Carb Tuning with 02 sensor - Alfa Romeo Bulletin Board & Forums
LinkBack Thread Tools Display Modes
post #1 of 22 (permalink) Old 05-13-2012, 10:02 AM Thread Starter
Registered User
euro1750's Avatar
Join Date: Jun 2006
Location: Los Angeles
Posts: 273
Carb Tuning with 02 sensor

I thought that some might find it interesting to see graphs and data that I collected today trying to dial in the jets on my carbs; the O2 sensor really reveals a lot. First, here are the stats on the engine:
2 liter engine, with C&B 12mm 292° cams, stock pistons (I think!!), and IMASAF exhaust. The carbs are the original 1750 40DCOE32’s with the following jets 130 main / 45F9 idle / F9 emulsion and 220 air corrector. Oh, and 32mm chokes. The 220 air correctors were still from previous tests, they should be 200, also was playing a bit with the idle jets (stock is 50F8). Ignition is the MSD 6AL-2 programmable version, with a blaster SS coil. NGK B7ES plugs, with 1mm gap. The advance curve is shown in one of the graphs below. Stock air box/filter.
I started with the idle circuit, basically using the fast idle method. As you can see in the 1st graph below, the fastest idle was at just less than 1 turn of the mix screw (Position1 : 7/8 open). I tried this out on the test road, but noticed that the transition phase was a bit cranky (stuttered a bit), and also had exhaust “popping” on decel. I knew the popping on decel is a symptom of a lean idle cct, so I opened the mix screw by 1/8 to 1 turn (position 2). Tried again, it was better, but still popped on decel, again opened it by 1/8 to position 3 (1 and 1/8). Now there was no popping on decel, but the idle was too low.

To get the idle up, I could have just used the idle screw, but previous experiments gave me the impression that the transition phase (starting from a traffic light for example) got worse if my idle screw was holding the throttle open too much (exposing transition holes I suppose). So, since I have a handy programmable ignition, I decided to push the idle speed to 1000 rpm by rapidly increasing the advance below 1000 (thanks to Bjornbusen’s post in the “Programmable Ignition” thread). Works like a charm. Now the idle is at 1000 rpm and the lambda is 0,82. The transition is still not perfect, I will admit, but compared to what I had, a massive improvement.
Next moved to the main circuit. I wanted to see what the main jets were doing (so watching lambda at WOT), and also curious to see what happens are part load, when the throttle is barely open, like when cruising on the highway at a steady speed. I am guessing that in this case the idle circuit is affecting the mix, even at mid-high rpm. Anyway, took some data and here is what I got (second graph).

I’ll admit that the 15% throttle estimation is pretty raw. I don’t have a throttle position sensor, so the 15% is truly a guess. I did have a chance to look at the lambda sensor while driving steady on the highway, and got the following:
3350rpm =0,84 / 4250rpm=0,88 / 4430rpm=0,86
Hard to estimate what the throttle position really was for these number, somewhere between 10% and 30% I would think.
So the steady state driving perhaps does not have such an exaggerated dip rich at 4000rpm.

Next step was to change the air corrector jets, as I had a set of 190s with me, to see the difference. I was anticipating that the tail at high rpm would become less lean with the 190s but what I got was more of a horizontal shift to the right (and a bit richer) - 3rd graph. Interesting. Honestly, I could not really “feel” a difference with the change in air corrector jets.

Any comments on which direction to go? Leaner idle jet? Leaner main? Smaller air corrector? I think there could be some interesting things to try with the idle cct, to keep it lean-ish at idle, but not too lean at faster rpm under light throttle.
Attached Images

