Emulsion tube testing

[1QuickS]

My quest for emulsion tube interaction with the various circuits is getting closer to being satisfied.

I've finished modification of my test Weber carb ( a twisted & partially broken IDS) which allows viewing of the emulsion tube while running. I also have vacuum taps for the following locations:

• Waist of main venturi
• Waist of auxiliary venturi
• Progression port chamber

The Weber has two throttle lever arms so the long shaft with two throttle valves is independent from the third throttle valve. The third valve has a pointer and degree wheel attached so throttle plate opening angle is known. This helps deterimne location of the throttle plate as it uncovers progression hole in wall of throttle body.

I can run the carb on my test engine and use two cylinders against the third to allow variable vacuum at the test cylinder; no need for engine dyno that way.

I have yet to install AFR meter into the exhaust pipe for the test throat.

I have venturis from 21mm through 34mm but expect to test emulsion tube response to vacuum at main venturi so the 21mm venturi will be used for that; more vacuum with less test engine stress.

I have a few more tasks to complete before my test program begins but I plan to research:

• main air correction jet size upon main circuit timing
• emulsion tube response to vacuum
• main circuit timing differences between short & long auxiliary venturis
• idle/progression circuit overlap with main circuit
• idle/progression circuit timing resulting from idle air correction jet size
• air flow comparisons between main venturis which will help tie all data together

Photos of carb as configured:



General view of sectioned carb showing viewing port, fuel feed pipe from float bowl, emulsion tube, vacuum taps and special shaft coupling.




View through port at transparent E-tube.




Degree wheel & pointer for throttle shaft rotation.




Assembled carb ready for testing.

[356RS]

Paul, that blows my away. Great idea that will provide some very useful data.

[1QuickS]

All I need is R&D funding, any help?

[356RS]

I hear that Paul. Been there and still there. Closer to Salem yet?

[1QuickS]

Plan is to close shop & sell in Jan & move then. See you there...eventually.

I wish I had started the R&D project 50 years ago when the info was more relevant.

[Fleming]

Any help?


You bet, PM sent.

All your efforts to keep our old cars running in just so appreciated.

Love your testing rig.

[1QuickS]

Thank you Bob for the support! I was really just kidding about funding, the R&D is to satisfy those nagging questions I have about the nuances of carb operation that I "think" I understand but don't actually have my own, real knowledge generated by a R&D testing program.

Little commercial value, this effort is; mostly intellectual. The majority of my customers are very happy with "close enough" and standard setups work well.

I have experienced emulsion tube changing on my car with associated jetting changes to make them work & the answer is: "There are multiple ways to skin a cat."

I figure that if a data base is generated consisting of interactions of components, all of which are based upon air flow through the main venturi then it is possible to provide tuning of carbs in a manner similar to re-programming your EPROM for your EFI system.

That is the plan, at least.

[j-dub]

Hey Paul,
This is Awesome. Reading the theory about how the emulsified fuel interacts with the Emulsion tube is one thing but actually seeing it is another. Really curious as to what you find.

Here is a you tube video of I believe Doug Berg doing a similar test with a 48 IDA.

https://youtu.be/Z65bYqfFo-A

[1QuickS]

j-dub,

Thanks for the video link, this is essentially where I am going but will use AFR readings to quantify fuel response vs. air flow/vacuum. The grand plan, that is.

I plan to make real progress by end of year but need to service some Webers/Solexes as my first priority.

[micheloaks]

Mr Abbott

your efforts do not go unnoticed.

BTW shame you ceased the ZENITH program.

Cheers Michel

[ed mayo]

Paul, you're about to uncover the "black magic" part of carburetor tuning,,,,,the mysterious 'emulsion tube' Eduardo would be so proud,,,,,hell, he probably had a similar rig!

[l_turn9]

Hey Paul - Great approach! Wish I could help with funding.
Larry ;-)
The Oil Analysis Co.

[l_turn9]

Paul - I hear all the time from people wondering what e tubes to use. All I can offer is the chart on my website and little else. I have some reference material but you are doing empirical testing which I haven't seen before. It's still relevant but probably to a smaller audience.
Larry

[1QuickS]

Thanks for the positive comments. I have a bit to go with instrumentation of my test setup but all items on the bench. Just need to steal a little more free time.

