weeks to get an original Porsche R&P from the dealer. Think it came over on a luxury liner, in the Captains Suite. The Dyno tuning is something I want to get done after it’s in and at a baseline. Thanks for the amazing write up. I’ve been quietly following your progress.
Tony

[KMoore68]

Dan,

That's impressive, well done!

As a fellow 2.7 owner that has been following your journey, would you mind summarizing the changes you made to your engine?

I plan on replicating a lot of your changes.

Thanks,
Kevin

[LukasM]

Hi Dan,

Impressive gain to have an extra 18% power just from (additional dyno) tuning!

As an inexperienced noob, I'd love to learn here:
- I thought your air fuel ratio was already fairly close to ideal (e.g. 12 or 12.5:1 at 100% throttle, leaner for cruising), as you had been tuning with a wide band already. What parameters did the tuner change, to get that much extra power?
- Given that the power and torque built up more linearly previously, what is causing the drop in the mid 3000 RPM range, and could it be tuned out, without affecting the high RPM range?

Cheers,
Lukas

[KMoore68]

Dan,

Thanks for the info, much appreciated!

Can you comment on the reasoning behind your decision to upgrade the cylinders? Could you have used the JE 92mm pistons with your existing cylinders?

Also, I'm guessing the extra deck height is what caused your compression ration to go from 9.5 to 9.8?

Cheers,
Kevin

[Tom_in_NH]

Hi Dan,

That certainly is a significant dip in VE between 3k and 4k. I'm wondering, that since its basically inefficient cylinder fill and slow to light-off, if increasing your ignition advance in that area might help. Similar to cruise state low VE timing.

[Tom_in_NH]

Ah, a good dyno tuner is hard to find. Tuning to MBT is the only way to properly tune. Sounds like you have the right guy!

[Jeff Alton]

Sounds like a great build! Congrats!

Can you explain the last dyno pic? Legend shows TQ as blue and HP as red, but on my screen it shows red and green lines? Maybe it is my laptop. Can you post just a final run HP/TQ with no previous data overlay? That plot looks like a big loss in the 3000-3650ish range?

Maybe also post an AFR plot to go with the power curve.

Cheers