69 1750gtv / 73 2000spider / 91 Spider (daily driver)
euro1750 is offline  
Sponsored Links
post #2 of 22 (permalink) Old 05-13-2012, 05:13 PM
Registered User
Join Date: Dec 2011
Posts: 231
From experience I would try a 120 main next. The part throttle should not be affected till 3500 rpm, but the rich dip at 3600 may get leaner. To keep up at full throttle, the air jet should need to be smaller then.
The 45F8 seems to be very small. You may have to go back to 50F8 at some point to keep transitions smooth.
Successful carb tuning is less logical thinking than try and see. No game of chess with plans for the next dozen moves.
You are in the wonderful position to have a lambda sensor and data logger, imagine trying to tune without this help. All you would make it was complete over fuel, as even extreme rich is still running.
If you don´t get a good result just with main and air jets, a change from F9 to F16 is worth a try, if you have these emulsion tubes already.
The engine seems a bit choked with these carbs and the high lift cams, which may lead to strange settings. Usually the venturis are more on the large side with these.
An option for more performance: A 34 mm venturi (largest I use before going 45DCOE, but 36 mm are available too) will give you more top end without negative effects. Trumpets, even short ones in the air box, will extend the top even more noticeable. Without them there may not be much profit from the cams in the high region.
But if you go for any of these options, the jetting will be different, again. So don´t jet to perfection first and change air flow improving parts later, do it the other way round.
You didn´t mention the acceleration pump jets, they play a vital role to enrich the top end. I suggest you use 35.
Anyway: Don´t change two things at a time, this will always confuse you.

Webber jets are very expensive. If you like to change things at your car it might be worth getting a set. If this is your only DCOE project and you want to finish it in short time, maybe the better way is to use a specialist. He should have all the parts for the carbs at hand.
There is always a perfect setting to be found, but there are some more, confusing parts in the carbs that can be changed . To try them out can be very expensive and frustrating.
The more logs you post here, the easier it is to help you, even if this help is limited over the net.

Last edited by AlfistaChris; 05-13-2012 at 06:08 PM. Reason: got something wrong
AlfistaChris is offline  
post #3 of 22 (permalink) Old 05-14-2012, 03:50 AM
Registered User
kredden's Avatar
Join Date: Jul 2003
Location: North Bend, WA
Posts: 3,334
Great timing on this thread! I just got an O2 sensor bung installed on my Super this weekend and was playing around with the Innovate LC-1 I got for use with the Megasquirt that I build. Just using the LC-1 in this case though. Had I known how proprietary Innovate was with their other connections I probably would have gone in another direction, I've been considering getting a Zeitronix as it looks to be better (from reviews) and easier to hook up an RPM signal.

I'm new to carbs so figured this would help me figure out how it is running. I have no clue what it has in there for jets and such, do those parts have identifiers stamped on them?

I only spent about 10 minutes playing with this yesterday sitting in the driveway (no driving around). I might have a sensor or ground issue as I was getting occasional sensor dropouts saying "low voltage" and it would have to warm up for a few seconds.

What software and wideband O2 kit are you using Chris?

Here's my first log on the Super.

Attached Images

Currently Alfaless. Ex: 87 Verde, 86 GTV6, 67 Super, 86 Spider, 88 LeMons Verde
Milano registry custodian(
To view links or images in signatures your post count must be 0 or greater. You currently have 0 posts.
kredden is offline  
Sponsored Links
post #4 of 22 (permalink) Old 05-14-2012, 11:26 AM
Registered User
Join Date: Dec 2011
Posts: 231
I use the Innovate LC-1 with XD1 gauge, the LMA-3 AuxBox, the new LM-2 and a very good and dead chap combined controller, data logger, rpm and 5V input, two programmable output device from this guy in Canada:


His products are fantastic and as he does them allone, he is probably a genius.
I went for his older DIYS products, when I was feed up with Innovate, who ignored problems me and a lot others had with one of their expensive products. After 3 repairs they finally got me out of the warranty time without their LMA-3 even working once as it was supposed to.
I don´t like paying a lot of $ for defective stuff.

With the 14point7 devices you don´t have to connect different units like you do with the Innovate stuff, what is the cause for most problems with these products.

You just wire the 14.7´s up cleanly and they work for ever (mine did for 2 years now).

