Su Carbs And Fuel Mixture Displays- Ideas And Info

[altezzaclub]

For something so mechanically simple, the SU carb has a complex system of operation. It seems very straightforward when you look at a cross-section of the carb and have a read over the internet, but I've found that it does not behave in such a simple way.

 

Looking at the diagram, it is made from a body that contains the throttle plate, with a 'top hat' screwed on top of it. That encloses a cup-shaped hollow, floating piston that holds a needle in the fuel jet screwed into the bottom of the body.

 

The little ramp in the body where the jet sits forms a low pressure area behind it and air flow over the venturi ramp sucks the floating piston upwards by forming a vacuum above it. That's what the blue arrows are showing, as air is sucked out vents in the bottom of the floating piston and atmospheric air enters where the red/yellow arrows are, under the rim. The gap between the piston and the top hat is crucial to minimise air leaking upwards but also the keep the piston floating on its air cushion freely. So the theory is, the faster the engine goes, the greater the airflow over the venturi, the greater the vacuum so the higher the piston is sucked up and that allows more airflow.

 

Fuel is added by travelling up the jet and past the needle to be sucked into the airstream by the venturi vacuum. It normally sits within a couple of mm of the top of the jet, so it is sucked out easily. The volume of fuel used is determined by the tapered needle shape and the jet. The jet is just a straight-sided tube but the needle has a taper down its length. At idle the thick part of the needle blocks most of the jet mouth, and as it is lifted by the floating piston the smaller section of the needle allows more fuel up and into the air mix. The needle taper will give you a constant fuel mixture at all revs, or any mixture you want really. You can have a very lean idle that will richen at high revs.

 

The damper in the dashpot slows the rise of the floating piston just as a shock absorber in a strut works. Slow up, but fast down, you can see it rise in one of the video clips then slam down as I lift off the throttle. The rate of rise is governed by the thickness of the oil, and its is there to richen the mixture on sudden throttle openings. When you open the throttle, the airflow increases and more fuel is sucked out. Obviously the piston wants to rise, but the oil only lets it up slowly, giving a sudden enrichment that tapers off as the piston lets more air in and it matches the new fuel volume fed in by the needle. So the mixture goes back to whatever cruising ratio you have the needle set to. This is the SU equivalent of an accelerator pump, and is why it is a demand-fed carb, not a throttle-fed carb like the stock Aisan..

 

There is a long soft spring down the length of the floating piston core, and this compresses as the piston goes up. You can get several spring rates and they automatically richen the higher rev mixtures by not letting the piston up so far for a certain amount of airflow. It makes needle selection easier as the needles can be thicker in their 'wide throttle' positions instead of being tiny and delicate or in having bigger jet diameters.

[altezzaclub]

I have my setup described in the build topic here-

 

http://www.rollaclub...the-girls-ke70/

 

Basically twin inch & ¼ SUs on the 4K with a 270deg cam and extractors. As well, I sanded the needles to suit the engine, fitted a cold-air inlet that pressurises the carbs and then a pressure release flap. In the build you can see photos and videos of how the carbs work on the road. The most important mod was fitting a pair of Commodore oxy sensors and a fuel mixture display, showing me the mixture from each carb over the whole range. It woks well enough that you can see the effect of a pothole or a bump in the road as the floating piston bounces up and down, changing the mixture.

 

Sanding the needles was easy, and I just richened the off-idle area by making the needle thinner a few thou. Other odd things that have been done by people is to weight the piston with a few bits of lead shot or to change the shape of the venturi edge and piston front for smoother airflow.

 

However some things still puzzle me-

 

The theory is that the floating piston controls the mixture by keeping a constant air velocity under it and hence letting the needle taper set the mixture at any air flow. But, if the car is running along a flat road and I jam the floating piston in one spot, when I accelerate the increased air velocity over the venturi ramp will richen the mixture. This is just like having the dashpot damper hold the piston down momentarily. The question is, why does it richen when it must suck more fuel AND more air at higher flow? Why does it suck proportionately more fuel than air.

 

While the cold air inlet pressurises the airbox slightly, this is not constant. Under say, 70kph, the difference is nothing, but at 100kph I have greater pressure that atmospheric on the airbox side. So now I have a greater pressure difference over the venturi and it should suck more fuel to compensate for the extra air. But, does it rely on a pressure difference, or does it all rely on air velocity, or does it also rely on the density of the air as well as the velocity?

