Mini 3 Port Injection

spag · Post by **spag** » Mon Apr 25, 2005 7:06 am

Shirly 2 TBs is going to fuel differently to 2 carbs because with a TB the fuel appears when the ECU says 'squirt', but with a carb the fuel appears when the inlet port opens and the fuel is 'sucked' out.

By my feeble reckoning you could get away with pretty uneven spacing between 'suck' events and still get consistent fueling for each (subject to the dashpot damping on an SU)

arnold · Post by **arnold** » Mon Apr 25, 2005 10:41 am

I have an A-engined (1275) Austin Healey Sprite. I ran till recently a Eaton M45 blower, blowing into HIF (SU) 44 turbo carb (from a MG Metro Turbo). I swapped the HIF44 for a TBI unit (Rochester) and I got it idling and revving. But as the TBI unit is mounted sideways (turned 90 degrees), the fuel leaked out of the unit while idling (because the fuel is sprayed against the closed throttel).

I guesed / hoped tha the blower would help to keep the fuel in, but it didn't. Because of the slight boost the fuel was also forced out of the TBI. I thought that the TBI unit would be ok to have the boost, but is wasn't. It leaks fuel where the throttle axle comes out of the housing. So, I made up my mind and I removed the trottle from teh TBI unit and will close the holes... I will put the throttle valve upstream, in between the airfilter and the supercharger

But.... could this lead to other trouble? Am I doing port injection then ??

Is this gonna works??

See details here: http://kmnet.demon.co.uk/~arnold/

Arnold

twentyover · Post by **twentyover** » Mon Apr 25, 2005 11:17 am

Kudos to mikesb and cng1 for trying the siamesed port deal. I'd like to see pictures of the manifolds and injector locations, if possible. Thinking about the anemic 78 B in the barn and possibility of turbo injected hp. help

racingmini_mtl · Post by **racingmini_mtl** » Mon Apr 25, 2005 12:01 pm

I agree that it's interesting that cng1 and mikesb are trying this. However, I may be wrong but I don't think that Mike is going to have as much of a success at it as cng1. From my understanding, with cng1's setup, the presence of the plenum and the use of 4 squirts means that all cylinders will get fuel from both injectors. It will be interesting to see how good of a fuel distribution he can get. I'm also curious to know if you can get constant fuel distribution with regards to rpm and load. I would assume that you would get worse distribution with lower load/rpm which means that fuel economy will not be very good.

With Mike's setup, there is no plenum so each cylinder pair is on its own and you get the full effect of the charge stealing. I don't know how many squirts are used but with 2 or 4 squirts you will have distribution problems. Do you have 2 widebands like cng1? Again from my understanding, with 4 squirts, you will get a constant unbalance between the 2 cylinders and with 2 squirts, you will get an unbalance that will vary every time you start the engine depending on which cylinder fires first (or first trigger point). I'm more that willing to be proven wrong if your results go against theory so let us know what you find out. However, I think you would be much better off using only one of those RC51 TB and swapping larger injectors. You would get much better fuel distribution. By the way, what kind of injectors are these? Are they the same as automotive types so that you can easily swap them?

In any case, I think that Arnold's setup is much better for a supercharged/turbocharged setup since you get a much more predictable fuel distribution and you don't want lean mixture under boost. By the way, I think that you'll be fine with your setup as long as your fuel pressure regulator sees the vacuum/boost seen by the injector.

Regards,
Jean

twentyover · Post by **twentyover** » Mon Apr 25, 2005 2:28 pm

If mikesb had a large enough wetted volume for each port, say maybe 1.5-1.75 times single cylinder displacement, and multiple squirst/rev?

Just calculated- a 1.5" id pipe would need to be 15 1/5" long to equal 1 cylinder volume. Maybe not so practical.....

1 1.75 id pipe would be 11.4" long- maybe starting to approach practical

mikesb · Post by **mikesb** » Mon Apr 25, 2005 6:37 pm

This should be pretty easy for me to check - it'll be obvious from the plug colours. I'll have a look as soon as I get some tuning done (probably on the weekend). My wife has a good camera, I'll try to get some pictures.

I've got a wideband on the center branch (which will be a narrow band for every day running some day), so it's not easy to check the outer cylinder mixtures with it - that's a worst case option.

I understand the point about the SUs only drawing when there's flow. I wonder whether the air flow at the TBs truly switches on and off violently. It could be stagnant when both valves are closed - no fuel pulled from the carbs, over richened by the injectors & most fuel flowing into the first valve to open. Or it could act more like it's a dampened spring, slowing down and speeding up, and at TB mounted jets the pulses are pretty smoothed out. Or somewhere in between those two extremes...

I've got a bunch of fall back positions - a hefty balance tube like in the original SU manifold, switch to use one of the pair (it'd be a sweet package if I can fit larger jets, all the bits will swap over), TBI from a buick (yuck).

exupturbo · Post by **exupturbo** » Thu Apr 28, 2005 2:23 pm

Any help ?

http://www.emeraldm3d.com/em_projects_Aseries.html

Not Megasquirt I know, but maybe some usefull info.

Mark

mikesb · Post by **mikesb** » Fri Apr 29, 2005 6:36 am

Interesting site. They use a single carb, I don't think there are any concerns with using TBI on a siamese port as the issues are eliminated with a single wet manifold, the air flow should be pretty constant - no lopsided speed changes. It's probably still best to use as many injection pulses as possible to emulate a carb. Our issues are with port injection, or with a hybrid like mine (motorcycle TB injection, further upstream than port injection).