[al lkosmal]

Paul,
Thanks for taking the time to dive into this. While I tend to focus on EFI these days, I still love the Webers and, like most everybody else, tend to tune them by trial and error. The tech info that you provide on your website has improved my understanding of how these things work..............but the emulsion tubes are still a mind blower..............

regards,
al

[DWit]

After observing some very wonky mixture results on the dyno, I used the Weber manual and did my best to categorize the common IDA emulsion tubes. My setup was this: Weber 46 IDA's with 38mm venturies and F24 tubes. Based on the results, I drew lines representing the RPM ranges where drastic mixture changes seemed to correlate with the holes in the tube. The test of the F24's showed lean below 4000, rich between 4000 and 6000, and perfect above 6000. I will be switching to F7, and will report back on the results. Keep in mind that tube diameter is also a factor, but only in relationship to main and air jet sizes. So this specifically addresses hole placement vis-a-vis RPM range.

This is by no means a scientific study, so let me know if you have any supporting or contrary information.

Dave

[1QuickS]

The selection of e-tubes is basically as you have deduced with the lines and having all e-tubes "justified" with their tops held at the same level. I have a block of e-tubes that allows me to visualize their relationships with each other. Another line to draw is the line representing the fuel level in the float bowl which is the same in the e-tube well. Holes at this level affect timing of main circuit initiation.

I have acquired instrumentation for my test engine and have the exhaust bends so I can make individual runners so I can install an oxygen sensor in the single barrel that is independent from the other two. I have made the flange for this header so there is not too much left to do before I get the test engine up & running.

I also just received an AFR data logger so I can make comparative runs in my car. My plan is to establish a baseline and then swap e-tubes & make 1/2, 3/4 & full throttle runs on a local road with a nice, continuous incline. I hope to achieve a very good set of comparative data charts that will illuminate e-tube selection.

By performing similar runs on my test engine & up the inclined road and by blanking off main circuit & then idle/progression circuit on one bank of carbs I hope to shed more light on the interaction of idle/progression & main circuit.

Problem is time. I'm backlogged, behind schedule & needing to prep our house to sell & then relocate to Salem, OR. Also, I intend to sell my car but not before data is collected.

Oh, yes...holidays are here with family commitments. It feels like looking at a Thanksgiving table of yummy food & thinking I can swallow all of it.

I'll make progress but gotta keep all the balls in the air as best I can.

[911pcars]

Paul, all I can say is "excellent commitment" to reveal the intricacies of of Eduardo's carb. Much as I suspected over the years, there are more variables at work in maximizing these carbs than mere engine size. The widespread use of WB sensors allows owners to better observe the air/fuel mix these carbs deliver under various road and driving conditions.

[Gordo2]

Originally Posted by 1QuickS

... I have acquired instrumentation for my test engine and have the exhaust bends so I can make individual runners so I can install an oxygen sensor in the single barrel that is independent from the other two.

Great experiment Paul - adding some science behind the black magic of e-tubes.

I was going to comment about using a wideband, but you were obviously already ahead of that one. The wideband is going to provide a lot of data and insight as you walk though the various configurations.

BTW - your tip on removing the main jets, and doing a test drive monitoring/recording AFR really helped me isolate how my idle circuit was performing. I went from thinking my idle circuit dropped out at ~ 3,400RPM to seeing (via AFR's with no question if I was getting contribution from the mains) that it continued to contribute all the way to ~ 4,500 RPM.

Looking forward following your progress.

Gordo

[1QuickS]

Vacuum is the "Rosetta Stone" for understanding fuel delivery for our Webers. Since vacuum resulting from high speed air flow (through the space between the partially opened throttle valves and the throttle bore where the progression holes are located and from the high speed air flow through the waist of the main venturi) draws fuel into the engine it stands that correlating vacuum signal at the progression port and at the waist of the main venturi to fuel flow in these regions will provide critical insight to carburetor tuning.

The data logger I acquired will provide vacuum as a function of engine RPM. At WOT, vacuum readings are directly proportional to engine intake CFM (basically a function of displacement & RPM) and venturi size. A mapping of RPM vs. vacuum for various venturi sizes provides correlation between vacuum and AFR (fuel delivery rate) for the idle/progression and main circuits. With this info and knowing cylinder displacement, engine RPM & main venturi size the fuel delivery is determined. The combining of this data to make it accessible for use will be interesting.

I also want to quantify the effect of the waist of the e-tubes which comes into play when opening the throttles from a cruising condition and another item of interest is to fabricate some e-tubes with holes along their length in evenly spaced intervals so that isolation of effect to mixture vs. vacuum may be determined as to where; along the length of the e-tube, the holes are located.

Per Gordo's comments regarding isolation of idle/progression vs. main circuit: this project should generate the road map for tweeking idle/progression parameters relative to the main circuit to smooth transition ills.

I have a EE friend who is eager to help setup the system and help collect & analyze the data so tangible progress with this project may be a bit closer than this Spring.