Weber/PMO throttle and venturi sizing equations

[amallagh]

Hello All,
I have been finding some very conflicting advice on-line regarding throttle and venturi sizing for Weber carbs on early 911 engines and would really appreciate some back up advice from anyone who really knows what they are talking about. Fundamentally I have found two apparently reputable sources of guidance for carburettor sizing which both refer specifically to early 911 engines. The problem is that they come out with very different answers for the same engine.

First let me describe the engine in question which is a 2.5 twin plug short stroke that will be going is a period 2.5ST tribute car (but the same principles/issues can be applied to any early 911 engine on carbs). The target car will be driven on track and on road so a compromise specification is required which I normally achieve by building an engine close to the original race engine spec but with milder cams and slightly smaller ports to give it better driveability on the road. So here below is the engine spec -

Capacity 2518
Bore/stroke 90/66
Valves 46/40
Inlet port size 38mm
Exit port size 36mm
Header diameter 36mm ID feeding into 36mm ID headers with twin outlet free flow sports muffler
Compression 10.5:1
Ignition twin plug dissy with programmable advance
Rev limit 8000rpm
Cams DC40 (or DC60 cams)
- duration@1 degree 266/249
- lift 0.474"/0.440"
Peak power estimated to be 7000-7300rpm


The first link of note that I used was this one from Performance Orientated which suggests some simple equations to be used to predict optimum venturi and throttle sizing based ion cylinder displacement and peak power RPM.

http://www.performanceoriented.com/t...enturi-sizing/

The key equations are copied below -

Throttle bore diameter D = 0.85 X (V X n)^0.5

Venturi diameter d = 0.8 X D

D = Throttle bore diameter (mm)
d = venture diameter (mm)
V = cylinder displacement (litres)
n = peak horsepower RPM

So for this 2.5 engine with DC40 (or DC60) cams, then -

D = 0.85 X (0.42 X 7000)^0.5 = 0.85 X 54.2 = 46mm ideal throttle size

d = 0.8 X 46 = 36.8mm ideal venturi size

Inlet valve diameter = 46mm
Inlet port diameter = 38mm

It says that typically there should be a 0.8-0.85 relationship between the 2, so this is spot on at 0.826. (0.85 for very high lift cams and 0.8 for more std cams)

It does recommend rounding down in size from these theoretical equations to get to the nearest available size.

The calcs on the previous link seemed to come out with the correct venturi sizes for some known engines right from 2.0S (32mm venturi) up to a 3.6 (42mm). The calcs predictably come out between about 0.5 and and 1.3 above the venturi of choice for each engine. So the logical conclusion was that these equations were a good guide right across the capacity and rev range for 911 engines.

So these equations would indicate that the ideal is 36mm venturi and 46mm throttles.
If 36mm venturi are correct for this engine then 46 mm throttles does also sit right in the middle of the recommended range for this size of venturi


Simple - NOT EXACTLY !

As ever with the internet here are alternative equations that can be used to calculate optimum venturi size and the answers can be quite different.

I've also referred to Bruce Andersons well known 'Porsche 911 Performance Handbook' on page 154 where it refers to venturi and throttle sizing for Webers. This uses similar but crucially different equations which I have copied below.

Optimum venturi size in mm = 20 x ((cylinder capacity/1000) x (peak power rpm/1000))^0.5

For my 2.5 this comes out as 20x( (420/1000) x (7000/1000))^0.5 = 34.3mm, implying that 34mm chokes would be the optimum size. A very different answer to 36.8mm from the other equation on the previous link.
The venturi do set the vacuum for the main jet circuit but the throttle diameter sets the vacuum for the idle circuit which can be particularly important for any low speed or part throttle running.

A slightly conflicting bit of information in Bruces handbook is back on page 141 where it is referring to modifications on a 2.7 engine where it shows a power/rpm graph of some 40IDAs with 32, 34 and 36mm chokes for comparison. The difference is negligible for some reason which Bruce comments on but likewise finds hard to explain.


So what is the best answer -
1) 46s with 36mm venturi predicted by the equations on the Performance Orientated web link ?
2) 40s with 34mm venturi predicted by Bruces book ?
3) something in between with 40s to get the idle circuit control better and 36mm venturi to give a bit more peak end ? (36/40 venture are available but the Performance Orientated link does imply that 36mm venture are outside the optimum range for 40mm throttles)

I really don't know which calculation system to trust. The graphs on page 141 of Bruces book are for a car with E cams but the 36mm venturi seems to start showing some upside above 6000rpm when the E cams have pretty much peaked out. With the DC40 or DC60s then there is still a lot further to go so don't know if this would be different with different cams.

