Zenith TIN Emulsion Tube Testing
 

Updated 7/11/16 --> Emulsion Tube really don't behave as you expect! Time to rerun the Baseline. See post below.

Latest Update 7/10/16 --> Emulsion Tubes don't behave like people say they do: They meter from the bottom up.

The Zenith TIN currently has one Emulsion Tube configuration available based on the 4.0mm diameter E-tube. It also doesn't do a very good job of controlling AFR from 2000-4500; it's just too rich. And you can't fix it with jets, i've tried for many years.

I am modifying new E-tubes (made by alfa1750 on e-bay) and figuring out what will happen. Nice that alfa1750 is making new E-tubes, I can hack them up and still have my fall-back originals.

Here are two attempts (E1 and E2) to figure out how each row of holes affects the AFR on a WOT hill climb in 2nd gear. The red lines are from the original tubes. Probably not directly comparable, they have large chamfers on the holes, and were made 45 years ago. The Green and Blue lines are the modified new tubes. The thin lines are the actual data, the thick lines are 6-point moving averages.

For E1, I plugged one of the 1mm holes with solder. I enlarged the upper holes from .65mm to .8mm. This mathematically did not change the total hole ares, but just shifted it upwards a little.

E2 I unplugged the soldered hole and enlarged all (4) to 1.1mm. this enlarges the second row by over 60%. Barely moved the AFR.

But the 5000-6500 rpm remained unchanged between the Green and Blue. Interesting....

http://forums.pelicanparts.com/uploa...1466986993.jpg

Compared to Weber Emulsion tubes, there are 1/3-1/2 as many holes in the Zenith E-tubes. Looks like I need to add a lot more holes. But where???

Hi Mike,

I keep planning on running some experiments on my test engine to generate some data like that...not enough time to git'r done. Needless to say I'm jealous of your testing.

Fuel level drops in the emulsion tube well as fuel demand increases (higher RPM). Holes near the top affect mixture at initiation of main circuit & holes down low affect upper mid-range. Zeniths do not have stepped diameter tubes so no need to discuss that.

To peg WHERE to drill might be illuminated by simply performing a baseline set of data and then blanking off the top most holes & perform a run, repeat for middle holes & then bottom holes. This will generate a road map of what RPM regions are affected by what holes.

It is my understanding (not my knowing for sure) that the 5000+ RPM response is directly due to the interaction of main jet & main air correction jet and the mixing (emulsification) action from the holes in the E-tube; effective fuel level is below bottom hole of E-tube.

I think it would also be illuminating to run WOT vs. 3/4 throttle to see how the fuel delivery curves respond.

Paul,

All of my previous data confirms above about 5000 rpm I can change the slope by air correction jets. I recently chamged from 165 to 175 and got that nice flat 13 from 5000 well past 6500. The top end rush is awesome.

Based on the little that I've done and looking at weber emulsion tubes I'm thinking of going big. I'm already so rich plugging more holes might get me into a region where the AFR gage can't really respond accurately.

I think I'm going to do two things: 1) add small chamfers to the holes like the originals. 2) increase the upper holes to 1mm from .8mm. That should be noticeable.

That's a 56% increase on the top row and 40% overall area increase over the baseline. You'd think that might start to show a leaning out of the bottom.

One additional thing you may want to do, run a baseline with unmolested replacement tubes, ideally they should perform exactly the same as the OE's, but it may be worth it to check now and know for sure.

I'm going with 40TINs on my 78 SC engine this winter, so this is all very interesting, thanks for sharing!

J0hnny,

Too late. I already changed the new tubes. After the first run with new tubes in all location I found the changes were subtle and only changed the LHS for E2 testing. My AFR sensor only measures on the LHS.

I suspect they are not identical. On one tube, a couple of the bottom holes were not drilled all the way through. All I had was some .46 mm drills to open them up.

Overall they are what they are. Affordable replacements. But not 100%. The chamfers are the big visual difference. Dimensionally they are identical. I measured them.

Interesting Observations
1) Original E-Tube is still the best!
2) Increasing the total hole area leans out the top more than the bottom
3) Increasing the upper holes had less of an effect that expected

Why is the Original "Baseline" still best: Running an air-cooled engine WOT above 13:1 causes it to run hotter. Above 13:1 power starts to drop off faster. Around 12.5:1 (.85 lambda) is really the best since it is about at the peak power, with a little bit extra fuel to keep the temperatures down. So that is my overall goal: 12.5:1 everywhere.

http://forums.pelicanparts.com/uploa...1467640579.jpg

Here is another way to look at it: Delta ARF vs RPM. I selected 4300 rpm as the zero point because it seemed to be a common RPM where the AFR rise, pauses. This let's me see what the apparent AFR rise slope was, and maybe tell me more about the holes on the emulsion tube.