The new one even logs on internal memory, so no laptop has to drive with you.
Don´t get shocked by the many sensors available, you can use any other sensor that you have or pull from any junked injected car.

Basically good ground and clean 12V are needed for best functioning of any measuring device. Don´t go to a point the car has, but do a new one, at last 10" away of any other ground wire.
For 12 volt a fused wire from the battery, followed by a relay (switched from ignition) is best.
Be sure ground straps from battery to chassis and battery to engine are 100%. Better fit new secondary ground wires in parallel to have the least possible resistance.

Last edited by AlfistaChris; 05-14-2012 at 11:30 AM.
AlfistaChris is offline  
post #5 of 22 (permalink) Old 05-14-2012, 12:43 PM Thread Starter
Registered User
euro1750's Avatar
Join Date: Jun 2006
Location: Los Angeles
Posts: 273
Thanks for the input Chris!
I am having a lot of fun, so collecting jets is not a problem (I have a small collection already...). As you mentioned they are not cheap when you multiply each choice by 4, so would like to make it all a little bit less random that I have done in the past (anyone looking for 150 main jets??)

I will put the 50F8's back in, do the same idle circuit tuning, and make the same measurements (hopefully this weekend). I am just curious if 50F8's are really "richer" than the 45F9's, the fuel orafice is larger (0,50 instead of 0,45) but the air orafice on the F9 is quite a bit smaller than the F8, so richer or leaner? I am not really sure..having two values per jet to play with just multiplies the potential options (once again) by 2.

I will have to buy some leaner main jets, or borrow some from a friend for a test.

I haven't checked the accelerator pump jets, but assuming they are are stock (35).

This dip rich at part load around 3600 has peaked my curiosity, I did just make a 1500km trip from the north to the south of France (and back) last weekend, the fuel consumption was 13l/100km or 18 miles to the US gallon. gulp. 4000rpm is 120kph, so it seems I could make things a bit better in this region. Of course I want all the power I can get too, for those beautiful departamental roads :-)

Regarding equipement, I have the PLX imdf sensor module. I think it was a bit more than the 14point7 module, but have no complaints whatsoever, it works perfectly. My datalogger is a bit more rustic, A4 paper, pencil and me repeating "point eight two at foutyfive hundred!!" while controlling the car under decelleration. Both the PLX lambda box, and the MSD ignition box have output signals so the next step for me is start collecting this data on a laptop :-) should be a bit safer.

69 1750gtv / 73 2000spider / 91 Spider (daily driver)
euro1750 is offline  
post #6 of 22 (permalink) Old 05-14-2012, 01:47 PM
Registered User
Join Date: Dec 2011
Posts: 231
If you drive constant 130 km/h, it should be more in the region of 10 than 13 liters. The part throttle lambda explains.

The idle jets are confusing, as long as you don´t spent a small fortune to try them in any "F" version from 45-55 maybe even 40-60. So it makes 15xF x3 fuel x 4 per carb, just 180-270 jets at about 5$ each. Not much for a professional, but too much for a single car.
But even if you are able to spent that much money, you will not find anyone who stocks them all and is willing to sell. So you will be hunting around for some values the rest of the year or even longer.
The mains are cheaper and simpler, 115-160 should do almost everything, so 40 of them are enough. Air may be 140-270, so 14x4 makes 56. Get some emulsion tubes too.
Pump jet will come out 30-45 so just 16. Think of different main venturis, usually the large ones are of interest. Sometimes helpful are auxiliary venturis, if changed from 3.5 to 4.5.
Pump return jets are not so critical, 0-70 are only 16 of them.
Get some pump springs in soft, medium and hard as you want good pick up from low rpm. The long, medium and short stroke are good for this too.
Usually the rest can stay as it is, because you are not Alfa Romeo, but your own customer and don´t want the car to start at -35°C.

If you have all that stuff, you are well equipped for any case. Surprisingly there are often more than one good setup for one carb. Today, with data logging lambda, we can tell which is best. This was impossible in the good old times...