[altezzaclub]

If anyone has a fuel mixture device fitted, can you post what readings you get in different driving situations. I have it set so-

 

Idle is 1000rpm, 14.7 dropping to 18 or so momentarily then back up then down, never still. It follows the grumpy idle of the cam up and down in both carbs, one with a slightly wider range than the other. It is better with a richer idle than that, but I'm tight with petrol money!

 

Driving away shows “R” for rich for a few minutes until the sensors are well warmed up, then it is under 15 around town unless decelerating with my foot off. The chance of a constant throttle around hilly Orange is pretty slim, so it is constantly loading and unloading, so shows 11 to 1 to 14.7 to 1. That's using around 9L/100km.

 

Cruising at 100kph it shows 18-20, and will stay at that going up a hill with a slight throttle increase of a few mm. Pushing the foot down will bring the mix back to 14.7 and it will stay around there until I lift off and cruise again. I don't know if such a lean mix affects the motor, I was worried about burning valves, but it seems fine. Pulling the choke out pulls the jet down and richens it at any needle position, so I can richen it at 100kph to 14.7 or richer, but it makes no noticeable difference to the power it puts out. That's why it used less than 7L/100km on a trip, over 40mpg.

 

Lifting off and slowing down under a closed throttle goes straight away to 20:1, as lean as it goes, as the floating piston drops right down.. When the speed drops to about 20kph the car gets jerky as the grumpy cam idle affects the slowing down and the car surges. I've wondered if that is an airleak, maybe around the throttle shafts, but I don't know.

 

These are narrow-band sensors & display, so they read accurately around 14-15, then quickly move off to 11 at the rich end and 20 at the lean. The readings are very mobile in use, constantly moving up and down quite quickly so I don't worry about reading the figures, I just keep an eye on the bar graph beside them. Its only on a trip cruise that it settles down for long seconds at a constant mixture, and in the flat lands it will only move within 1.0 ratios over minutes at a time, say 17.5 to 18.5 range. Under power up a hill it will sit constant at 14.7 without varying.

 

I've never seen a wide-band, but would love to know what you actually see when driving if you have one!

Edited July 15, 2012 by altezzaclub

[snot35]

Interesting. When you accelerate is it richer throughout the whole acceleration?

 

Also, I was under the impression that these carbs more or less work purely on velocity. We're basically using a rough edged venturi. If you look the Venturi effect and the application of Bernoulli's equations, there is no reference to density at all, simply volume flow rates to give pressure differences. This would indicate to me that density isn't relevant.

 

Are your float bowls also seeing this positive pressure from the airbox, or are they venting to atmosphere (which is a little naughty, but we all do it :D )?

Edited July 15, 2012 by snot35

[altezzaclub]

Yes, it will hold 14.7 until I lift off, which is the way they wanted it to work I expect.

 

Well, driving along watching it I spent a couple of hours wondering how thin the air would have to be before it affected the mixture, velocity notwithstanding. Say 14,000ft in Bolivia would have fewer air molecules passng over the venturi, so maybe it would fill those spaces with petrol and enrich it.. I don't know, but it passed the time!

 

The float bowls don't see the pressure, so in theory it should push the fuel level down the jet slightly IF the higher pressure in the airbox translates to a higher pressure over the venturi. When I've caught up with everything at home I'll go and do some testing with and without the cold air tube connected. The hard thing around the Central West is to find a level road for consistent conditions!

Edited July 16, 2012 by altezzaclub

[snot35]

Wow, interesting. I have to say I didn't really expect that, but it kinda makes sense? Actually, you've confused me. Referring to the fixed piston scenario, are you saying it stays 14.7 all through acceleration?

 

It would surprise me that you'd have to re-jet even with an SU for altitude. You have to do it with Webers, even though SU's deal with things differently, we're still relying on vacuum to make a fuel signal, so the very basic principles are still the same.

 

I know what you mean about hills. I don't even have a flat spot to do oil changes and car setup, it's frustrating to say the least!

[altezzaclub]

Referring to the fixed piston scenario, are you saying it stays 14.7 all through acceleration?

 

Ah sorry- fixed piston is a thought experiment.. I know how to do it but I haven't tried it yet.

 

If I'm going up a long hill and I accelerate, it will go from 18 to 1 back to 14.7 to 1 and sit there constantly. Adding more throttle makes no difference as the needle is setting the mixture and its perfect for power at the top end when working hard.

 

I think I'll go read some of Prof Bernoulli's horrible work again (after 40years..) and assume it works on those principles.