SU Carb needle profiles

[Mike Pim-Keirle]

Hello again everybody.
Could anyone advise on which would be a suitable needle profile for a mildly up-rated 18GB engine?
I had this engine re-built last year, over-bored +0.060", which gives capacity of 1865 cc I think. New standard cam fitted, standard pistons, but both head and block lightly skimmed, so compression ration is probably up a bit from standard. Looking at the head, it appears to have been gas-flowed at some point but I don't know to what extent - I've owned the car 33 years and it has always performed well, although I have had the head off before, I never really noticed it was polished before now! K&N filters are fitted but otherwise it is all pretty standard. I use regular 95 octane unleaded petrol. I was running what I believe were number 6 needles - they would have been whatever MGB Hive supply with K&N filters. Last September I was tempted to have the car rolling road tested by JPR tuning, in Yate. They found the mixture was very lean at the top of the rev range and fitted type S6 needles to correct this. The engine was then tested and found to produce an astonishing 101 bhp and even more surprising 127 ft lbs torque - all this from what is basically a standard engine. I'm delighted with the way it drives, but very disappointed at the resultant fuel consumption, which varies from high teens about town up to 23 mpg on a 100 mile gentle run. I could understand this If I was using the extra power all the time, but I'm not. It is nice to have the power and more specifically, the torque, but most of the time I drive it like an Old Man ( partly because I am one!) and would like to get back to the days when I could get mid 30's mpg on a decent long run. There is such a bewildering range of needles available I don't know where to start, I've spoken to Burlen services, but they just say to go back to no.6, I'm a bit reluctant to do this as according to JPR tuning these needles gave such a weak mixture at high cruising speeds that they were likes to cause detonation at motorway speeds.
Any comments welcome.

[Peter Cresswell]

Hi Kim
Firstly getting the car on a rolling road was the right thing to do and I always recommend people do this when they modify the engine even if it is only minor mods. My B is also 1865cc and the torque figure you got is the same as my car was before I carried further engine mods last year.
Was the power figure the measured bhp at the rear wheels or a calculated flywheel bhp? Some rolling road operators don't tell which it is or they only quote the calculated figure as this is about 30% higher than the measured figure. I ask this because if it was at the wheels this would equate to about 145 - 150 flywheel bhp and is the reason why it uses so much fuel.
Now my engine is more modified with a Stage 3 head, Piper 285 cam with modified distributor and 10.5:1 compression ratio, I am using the No.6 needle as an interim whilst the car runs on HS4s. This was also recommended by Burlen as a good needle until a rolling road session. It pulls cleanly up to the red line in all intermediate gears and although it uses more fuel it is not far off what it was before the mods. I don't bother measuring/recording fuel consumption - I just put fuel in when it is needed - but a tank lasts about 250 fairly hard driven miles, so to get into the teens with your engine I think it must be running too rich in the 'normal' driving range. However looking at the No.6 needle compared to the L needle, the profiles are very different after the 3rd fuelling point with the L needle becoming increasingly weaker than the No.6.
You can check out and compare any SU needles by going to: http://www.teglerizer.com/suneedledb/ . The 'fatter' the needle is, the less fuel it allows and so runs weaker (leaner). Running weak or rich can produce high fuel consumption, so something doesn't add up here, which is one of your problems but I can't see why an L needle was supposed to cure the lean mixture at high rpm. Many rolling road operators now take a needle that is mostly correct and then modify it in the areas where it is a bit too lean simply by filing a bit off in those places!
You also have to remember that power costs money, and not only to get it in the first place with the cost of parts but also to run the car in fuel. When Autocar road tested DRX255C (the 3rd Le Mans car which finished 11th overall in the 1965 race) in September 1965 it recorded 14.3 mpg average over 476 miles. Incidentally, the engine also gave 101bhp at the wheels.

[Mike Pim-Keirle]

Hi Peter,
Thank you for your quick and very knowledgeable reply. Unfortunately I made a mistake in my original post (now corrected). I have just removed the needle and the number stamped on them is S6, not L. I had mis-remembered, the tuner had tried L needle first, then put in the S6 ones and then,as you said, modified these further with a file. Thanks for the link to the Teglerizer site, very useful. I have been able to check the dimensions of the S6, which seems very similar to the 6, just a bit leaner in the middle and a bit richer at the tip.
If you remember from one of my earlier posts, which you also replied to, I have also had problems with the float needle valve. However, my original needle valves are now back in the float chambers and I am happy that they are not leaking.
The power and torque figures , by the way, were calculated as at the flywheel, though measured at the wheels. I'm afraid I am one of these boring people that writes down the amount every time I buy petrol, so I'm fairly confident that my figures are right. Certainly at the moment the car is using more fuel than my 1974 MGB GT V8!
Mike

[ChargedAutoGT]

Mike
well the best way to find out what is happening is with an AFR sensor. I run one with my supercharger as it's crucial not to go too lean. It sounds like your needle is ok at idle, ok up top (good power), but far too rich at cruise type speeds. that's how i found most off the shelf needles and it took me a few weeks of driving on the same road over and over again to rub down a needle in the right places to get one that works across all RPMs and loads.

wideband sensors and gauges are a LOT cheaper than they used to be (narrowband are useless). mine was just over £100.

ironically, i probably spent more on the sensor and petrol running up and down that road that i will ever save in fuel costs with improved mpg. but at least i know it's running reasonably optimally. 118@wheels and 26mpg overall.

but why do you want a 1.8 when you have a V8? roadster?

G

[Mike Pim-Keirle]

Thanks for that. Actually I think I may have answered my own question! I am now running on number 6 needles, which is what is recommended for lightly tuned engines - eg just with K&N filters. Looking at the profile which I got from following the "Tegleriser" link in an earlier response, they are very similar to S6, about the same at the top, slightly leaner in the middle and slightly richer at the tip. I don't have an AFR sensor ( not even sure what on is to be honest) but I have been using a Gunsons Professional model gas tester, which probably doesn't show absolutely correct CO readings, but it does show relative differences after tuning. Its probably a bit early to say for sure as I've only used 1/4 of a tank since fitting them, but I have covered about 15 to 20% more miles than I would have done before, with performance unchanged or possibly even slightly better.

I've owned my roadster 33 years, so to me, it is the best MGB in the World! The V8, which I've had for 4 years, is a bit of a disappointment (though it is a very good one) and is currently up for sale. My third classic is a Lotus Elan +2S 130/5, bought in 2007 and gradually brought up to standard. It does everything the V8 does, only better and all from half the cylinders and less than half the capacity!

Thanks to all who have responded.

[ChargedAutoGT]

where is your V8 advertised? been fancying one myself.
a lotus elan+2S 130/5- you must be my other self! please don't tell me it has a full length sunroof. what colour is it.

An AFR is a Lambda sensor, O2 sensor. all same thing different names. i also have the same Gunson and tested it against my Lambda just last week.

The two agreed to within less than 0.5% which surprised me. but of course Gunson only tests idle whereas the sensor is there, with a gauge for all driving conditions.

G

[Mike Pim-Keirle]

Thanks for the information.
The V8 is on Classic Cars for Sale and also Car and Classic.
It's not cheap, but it is good!
It's Damask Red, reg no. JET291N