I would appreciate any solid engineering advice on this topic and particularly in respect to the best equations/predictions to use for throttle/venture sizing. I am looking for an engine which makes good peak power but is still tractable and responsive in the mid range. The cams selected are not extreme so I trust a reasonable compromise is possible.

Many thanks
Andrew

[zedsn]

For my rs spec 2.7 I was able to run 34 and 36 chokes and didn't notice much difference and kept the 36 chokes in my PMO carbs. I have a brand new set of 38 PMO chokes available if you want to experiment.

[roblav]

The performance difference between 34 and 36mm venturis is minimal. One engine I built up 25 years ago was a 2.7RS (big valves / ports, S cam) with Weber 40 IDA and 34mm venturi. It was a fun engine.

I have a very old Weber Carb book with graphs in it with venture size based on max HP RPM and cc/s per cylinder. There's no lines! Only wide slices.

IMHO, you're better off with the smaller 34mm venturi.

[Jack Stands]

My 2.7 RS clone uses 34mm venturis on Weber 40 IDAs. Obviously, oxygen sensors and a dyno is your best friend in setting up the carbs. Should be a fun engine when it’s done!

[Trackrash]

DC40 OR DC 60? That's the real choice, IMO. Once you decide on the cams then the carb choice falls into place.

I also think you should reconsider your port size. 36mm is what the 2,7 Carrera RS had.

Unless this is a dedicated track motor, I think DC 40s and 40 / 34 Webers is the way to go.

IMO horsepower on the dyno is great to brag about, but do you really want a motor that is a dog up to 5k rpm?

There are calculations that can be done on port size / port velocity. You might want to visit those.

FWIW, I have a SS2,5 with E cams, 35mm ports, 40mm Webers with 32mm vents. It has it's peak HP at 6,700 RPM and still pulls to my 7300 limiter.

[amallagh]

Thanks for the responses so far giving details of what others have selected for similar but not identical engines.
Main difference to my engine compared to the examples so far I think is they are running milder cams, slightly larger capacity and I guess might have lower compression and maybe not twin plug as it isn't mentioned. I also get the impression that most of the engine mentioned are purely road car engines but I might be wrong.
Votes so far in favour of 40/34s.

I've had a few other comments off line -
- 40/36 may have a reduced venturi effect due to the limited difference in diameters so maybe not the best options for that reason (?)
- Had other comments supporting the view that changing the choke size 34 or 36 hasn't had a lot of effect on their engine which would imply to me that 36 isn't too big and may give slightly more peak end performance.
- One vote off-line for 46s with 36mm chokes on the basis that 36mm venturi are the better size by calculation, and then 46s are better match to this venturi size with a 0.78 ration between them.

I would be realy keen to hear from anyone using 46s in a similar engine who can describe what it is like to drive on part throttle and how it reacts in the mid range, and even what jetting they used.

Also very keen to hear from anyone with some specific advice about the 2 calculation methods described in my first post and their relative applicability given they come out with such distinctly different answers.
Many thanks
Andrew

[roblav]

Naaaa.... 40's with 34 venturis. Your engine is too small for larger carbs!

With 36's or above, your not going to get the high velocity / low pressure effect necessary to draw fuel efficiently from the mains.

Twin plug or single plug matters not.

[Gordo2]

Mr. Abbott (Performance Oriented) - highly recommended the book "Weber Carburetors Tuning Tips and Techniques" (by John Passini) as I was tuning and configuring my PMO's.

A good book if you want to delve into theory.

Passini comments frequently that each engine is different (even though they may be built to the same specs / same parts / for the same purpose) - the calculations provide a good starting point to configure the carbs.

For my engine (3.2L with DC Cams GT2-102) - the calculations lead me to buying / installing PMO 46's w/38 chokes. I worked with this config for quite a while trying to smooth out the idle/progression to main circuit transition - swapping idle jets, idle airs, mains etc. I ended up backing the chokes down to 36's which I believe helped the transition, without giving up much for performance (the dyno shows HP drops off at upper RPM - but I didn't notice in my every day driving).

Good luck.

[hienz]

Originally Posted by Gordo2

Mr. Abbott (Performance Oriented) - highly recommended the book "Weber Carburetors Tuning Tips and Techniques" (by John Passini) as I was tuning and configuring my PMO's.

A good book if you want to delve into theory.

Passini comments frequently that each engine is different (even though they may be built to the same specs / same parts / for the same purpose) - the calculations provide a good starting point to configure the carbs.