Ever thought about programmable injection? DCOE´s are really a luxury product!
AlfistaChris is offline  
post #7 of 22 (permalink) Old 05-14-2012, 02:13 PM Thread Starter
Registered User
euro1750's Avatar
Join Date: Jun 2006
Location: Los Angeles
Posts: 273
Thanks again for the tips, I am really looking forward to getting some more data with the various setups, and consequently learning more and more about these amazing mechanical computers! You seem to have an amazing wealth of knowledge on these carbs, thanks very much for sharing!
I had a really difficult time switching even to electronic ignition to be honest, I like the technology level of the car, but finally it became clear that the ignition technology was really not at the same level as the carbs. Switching the DCOE's? ouf, I don't think that I could ever do that...even if the gains might be clear. I'll just get another car for that kind of fun :-)

69 1750gtv / 73 2000spider / 91 Spider (daily driver)
euro1750 is offline  
post #8 of 22 (permalink) Old 05-14-2012, 07:13 PM
Registered User
Join Date: Dec 2011
Posts: 231
There is a guy that did two books about Webers, one for overhaul and one for tuning. John Pasasini if I remember right.
He explains hardly anything. The funny thing about the books is, as long as you don´t do it by yourself, you have no idea how it really feels when you change a jet or other component. Sometimes you get a fairly good set up done quite easy in no time, but then any change works out wrong. The main point is, you never find a good set up by logical thinking. You can sure make a theoretical approach to have a start, but than the engine dictates the way to go. If you don´t have a basic set of jets it gets really frustrating. Race cars are easy, you just go for good power and high rpm transitions, dial in idle somehow and forget about the whole midrange. This is done in a few hours or even less.
But a well driving, economical and long lasting set up with good power at any rpm is something totally different and can take weeks.
I´m just fitting the complete data logging stuff to a GTJ, including throttle position, always exciting to find out what your first try may show in the light of reality. I perfectly understand that some dislike this cruel instruments. They never say "everything is great" but always show imperfections.
AlfistaChris is offline  
post #9 of 22 (permalink) Old 05-15-2012, 09:50 AM
Registered User
Join Date: Jan 2005
Location: Athens, Greece
Posts: 750
Those were the good old days! When there weren't any O2 sensors, we'd get a box of jets and a stopwatch and we would drive the same uphill section of highway tens of times at various rpm brackets to find the best jetting! 2500-4500 to find the main jets and 4500-6500 for the air jets. Idle jets were selected for best tractability and behavior from 1000-2200 or so in slight uphill conditions. Acceleration jets were a bit more difficult as there are many variables to play with -spring stiffness, rod length, jet size, pump return... Venturis were simpler, they were selected based on engine size and desirable rpm. Then, when the exhaust pipe analyzers became available, we'd drive around with those devices strapped to the bumper and the meter inside the car, looking for 12.5-12.7:1.
I'm very fond of my neat set of jets and still get to play with it when some of my friends show up with their...Jurassic cars! There was actually a 3-4 year period when I'd find a bunch old rundown Webers, rebuilt them completely and sold them for good money! I used to get the pair for ~15€, spent 5€ for all rebuilt parts and after ~8hrs of work I sold them in impeccable condition for 60€!! This was good money back in 1982! In those days there were plenty of carbs running around! Hell, I was 30 years old then ...that'll never come back!
Jim K.
Jim K. is offline  
post #10 of 22 (permalink) Old 05-15-2012, 01:01 PM
Registered User
Join Date: Dec 2011
Posts: 231
The most surprising thing with the O2 meters is how fast you get the pump jet setting right, even with all these variables. In fact you see clearly why there are so many options. Because of the sensor and LED´s reacting in a millisecond, you are perfectly aware what is happening in the engine.
The exhaust gas analyzer did not give any information in this direction, needing at least half a minute to stabilize. It was like a miracle when I got my first O2 gauge in 1984, about 8 years after Bosch produced the first sensors.
AlfistaChris is offline  
post #11 of 22 (permalink) Old 05-18-2012, 05:03 AM
Registered User
Join Date: Jan 2004
Location: GA
Posts: 80

