Alfa Romeo Forums banner

1 - 12 of 12 Posts

·
Registered
Joined
·
31 Posts
Discussion Starter #1
I have a good alfa nord 2 litre engine that has just been rebuilt. It has twin 45 weber carbs and puts out 140bhp at rear wheels. If I put ITB on instead of carbs what would the advantage be, if any?
 

·
1966-2013
Joined
·
13,741 Posts
You mean with programmable EFI control?

You'd get a higher degree of tunability and could make up for some of the weak spots in the range a carb might produce.

Though a single TB might be more desirable than ITB as there's some pretty radical vacuum loss with ITB's, or at least they operate in a much narrower range than what carbs or an ITB would (like say -12 at idle ~ 0 under load instead of the more common -21 at idle ~ 0 under load to throw some numbers)

Flip side is with ITBs you can tinker intake track length pretty easily through the use of stubs and stacks.

Whether it would actually make more power on the same engine :shrug:
Depends on how off the carbs might be now and how much dyno time you wanna spend tweaking at maps. They could certainly match the performance you're getting, but to go beyond it by an amount that would make the time, cash and effort worthwhile, dunno.

They'll sound much like the carbs though do if that's important to ya :D
 

·
Registered
Joined
·
256 Posts
Keep in mind that the stock SPICA throttle bodies have a 38mm bore, so this might be smaller than the choke size in your 45 DCOEs.

The biggest advantages of an EFI system might be with better driveability, better cold starting and driveability when cold, and maybe a broader torque curve if you integrate it with mapped electronic ignition control. But there's no inherent reason EFI could give you more power... certainly no more peak power.

It'll put a huge dent in your wallet, though. Add up your cost estimates for all the parts and machining and wiring and dyno time... then multiply by 3, and quadruple your time estimates. You'll then have a realistic estimate of the time and money it'll take to convert to EFI.

I've spent the better part of 6 months converting my SPICA car to EFI, and the past 1 1/2 years constantly fiddling with it... and it still doesn't run right. It runs better than a carb'd car when cold and under certain conditions, but is still not consistant, and has all kinds of annoying and wierd electronic and hardware problems.

In my experience, it's an interesting project if you're into fiddling with stuff and fabricating parts. If you'd rather just drive the car... stick with what you've got!

George

'74 GTV 2000, SPICA throttle bodies, Megasquirt II EFI
 

·
Registered
Joined
·
400 Posts
FWIW....I did a 2-stage conversion on a carb engine some while ago (not Alfa..but pertinent). It was a 1690 Ford crossflow, with the classic tuning mods of 11:1 CR, 40s, big valves and ported head, along with a fairly lumpy cam. It made decent power (close to 140bhp), was a bit temperamental at light throttle/low speed and had a bit of a hole in the torque lower down too.

First step, I converted from the optimised mechanical distributer to proper 3D ignition - the difference was pretty impressive all round.

Next step was ITBs, only 38mm, but bigger than the chokes in the 40s, another huge improvement. Sadly, I never did any rolling roads tests so it's all very subjective, but it was absolutely worthwhile - transformed the car. Much more tractable, better mpg, better power, totally smooth delivery. I never did take this engine to the rollers as it was always a half-way-house to the 2L Zetec, so I used a map from an engine with the same basic spec that had been on the rollers - it was close enough for my needs, although no doubt could have been improved upon.

The only real problems I've had with my aftermarket EFI conversions so far have been getting cold start/warmup right once the basic mapping has been done - this is a tedious and long-winded job and not really something the RR can do as it's a one-shot-per-day kind of thing. I've spent many journeys to work with the laptop sat on the passenger seat trying to get this bit right. Based on his comments, this is the sort of thing that George is probably still working on I'd guess. It can be tedious! I reduced the pain a bit by using a commercial aftermarket ECU rather than a total DIY job.

At least I do all my own work, so I 'only' have to buy the ECU and odd parts. Still not exactly cheap.

HTH
 

·
Registered
Joined
·
256 Posts
The only real problems I've had with my aftermarket EFI conversions so far have been getting cold start/warmup right once the basic mapping has been done (snip)... Based on his comments, this is the sort of thing that George is probably still working on I'd guess.
Ha! I nailed that part after about 6 months of daily driving the thing to work. From my experience, the Nord engine with ITBs seems to need a HUGE squirt of fuel to get started when cold. It exceeds the maximum pulswidth that the Megasquirt software can allow, so I cheated the software in other areas to get enough fuel in there when cold. Hot starts are much easier. But still, it's a compromise much like you'd make with carb jetting.

I'd guess that your Ford X-flow on ITBs made similar power to your Weber setup, but probably had more "area under the curve", ie broader torque curve and better low and mid-range driveability. On the road, this would feel like more power, which is not a bad thing. I'd guess that half of this is down to the mapped ignition.