Also per my calculations, 10F air temp increase is worth about 0.2-0.3 points in AFR DROP. The hotter days give results that are richer than if they were run on a cool day. 63-66F is my "cool" day. Thus E1 and E2 on the top end would be LEANER than Baseline if they had been tested at 65F. E3 was tested at 65F, and it shows how much leaner it was on the top end.

http://forums.pelicanparts.com/uploa...1467641444.jpg

E2 actually had a shallower slope, and E3 looked a lot like the original, just smoother. Opening the upper holes seems to be causing the whole system to get leaner. I think I need to decrease the bottom row of holes to keep the overall area in the emulsion tube constant.

My next test will do just that: Plug (2) of the (4) 0.65mm holes and see what happens.

More Data: E4 tested and plugging (2) of the 3rd row holes helped lower the 4500-5000 peak AFR.
http://forums.pelicanparts.com/uploa...1467661253.jpg

It looks like total area and how it is distributed really matters.

Next Trial: Take E3 and plug (2 or 3) of the 3rd row holes. That will lower the total flow area (richen the top end) and hopefully not give away benefit of the low end of E3 compared to the baseline.

Also, I pinned the original baseline tubes and it turns out the 3rd row holes are closer to .7mm verses the new tubes, pinned at .65mm.

I like this

Update on Testing has me rethinking how Emulsion Tubes really work
 

Here is a composite image of my latest work:
http://forums.pelicanparts.com/uploa...1468163081.jpg

A lot going on here.

Paul, here is the difference between WOT and part throttle. Completely different AFR curves. Partial throttle is still dominated by the transition ports. I even captured the tip in "stumble" although it was barely noticeable.

From what I have read, the general feeling is the emulsion tube effects fuel mixing starting from the top going to the bottom with increasing RPM. As the fuel level drops in the well, the holes become uncovered and bleed in more air. If that was true, every time I open up the upper holes it should get leaner. But the odd thing is, the top end does, but the bottom gets richer. Odd.

Then, I closed off (2) of the (4) 3rd row holes, and in two separate configurations, the reduction in 3rd row hole size enriched the mixture across the entire range. Really odd.

E5 was made from E3, and E4 was made from E2, with just (2) holes plugged. Air temperature wasn't even a factor, because E3/E5 and E2/E4 were run at the same air temps.

The data says: The 3rd row (nearer the bottom of the well) starts mixing in air at the very beginning of the acceleration run.

Which means the Emulsion Tube meters air from the bottom --> up. As the fuel goes up the well into the secondary venturi, it picks up air from all holes, as it goes by them. But more importantly, the bottom holes seem to influence the overall mixing, not just at the top end. For me, this is a complete re-think if what's really going on during a WOT acceleration run.

Other things I observed: It seems that increasing the holes sizes increased the variation from low to high RPM. The exact opposite of what I was trying to do. It also seems that increasing the holes sizes leans out the top end more, which is definitely not what I want. So, total hole area of the E-tube matters too.

E6 is going to be a complete departure from E1-E5. The data says to open up the 3rd row (and maybe the bottom most row someday) and close off some of the upper holes to keep the total area about the same or less than the Baseline E-Tube.

This will be interesting......

Confused
 

The data says: it is really hard to predict what will happen next.

The original E-tube is still the best overall.

Which means I need to reinstall it and verify I can recreate the baseline data with enough precision to be able to say the other changes are discernible.

http://forums.pelicanparts.com/uploa...1468275904.jpg

I made some pretty big changes in E6, yet the results only show I made it richer down low, and in the critical 3500-4500 rpm range. Top-end remains awesome.

E3 might be the best, but i would need to go to a larger (162-165) main jet to take out the peak leanness at 5000 rpm, and that would likely push the low-end back to where it is with every other modification.

I wonder, how much is this related to being able to measure AFR now. Was this always an issue for carburetors?

Are Webers and PMOs really any better verses the Zenith TIN when measured like this?

2017 Plan
 

After pondering the Zenith TIN emulsion tube data, I've concluded the next step is to open up the top most holes, and jump from 1mm to 1.25mm diameter: E8

Recap:
http://forums.pelicanparts.com/uploa...1488650755.jpg

Basically, the Weber F3 emulsion tube has a lot more flow area, and it seems the Zenith TIN tube would benefit from a lot more flow area at the top. So I'll try E8, with Weber sized holes on the top.

BTW, I created E7, put them in, and did not do any testing. Watching the AFR gauge gave me confidence E7 was OK. I would have expected the E7 would overall lean out the mid-range. However, summer happened and I just drove the car......

Testing of E7 and E8 will be sometime in April 2017, once the car is back on the road.

Scott Hendry just performed a dyno run on his 2.5 liter with S-cams and compared 40TIN to Weber 40IDA using same main venturis. Peak HP (220 @ crank) was within 2HP but Zeniths provided 10ftlb more midrange torque. I think e-tubes were stock Zenith & F3 Weber.

Basically the main circuit is a venturi & main jet. Air cannot flow significantly better through the waist of a venturi regardless of the brand name stamped on it. If the flow characteristics above the waist vary (auxiliary venturi flow disruption, intake air horn efficiency, bell mouth curve of the main venturi, etc.) then some flow advantage can be realized but I doubt there is that much to be gained.