In the text they say they improved from 75HP (flywheel) to 90HP, and it sounds like thanks to the FI. But in their performance graph you can see they get 90HP from a modified engine (ported head, fast road Piper cam, 1330cc), not really apples to apples.

The advantage of SUs and one reason why I'm trying twin M/C injectors is there's a straight shot at the intake ports, I think I'll get better flow at higher RPMs. Of course the real reason is the M/C TB's already have the correct sized injectors installed and it's easier to build 2 single manifolds than one dual - I'm lazy

.

tobyanscombe · Post by **tobyanscombe** » Thu May 05, 2005 5:11 pm

Mmm Read this with interest as I am looking at how I can get EFI set up on a midget. My current plan is to get the ignition sorted and then play with TB's and injectors.

Hadnt thought of bike TB's....Just the Jenvey site (single TB with 2 injectors)

mikesb · Post by **mikesb** » Sat May 28, 2005 5:56 pm

I think someone asked for an update, and I posted one - it looks like both were lost by the backup.

Had some things keeping me busy so I hadn't gotten back to this for a while. I'm still sorting out some things, like the surge tank, air filters, etc. But I'm still playing with it and I've even gone for a short drive.

From what I can tell, the idle is lumpier than with the carbs, and this could well be due to the siamese ports. OTOH, it could be my req_fuel, my pw at idle is pretty short - around 2.6ms IIRC.

In any case, as soon as I crack the throttle all of the lumpiness disappears and it really picks up. I'm looking forward to see how it goes when I've got everything sorted.

arnold · Post by **arnold** » Mon May 30, 2005 3:37 am

I am running!! A-engine with a TBI and a supercharger blowing! It works, I have beenm driving around, no real problems since I moved the throttle upstream in front of the blower. I have to do all 'fine tuning' (VE and cold start), but it works (wet manifold, 2 squirt per cycle).

mini1071s · Post by **mini1071s** » Mon Jun 06, 2005 5:19 pm

So how are the projects going? Any updates. Going to try the single TB dual injector wet manifold route. Are there any starter maps for this set up on the a series engine? Thanks

Peter

tobyanscombe · Post by **tobyanscombe** » Tue Jun 07, 2005 1:01 am

Peter - I will have the ignition running in early July and will then start to add a couple of wideband O2 sensors. Then I shall be in a postion to look at injector sizing and which TB to use - Currently the Jenvey 48mm single with 2 injectors is looking like the best bet.

Dyagciog · Post by **Dyagciog** » Wed Jul 06, 2005 11:37 am

Any one have updates on injection the A series????

tobyanscombe · Post by **tobyanscombe** » Thu Jul 07, 2005 11:51 pm

Pickup my engine with crank sensor today so should be able to check the 10Deg limp home mode over the weekend.

I have a RR session booked next week so I am looking to get ignition mapped then. Then a slow build of all the sensors and TB. I hope to get fuel sorted in a month or so.

Anyone got a base ign map for an A series?

arnold · Post by **arnold** » Fri Jul 08, 2005 3:26 am

I am running, A-engine 1293, supercharged. Did a bit of tuning (MS tweak), things look very promissing, I only have some spikes which I have to tackle now. But that has noting to do with the engine itself or the setup.

I have no leaner cylinder (checked the plugs). Got it almost good enough to make the trip from the workshop to my house (100 miles) to finish tuning

cng1 · Post by **cng1** » Mon Jul 11, 2005 3:30 pm

My latest update is that I've managed to pass our emissions test but for the most part it's running hugely rich. That's mainly down to the standard injectors I'm running being sized for sequential injection so they're twice as large as I really want to be running.

If I run 2-squirts per cycle I gain enough resolution to lean it out, but I want 4-squirts to eliminate any charge-robbing. MS2, with it's high resolution code will cure the problem, as would smaller injectors. As it stands though it is driveable albeit a bit jumpy in traffic and it is mappable, albeit slightly painfully. I'll give a further update when I have some more news.

cng1 · Post by **cng1** » Tue Jul 12, 2005 1:38 am

"cng1" wrote:
> That's mainly down to the standard injectors I'm running being sized for
>sequential injection so they're twice as large as I really want to be running.

In fact having just reviewed the logs, my peak duty cycle is 30% which does
help explain why I'm struggling with the resolution. Time to order some
smaller injectors I guess.

Posted by email.

tobyanscombe · Post by **tobyanscombe** » Sat Jul 16, 2005 6:20 am

Not quite as sexy or advanced as Chris but Ladies and Gents - we have ignition!

Any pointers to how to sort a map out or better still a base map for the A series?

Oh, and how to not blow a hole in a pistion while playing....

cng1 · Post by **cng1** » Sat Aug 06, 2005 2:17 pm

I've finally had a bit of time to try mapping my mini. I've swapped down to some 208cc/min injectors from the monsters that rover use from the factory.

The smaller injectors seem to be ideally sized but running 4-squirts per cycle it's very lumpy pulling away from idle, MS2 will fix that though as it's just down to resolution. The TB is also a bit on the large size which means that a gentle cruise is at 90kpa which isn't helping but I think we can cope with that.

There seems to be very little power beyond 4000rpm which is either down to needing more advance or a "towing spec" camshaft.

Mapping is a trickier than other cars I've done but all in all I think we've thoroughly disproved the theory that you can't do port-injection on a siamesed engine. Now if anyone from this date forth starts talking about charge-robbing, sequential injection or how you need to run a wet manifold then I shall personally hunt them down and blow a rasberry at them.