I worked with a race engine builder years ago converting Formula Atlantic engines from Webers to EFI and ITBs. For the first year or so, the engines made a bit less power than the carb'd versions. After a year or so of development and countless thousands of hours on the dyno, the engines started making a bit more power... up from 240-ish to 245-ish hp.

My biggest problems have been with making the thing run consistantly under all conditions, and getting the subtle throttle roll-on response right under light and medium loads at all rpms.

Some of this (mid-range roll-on response), I'm guessing, is down to ignition advace (I've still got my stock MarelliPlex dizzy), but a big part of the mystery lies in the built-in temperature compensation tables in the EFI software. I've got 8 different tables to comprehend and fill out to deal with air and coolant temp compensations for different conditions. Its maddening how differently the car can run at different times of the day. Nothing a few thousand more hours with the laptop can't solve!

George
 

·
Registered
Joined
·
400 Posts
Hi George,

Absolutely - a big slug of fuel on startup usually helps with most engines it seems. I feel your pain on the many tables too - I'm not using Megasquirt, but there are tables for cranking enrichment with engine-turns & temp factored in, basic temp maps of course, accel fuelling, idle air control, extra air at start....endless bloomin' list of things to get wrong.

The Xflow definately made a bit more top end due to the bigger bore of the TBs compared to the chokes in the 40's (long time ago, but I think they might have been 34s) but for sure the biggest gain by far was the area under the curve and the good manners.

WOT is easy - it's the transitional stuff that's hard as you say, and the warmup phase along with different ambients - no wonder the OEMs spend years doing maps. I suspect you'd get some help by moving to mapped 3D ignition though - makes a big improvement and you can play tricks with it to help fill holes.....
 

·
Registered
Joined
·
272 Posts
Never noticed any difference between carbs and EFI flat out.
Hillclimb times were virtually identical.

But did notice a huge difference in low end torque and overall engine smoothness.
Hills that I use to have change back to 4th I could accelerate up in 5th.

Also, starting was a lot easier hot or cold.

But as others have said, it took a long time and many hours of tuning to get there, especially with ITBs, they are very tricky to setup due to the weak and erratic vacuum signal.

I had a SB chev tuned in a weekend, the alfa took years of constant tweaking.

A wideband 02 sensor and good knowledge of the nord's preferred AFRs are a must for the home tuner.
 

·
Registered
Joined
·
400 Posts
I use Alpha/n on mine so don't have any problems dealing with the wild MAP signal, but a friend did his engine with a Megasquirt and found a blended setup worked well. I don't know any details I'm afraid, other than he's now happy!
 

·
1966-2013
Joined
·
13,741 Posts
OT tangent warning:

FWIW, a good %-age of the at idle pulsations that occur in with ITB's can be eliminated by 1st building a small 'vacuum chamber' to tap all the cylinders into, then on the service hose to the MAP sensor that comes from that chamber to the sensor, orafice the throughhole down to something in the wee tiny range, (.062" is around the generally suggested size range) which in turn helps slow the pulsations even more.

Many aftermarket injection places reccommend/describe/suggest the above method to help things along for we of the ITB, while others do not and you're left more or less hanging with radically pulsing MAP readings to mess things up if you're not using TPS for mixture mapping.

Suffice to say you never want to run a MAP sensor off just one cylinder regardless.

Oh, and just because you have one brand/make/model programmable EFI system doesn't mean you can't read other complanies install manuals to glean an idea or three.
 

·
Registered
Joined
·
2,984 Posts
This is what I did. I cut notches in line with the axle of the butterfly.
This is where the air speed is the slowest.
The other problem with ITB is the MAP ports can ether have air blowing into them then snap and act like venturi sucking on the port.
 

Attachments

·
Registered
Joined
·
272 Posts
The problem with commoning up the vacuum lines from each individual runner is that you effectively make up a mini manifold whereby air flows back and forth from one port to another.

It tried all sorts of combinations and the best was to not have any capacitance chamber as this seemed to cause too much lag.

I used .6mm mig tips in the line from each runner and then a 5th one in the line to the MAP sensor.

On the MS website is an interesting discussion called synchromap about using a MAP sensor for each port with some sort of low select circuitry.
Those that tried it reported better MAP resolution and less of log type VE table.

I tried blended SD/AlphaN but couldn't get it to work successfully. In truth, after all the effort of getting just plain SD to work I just wasn't up to starting all over again.

I haven't heard of anyone trying a MAF sensor with ITBs but I reckon it would work and would be easier to tune.
 

·
Registered
Joined
·
2,984 Posts
I think some of the OEMs with ITB use a MAF. I think BMW is one.
I was running MAP only with that setup until the RC servo died in the spica.
I need to redo it with a stepper just been busy with other things.
but I did still have some strange ups and downs. I used a small table. the larger table would have helped. I will try MAP down low and TPS up high the next time I get it together.
I had no pulse problems. but I think I was still seeing the venturi affect. that needed steep ups and downs at certain RPMs in the VE
 
1 - 12 of 12 Posts
Top