Nice to see graphs -especially for the idle jets we drive on most of the time ( fuel & air holes plus the well size make many different shaped fuel slopes) . F-16 emulsion tubes are leaner down low & helped my mid range richness from F-9's .Lets see more graphs as you make progress ! Roy
roy is offline  
post #12 of 22 (permalink) Old 05-19-2012, 10:16 AM Thread Starter
Registered User
euro1750's Avatar
Join Date: Jun 2006
Location: Los Angeles
Posts: 273
Once again, thanks for the comments. I had a chance to make some runs today, basically trying all the idle jets that I have (50F8 – 45F9 – 55F8). So first to the idle circuit. The first graph is comparison of the three different jets. I was wondering before if a 45F9 was indeed “richer” than a 50F8 (at least at idle speed), since the fuel hole of the also smaller than air hole (1.00 on the F9 and 1.20 on the F8). Seems like yes, a 45F9 is still richer than a 50F8 . Not sure how much can be concluded from the graph alone, but I can say that the engines seems to like the 50F8 better, the idle is smoother, and the transition better (old school measurement!). Although the other two jets will get a faster idle “naturally” (ie. without the help of the idle speed screw), this doesn’t seem to result in better behaviour. Fast idle is at about 1 turn, and it seems that it always wants an extra half turn rich to not “pop” on decel. So I went to 1.5 turns. To get the idle up, I made a combo of ignition advance and an extra twist of the idle speed screw, now at 7/8 of a turn.
I was curious about the effect of ignition advance also on the idle speed and lambda, so…. another graph. I thought a some point the idle would start to collapse, but no, it just keeps increasing the speed (at least to 32 degrees of advance). This was done earlier with a 55F8 idle cct.
The full throttle (WOT) shows that the idle jet seem to have no influence above 3500 rpm (no surprise here). I should have taken some more data points at lower rpm, as they seem to diverge a bit, but hard to say with only a couple of data points.
The part throttle graph is quite interesting. The 55F8 is definitely richer than the others, and does seem to give any benefit. The 45F9 and the 50F8 are similar, even thought the 45F9 is richer at idle. This sort of makes sense, as both the fuel hole and the air hole are smaller, so the “impedance” at higher flow, should be more, therefore the jet being a bit less generous with the fuel as the flow gets higher.
So for the next step, I think I will try one step leaner on the mains (125), and smaller air corrector jets to keep the top end from leaning out. Also I think that some 45F8 idle jets might be interesting. I think pretty much what AlfistaChris suggested!
Or…I could just get 34mm venturis and start this whole exercise all over…  I think several have mentioned this based on the cams that I have, and it is also in Jim’s K book (great book by the way! A reference on my bookshelf – thanks!)
Looking forward to your comments!
Attached Images

69 1750gtv / 73 2000spider / 91 Spider (daily driver)
euro1750 is offline  
post #13 of 22 (permalink) Old 05-19-2012, 09:04 PM
Registered User
Join Date: Mar 2007
Location: San Antonio, Texas
Posts: 4,121
Originally Posted by AlfistaChris View Post
There is a guy that did two books about Webers, one for overhaul and one for tuning. John Pasasini if I remember right. He explains hardly anything.
I have to disagree with this, Chris. There is information in there but Passini reads like he recorded ramdom thoughts and then just typed up what he had. The real problem is understanding what the guy is trying to say. Buried among the random words and commentary are some good ideas but I found that I had to read the d**n books several times before he made any sense at all. I've always thought that understanding Passini is a good litmus for true Weber conoscenti.

The funny thing about the books is, as long as you don´t do it by yourself, you have no idea how it really feels when you change a jet or other component.
Oh, yeah? Well, what about "suck it and see" . . ?