What Webers have over Zeniths is: parts availability for tuning, an accommodating progression circuit design and a less complicated accelerator actuation design.

Paul,

I've looked as a few AFR curves for F3 equipped Webers, they seem so much leaner at the low revs. I just need to lean out the 3000-4000 range under full throttle. So bigger holes it is!

Fun part is, I'll find out what happens.

Weird to me since holes up high on the e-tube add air to fuel mixture at low RPM. F3 E-tubes have first holes somewhat below fuel level in E-tube well which would make them richer.

Holes lower than edge of paper are actually above fuel level in E-tube well when installed.

F7 is same as F3 except lowest holes are closed (soldered shut in this example).

The following two pictures show my block of E-tubes for Webers. The strip of paper indicates fuel level in the float bowl under idling conditions. The E-tubes are upside down which justifies them all to be comparable as installed. The second picture shows which tubes are shown in the elevation view of the block of E-tubes.

http://forums.pelicanparts.com/uploa...1488673935.jpg

Elevation view of E-tubes "on their heads".

http://forums.pelicanparts.com/uploa...1488673972.jpg

Top view of block of E-tubes showing identifications:

• Bottom row, left to right: F3; F7; F23; F05; F24
• Top row, left to right: F1; F26; F2; F6

That's very interesting. Thanks!

The stock zenith and my mods are very similar to the F3. Which is odd. As I fixed the acceleration jetting to drop the initial spike, the engine would bog down under too much fuel. Afr of 10 is pretty bad.

I have never seen the fuel level in the F series. This picture is exactly what I need to think about. Great info. The waterline visible on my old emulsion tube says it's about 4-5mm above the top hole. The hole sizes in the F3 are much larger than the zeniths. As you can see from my plot.

Inspires me consider another row, higher up. Start small at 4 x 0.45mm, just to see what happens. Also bigger holes.

A little history regarding the 2.0 engines and their jetting:

• the original 2.0 used 30mm venturis, 125 mains, F26 E-tubes and 180 air corrections
• the original 2.0S used 32mm venturis, 135 mains, F26 E-tubes & 170 air corrections
• the later 2.0S used: 32mm venturis, 125 mains, F-3 E-tubes & 185 air corrections plus a high speed enrichment device when the 40IDS Weber was developed

Note that the later "S" used same venturis size of original "S" but: smaller main jets, a F3 E-tube and larger air correction plus the high speed enrichment device.

The importance is the same engine was jetted first with F26 tubes and then with F3 tubes. The F26 has holes up high and used a fatter main jet and smaller air correction. The later "S" used a smaller main jet but the F3 E-tube is much richer at initiation of the main circuit (no holes up high on the tube) and high speed mixture was strengthened by the high speed device.

I think it interesting that you noted fuel mixture is controlled by main jet & air correction above 5k RPM. I have come to believe this as well; the E-tube is 100% active once RPM is high enough and then it is main jet & air correction that determines peak RPM mixture.

Larger venturis will shift the region of E-tube effectiveness since, for a given engine, high speed air flow past the waist of the main venturi determines when the main circuit becomes active. A larger venturi will require more air flow to generate the signal to initiate main circuit than a smaller venturi hence a later initiation of main circuit.

The Zenith also benefits from high speed enrichment from the accel jets. At least, that's what the operating diagrams for zeniths say. So I'll keep that in mind.

I've jumped between 165 and 175 air correction, with the bulk of the emulsion tube testing done with 175s. 175 seems to be really close as you can tell, except as the air temp drops, 165 are more appropriate. I drive as low as 25-30F. Which is where I'm at now for fall 2016 driving.

All very cool info.

7/8/17 Update
 

After another round of testing my E7 hole combination seems the best.

I did two major things this spring:
1) I changed to 130 idle air jets vs the 125s. Much nicer cruise
2) I added a "hatchet" style bellcrank to operate the acelleration jets sooner in the throttle opening.

Here are the two summary slides. I can't explain why E7 run 2 seems to start 400 rpm sooner.

http://forums.pelicanparts.com/uploa...1499530101.jpg

http://forums.pelicanparts.com/uploa...1499530311.jpg

My "Hatchet" Style Bellcrank, made from brand new bellcranks available from alfa1750 on e-bay
http://forums.pelicanparts.com/uploa...1499530827.jpg

I was just browsing google emulsion tubes and stumbled across this thread. Mike, thanks for all of your experimenting and documentation. Any recent testing?

I can update this thread with the last bit, but E7 came out the best.

I finished 2017 with that as my preferred tube.

Winter 2017-18 saw another topend and bigger cam. I went back to the stock tube for the first runs while I dialed in the carbs to the cam. I.e., big stuff first.

2018 was a very busy year and testing got pushed off. Plus the heat wave sapped my desire to do testing. Optimizing in 95F and humidity is pretty useless for a New England climate year round.

Biggest takeaway is the hatchet style bell crank with 0.5mm acceleration jets really helps with tip in stumble.

Im also waiting for alfa1750 to have a sale on a new set of emulsion tubes. Then I can have a fresh set of E7 to start over.