For my engine (3.2L with DC Cams GT2-102) - the calculations lead me to buying / installing PMO 46's w/38 chokes. I worked with this config for quite a while trying to smooth out the idle/progression to main circuit transition - swapping idle jets, idle airs, mains etc. I ended up backing the chokes down to 36's which I believe helped the transition, without giving up much for performance (the dyno shows HP drops off at upper RPM - but I didn't notice in my every day driving).

Good luck.

I just bought 40IDT came with 32mm chokes and 36mm chokes (two sets of chokes) and PMO intakes, and PMO fuel regulator to put in my stock 911s 1975. Which main jet, idel jet and which emultion tube is best if I installed 36mm chokes?

[1QuickS]

Those equations on y web page are from Weber sizing formulas and other correlating information published from the 60's & 70's. As such they are guidelines and the factors provide a range of sizing possibilities to select from. I chose median factors but provide the range so you can make your own sizing guestimates. I have logged quite a bit of customer data for carb sizing vs. engine particulars. In fact, the carbs shown on my home page were custom built for an Italian customer who dyno'd his engine. It was 2.5 liter with GE 60 cams, twin plug and 911R muffler. It made 251 HP @ 7034 RPM. The Webers were 44mm with 36mm chokes.

My personal car had a 2.25 liter with S-cams, twin plug & 911R muffler. It had 43mm Webers with 35mm chokes and was a street car without driving faults.

Both engines had their Webers sized using the median formulas and both were happy combinations.

My formulas (from other sources) and my graph for venturi sizing seem to correlate to my customer data for engines from 2.0 through 3.4 liters with all kinds of carburetors sitting on top and for applications from street only through track dedicated cars.

I think Bruce Anderson's sizings are practical ones based upon easy to live with cars for street usage. I enjoyed my car as having a bit more spirit and that is how I set it up. Did 35mm venturis make a sizeable difference from 32mm or 34mm? I don't know or really care. I sized them in the same way the carbs were sized for the 2.0 S engine so nothing fancy there. The 2.0S was faulted for a peaky engine, larger displacement will tame the peaky-ness of DC40 (S-cams), the DC60s will want to rev higher to get the power and higher RPM means more potential to flow air.

So, 34mm for DC40 cams & 36mm for DC60 cams is what I would recommend as a starting point. Again, my 2.25 ran flawlessly on the street and had a very nice power band without glitches or making demands from the driver. I did tend to keep the revs at 3000 RPM during tooling around but that was not a requirement, I just liked driving it that way. I could put it in 3rd at 2000 RPM & feed the throttle without it bucking; this is with 35mm venturis in 43mm bores.

Something that can be done to maintain WOT power with a smaller throttle bore is to thin the throttle shafts which provides less blockage to air flow so 40mm bores with round throttle shafts will be more like 41mm bores at WOT.

Please remember that carbs are not EFI and typically do not like smacking the throttles open, regardless of configuration. Per Passini I paraphrase: Jet your carbs as you think they should be then "Suck it & see". If the engine is happy then fine, if not then alter jetting. The Dyno is your friend to assure WOT will not melt your pistons.

Most people don't get much benefit from having peak HP at 6200 RPM in routine driving and would rather a Solex cam for all around driving with more usable power through out the power band.

[amallagh]

Again, thankyou for all the responces. A lot of the responces are contextualised in regard to road cars where mid range response dominates completely. Not surprisingly then most of the recommendations are for 40s with 34mm chokes. Although this engine will be used on the street and track, it is ultimately a race replica engine with a milder cam and peak power and characteristic is important.

Paul, I am particularly interested in your detailed response, crucially because it builds on the links I'd mentioned in my first post. Thankyou.

You mention a '2.5 liter with GE 60 cams, twin plug and 911R muffler. It made 251 HP @ 7034 RPM. The Webers were 44mm with 36mm chokes'. This is course is the most similar example so far to this engine I'm building.

You mention, '34mm for DC40 cams & 36mm for DC60 cams is what I would recommend as a starting point'
The question I really have is what carb body to put with it. You mention in your reply examples with 43 & 44mm bores/throttles. I can only assume that these were bespoke made carbs. Is this correct ?
The problem I have, given that 34mm or 36mm chokes is where I need to experiment is that I need to choose 40mm or 46mm throttles as I don't have access to bespoke throttle/bore sizing. Once I've made this decision I'm fairly stuck with it.
The problem then comes that 34mm chokes are at the upper end of the recommended sizing for 40mm bores/throttles, and 36mm chokes are at the lower end of the recommended sizing for 46mm bores/throttles.
So which carb body do I choose before the engine is tested is my problem ?
Do you have any further advice on this ?