Jim . . . '72 Super 1300, '70, 1750GTV, 2nd series,
'62, Lancia Flaminia Zagato3c, 2nd series

Last edited by 180OUT; 05-19-2012 at 09:11 PM.
180OUT is offline  
post #14 of 22 (permalink) Old 05-20-2012, 03:00 PM
Registered User
Join Date: Feb 2007
Location: Austin, Texas
Posts: 187

Since you're soliciting comments, here's a few for your consideration:

1. What kind of redline are you going to be using and how do you drive? Will you be autocrossing or doing HPDE events? Are you willing to give up low speed tractability for higher end hp? I haven't seen any curves above 6,000 RPM.

2. If you're not going to be routinely using a 6.6 - 7k redline, I don't know if I'd switch venturis. The factory chose 32s for great low end flexibility, although I have successfully run 36mm venturis in 40 DCOEs in an autocross and street car with a setup not much different from yours (300 degree Alfa factory competition cams). I was a demented youth at the time, but the car was very tractable (and fast ).

3. FWIW, you might consider shortening up your exhaust cam timing a bit. As Jim K pointed out in his book, the exhaust flows better than the intake. You can get both better top and bottom end performance by shortening up your exhaust timing relative to the intake. Maybe trade a cam and lifters with a bud?

4. Since you're undertaking a methodical approach to dialing the carbs in, given your comment about changing both main and AC, my comment that you should make only one change at a time is, no doubt, superfluous. In the same vein, the purchase of a set of jet drills might save a bunch of money in idle jets until you get close, allowing you to blow your funds on emulsion tubes, pump jets, pump exhausts, etc. Further, you could drill either end of the jet and make one change. Go too big? Solder 'em up and re-drill. Replace with factory jets when you're happy.

5. Will probably find peak hp in the vicinity of .86-.90 lambda.

6. What is your regular cruising rpm? If we knew that we might be able to point you in a direction that yields a nice, relatively lean cruise without compromising throttle response or power.

7. Those 40 DCOE 32s have a really nice progression curve. Try to keep the throttles close to closed at idle.

Thanks for posting your data and good luck. Hope some of this is of use.


Gone but not forgotten: 70 GTV, 76 Alfetta, 82 GTV6 Balocco, Step front GTV race car. Current: Giulia Spider Veloce race car, 105.51 GTV race car to be.
ciscotex is offline  
post #15 of 22 (permalink) Old 05-20-2012, 04:02 PM
Registered User
Join Date: Dec 2011
Posts: 231
Yes, you are right, it´s not easy to read but as it is the only use able book on the planet, it is the best you can get. I sure read it 25 times, some passages even much more often.
But you can´t learn to swim from a book, you need some water. So you have to try out in your personel reality what we writes and see the results.

Thats what I wanted to say. So any time you tune a real carb, you can understand a bit more of the book, but maybe not. It sure is no " Five easy steps to the perfect tune" book.
But is has a lot of important pieces of information in it, facts you will not find out that easy, left alone.
As long as people don´t have all the jet´s available, most hope tuning is mostly a hardware problem. If you have them all, you realize it´s not.

Today the way to a well tuned engine is different, but not easier. Because of hard facts that show how far you have come, you are not satisfied if the engine runs well, it has to have the real fueling too. This is a good thing, but also makes it harder to find the point to finish.
As I do more programmable injections today, I´m used to get everything spot on. This is very hard with any Weber.

Euro 1750,
I have a some problems to read the right things from your logs right now, so I will not comment on them.
AlfistaChris is offline  

Quick Reply

Register Now

In order to be able to post messages on the Alfa Romeo Bulletin Board & Forums forums, you must first register.
Please enter your desired user name, your email address and other required details in the form below.

User Name:
Please enter a password for your user account. Note that passwords are case-sensitive.


Confirm Password:
Email Address
Please enter a valid email address for yourself.

Email Address:


Thread Tools
Show Printable Version Show Printable Version
Email this Page Email this Page
Display Modes
Linear Mode Linear Mode

Posting Rules  
You may post new threads
You may post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off

For the best viewing experience please update your browser to Google Chrome