(Another factor is that the 46s would allow more scope for increasing capacity back to 2.7 litre but this is not what is driving the decision for selecting the carbs for the engine as it is so ignore this - It just means that if 46s with 36mm chokes would work then the customer would prefer it)

Many thanks
Andrew

[1QuickS]

36mm venturis are fine in 40mm Webers but when you get to 38mm venturis you need the 46mm Webers since the 40mm bore with the blockage of the throttle shaft restricts air flow capacity. Both the 40mm & 46mm Webers share the same inlet diameters above the venturis. The custom bores were performed to take advantage of tailoring those two projects to be as balanced as possible for flow through the carburetor, a calculated design as opposed to a moderated approach.

Just to be clear: a 40mm throttle bore has the area of 1257 sq mm. Subtract the blockage of a 8mm throttle shaft that is 40mm long (320 sq mm) and you have a cross-sectional area of 937 sq mm. A 36mm venturi has a cross-sectional area of 1018 sq mm. So, a simple comparison of cross-sectional areas shows the 36mm venturi has a larger flow area than the 40mm exit bore. However, this is not an issue since the air flow through the venturi is at a higher velocity & associated lower air pressure than at the exit diameter so 36mm venturis are fine in 40mm throttle bodies. (A simple comparison avoiding the math associated with fluid dynamics is sufficient I believe.) However, 38mm venturis would be too large.

The issue with larger throttle bores is that the progression circuit becomes less responsive to air flow for smaller displacement engines. So a 2.0 with 46mm bores would be challenged to provide a good transition onto the main circuits. This is not an issue if the car is light and the operational RPM of the engine is routinely above 3000 RPM. This would be why the 906 (1300 pounds) was fine with 46mm Webers and 42mm venturis...not much need for below 3000 RPM operation when you are running for FIA points in 1966.

There are some tricks that I use to help tune out transition issues:

• Install tune-able idle air correction screws. These allow changing the "slope" of the transition fuel mixture curve to help prolong its effectiveness without installing overly large idle jets.
• Use tall auxiliary venturis. These help advance the initiation of the main circuit. This, in conjunction with the use of tune-able idle air correction jets allows the progression and then transition to have proper fuel mixture.
• Use full radius intake air horns. I was personally surprised in how a very minor "tick" at 3000 RPM in my engine went completely away when I installed performance air horns.
• If tall auxiliary venturis are used I recommend the tops of the throttle bodies be milled by approximately 0.010". This allows clamping of the ends of the aux venturis by the top covers which will prevent the aux venturis from loosening over time. (The tall venturis have their center of mass above the mounting interface (the rectangular end plates on the aux venturis) so the natural, rocking vibrations of the 911 engine tends to "beat" the end plates until the venturis get quite loose. The short, OEM venturis do not suffer this ill since their center of mass is at the same level as the mounting ends and no rocking dynamics are imparted to them.) When these become loose they allow "false air" to be drawn into the interface between the end plate of the hollow arm in the aux venturi and the interface with the body at the fuel transfer port from the emulsion tube well. This false air can/will delay main circuit initiation.
• And one last item: Adjustment of main jet, main air correction jet & emulsion tube will affect timing of main circuit initiation.

36mm venturis in 46mm bodies can be made to work but if the plan is to go to 2.7 liter (crank changed for later crank?) then that cost would be somewhat larger than the cost of enlarging the 40s to 46 or acquiring 46s at that time. If the 2.7 re-configuration does not happen then you end up living with compromised carb performance. Another thing to be aware of: the early 46s and those made after Porsche stopped using carbs on race engines have different progression circuit design. Those with serial numbers will have the early progression and those with alpha-numeric serial numbers will have a different pattern (3-hole or 4-hole) and are better suited for street driving.

If 40mm Webers are available then avoid IDTP bodies. IDA, IDS, IDT and IDAP are all good. If the car is to be tracked seriously then fuel well baffle plates are needed but do not vent the tops of the top covers using the PMO "anti-percolation" jig. This is a Ferrari mod as developed for the 365BB & 512BB. The venting spills fuel past the idle air correction jet which will flood your engine making hot re-starts very difficult. Also, during sustained, high-G cornering the flood will flood the idle circuit of the carb on the outside of the corner. I have developed an alternate venting that is centered on the carb that vents without flooding.

[amallagh]

Thankyou Paul, I found that very interesting.

The reassurance that 36mm chokes work well on 40mm throttles is good to hear, but I'm still not sure how the air speed in the 36mm choke is higher than across the throttle when the cross sectional area past the throttle valve is 8% smaller ?

I'd love to go down the bespoke throttle diameter route to get everything better matched for 36mm chokes but that option if too expensive for this projecct. The comment about going back to the 70.4mm crank is only speculative so not particularly relevant and the engine case was originally a 2.7S so the customer has the 2.7 crank anyway. The cylinders are cut down 2.7S.

So I think you are suggesting that 40s with 36mm chokes (or 34mm chokes) is maybe a better bet for this 2.5 compared to 36mm chokes in a 46.

Regarding your other suggestions about how to improve the transition -
- I think the PMOs have idle air correction screws so that is good
- I don't know how the PMO venture compare to the webers and whether they are taller ?
- Interesting what you say about full radius air horns. As you say, very surprising that that could have any marked effect. I'd just be using the big PMO ones which I'm assuming aren't what you mean by full radius.

Regards
Andrew

[roblav]

So you're running a long throw crank with smaller pistons?

Why?

In any case, all the more reason to run 40s with 34 venturis.

And full radius air horns are very effective.

[zedsn]

From my understanding tall horns are for torque and short horns are for top end. I went the 40 34 IDA route but ended up going the 40 36 PMO route with short horns and was very happy with the result for my 2.7 RS motor. The newer style hatchets did make a difference with low throttle transition and think new style are standard with PMO carbs now.

[Trackrash]

FWIW. I changed from 32 vents to 36s in my Weber 40s when I moved my carbs from my 2,5 to my current 3L. No change in the way the carbs operate or in throttle response.

[1QuickS]

Regarding the 36mm chokes vs. 40mm bore diameter with throttle shaft blockage: brain fart. Air velocity increases as it is passing waist of venturi then slows as it expands into diameter of throttle bore. Then the velocity increases as it must pass the obstruction of the shaft. This is at WOT.

40mm throttle bores will provide great low speed throttle response in larger engines but cannot support venturis larger than 36mm.

Low speed throttle response is directly affected by airflow past the progression holes while the throttle valves are almost closed, this is referred to as the "Low speed circuit". For a given engine, smaller bores will have higher air flow velocities past the edge of the partially open throttle valves as compared to slower velocities of larger throttle bores. Higher air flow velocity generates lower air pressure which sucks fuel from the progression ports more effectively than lower air velocities of larger throttle bores.

Full radius horns can account for a few % extra air flow over the tulip horns of the Webers, something like 5%. What surprised me was the improvement in low velocity performance on my engine.

Tall air horns vs. short air horns in the triple Webers is only 6mm difference in height. I doubt any performance enhancement is detectable here. I think the tall horns were adopted when the tall auxiliary venturis were introduced to keep the tops of those below the top of the air horns. To gain significant advantage in height of air horns you would need inches of difference between tall & short horn height.

The newer style "hatchets" are actually from the IDTP Webers as used on the 1969 911T & 914/6. I know this since I recommended that Richard adopt them on his PMOs to help one of his customers overcome slow speed grasp. I also revised the progression pattern for his 40mm, 46mm & 50mm carbs. The IDTP "hatchets" have a steeper "ramp" than the standard "IDA" cam-levers but total lift is the same. So, total fuel squirt amount is equivalent in both types but the IDTP cam-levers deliver it quicker at a lower engine speed.

I thought we were discussing Webers. So now we are discussing PMOs? I think all PMO carbs use a 50mm top bore vs. 46mm for the Webers. The larger opening should provide a smoother inlet air flow than the Webers would. 46mm throttle bodies would probably be fine with 36mm chokes but you would not have the easiest path for good transition as with 40mm bores. If you have Webers already then you are ahead but they would probably need service. If they are 40mm bores and you wanted 46mm then they could be enlarged. If you do not have carbs then PMOs are the lowest cost path forward. I'd contact PMO for their recommendations on carb sizing for your project.

[amallagh]

Originally Posted by roblav

So you're running a long throw crank with smaller pistons?
Why?
In any case, all the more reason to run 40s with 34 venturis.

No. As the post says, short stroke 2.5. That is effectively a 2.7 with shorter 66mm crank.

[zedsn]

Here is a photo of the early style hatchets and I tried both and later worked for me better for my PMO.

[amallagh]

Thanks for that, I hadn't realised the difference and wouldn't have noticed if you hadn't pointed it out. Pauls last post explains the merits of one over the other.
Thankyou

Originally Posted by zedsn

Here is a photo of the early style hatchets and I tried both and later worked for me better for my PMO.