Advantages of Port Injection?

[Vic_N]

What are the advantages of port injection over throttle body injection? With port injection, primary considerations for OEM's for MPI over TBI are instant starts and idle, rather than 2-3 cranks until the fuel reaches the cylinder.

 

With TBI, there are fewer parts, with only two injectors. Air and fuel are completely mixed in the plenum and then drawn into the cylinders. Fuel will also be vaporized, rather than droplets. With port injecting, injectors running at more than 30% duty (as soon as you get into boost) stay open longer than the valves. You are spraying fuel on a closed valve, and it puddles, then the puddle gets sucked in next time the valve opens. At 60% duty, half your total fuel is a puddle on the valve. By 90% duty, 2/3 of your fuel is drawn in as a puddle. Not great for vaporization. It has also been shown, the more vaporized the mixture, the better it burns, the lower emissions, and the higher mileage you get.

 

So what are your opinions on TBI and MPI?

[Boosted_One]

TBI is better than carberated. Better atomization of fuel and more precise control over the fuel metering.

 

In the same fashion MPI is better than TBI.

 

Better air and fuel distribution due to the seperated runners and an injector based right near the head port. Fuel is shot almost directly at the valve so puddling in the plenum is not an issue.

 

Overall MPI is leaps and bounds better than TBI. No comparison in overall efficiency and bottom line results.

 

On my dyno runs the MPI Magna made 50 WHP more than the former TBI set up. Proof is in the pudding.

[shift1313]

even if the fuel does puddle on the back of the valves which most mpi systems do its still better then tbi.  tbi still does the same thing and also is horribly inefficint.  i cant remember cfm numbers but i think the exact same manifold running mpi flows double the cfm then a tbi setup.  

[Ben_D]

The real hot set up is direct fuel injection (as seen in a 1954 M-B 300SLR (gullwing) But thats far away from the starquest world     But another note is even with an injector firing while the valve is closed the fuell will not puddle due in part buy the high tempatures of the intake valve. The heat will vaporize the fuel. But the A/F mixture will still wait untill the valve reopens but now rather then there being atomized (spelling?)fuel  there is a complete a/f mixture.  Another key advantage to to a MPI set up is thata "wet" manifold carb or TBI will flow around 5% less air (CFM) then a dry manifold. Due to the expansion of the fuel and the weight and volume of it

Fun stuff, some very smart people on this board!

Ben D

[eeaston]

What is all this talk of puddling?

 

As has been mentioned, any liquid fuel that does happen to make it to the back of the intake valve will vaporize because the valve is so stinkin hot.  Formula and other high-tech cars use injectors that don't atomize fuel at all, it squirts out as a solid stream.  They time it (sequential injection) so that ALL the fuel is sprayed when the intake valve is closed and it's aimed at the backs of the intake valves!

 

The other thing to realize is that the time between valve openings is so small that the fuel barely has a chance to do anything before it's drawn(pushed) into the cylinder.  If you're running at 5000RPM, only .024 seconds pass between one cycle and the next on any given cylinder.  Figure that the intake valve is open for around a fourth of that time, so it's only closed for about 18 milliseconds.

 

TBI systems produce poor fuel distribution because the runners cannot be exactly the same(length, shape, restriction) for each cylinder, plus there is typically more restriction due to shared runners and extra bends.

 

One more thing-a properly designed MPI manifold can be tuned to give a power increase through a given RPM range just like a tuned exhaust.  The runners can be as long as necessary without any change in fuel atomization.  Long runners on a TBI setup are where fuel really puddles.

[importwarrior]

at what temp does gas start to boil?

 

if it is sitting on the hot steel valve will it boil or is it too fast of a cycle to even get to that point?

[shift1313]

one thing to remember, say with water.  that the boiling point is a certain temp(as with fuel)  say you have a hot exhaust( something everyone has)  and you squirt some water on it, does it boil or just vaporize?  same thing with the fuel on the valves.   the temp of the valve is way past the threshold so it vaporizes.  and as was said there is a very small 'window' where the valve is closed.  

 

combustion temps are 1200+ F  , i dont know for sure but i would imagine anything over 200 degrees would cause vaporization.  anyone know some actual numbers?

[importwarrior]

thanks for the reply.  

[heefner]

I was reading somewhere about a cool can for fuel, basically it cools your fuel with ice.  People were saying that on a dyno they were noticing an increase in power.  If the fuel is going to vaporize the second it makes it to the head, whats the point.  

[Joel]

The point is that cool fuel is not boiling fuel, its not already vaporizing, and thus allows a more dense fuel to be pumped and injected before it gets vaporized.

 

Sequential injection has some decent advantages. One of them is called injector phasing. Injector phasing allows you to "phase" an individual injector to be timed for fuel hit the back of the valve just before it opens. With batch fire/MPFI firing batches of injectors, phasing isn't allowed, and thus doesn't see the benefits... its hit or miss.

 

When comparing the two for max power, you won't see a huge difference, we aren't even talking 15hp on a 500hp motor. We are talking less. On the other hand, if you want decent low speed and idle quality, sequential injection does it much better, and the reason is mentioned in the previous statement.

 

The problem with TBI/Batch fire is that sometimes the miss (as opposed to the hit), you don't hit the back of the valve, but rather you hit the valve after it opens. This ruins the mixture quality (don't read as lean/rich, but the actually quality of a homogenous mixture), thus can effect low speed/idle conditions greatly. You will even see that proof in the stock good running conquest. No matter how well you make it run, the quest just doesn't run as good as it goes at WOT.

 

WOT is less effected by the vaporization of the fuel at the valve. The reason is at higher throttle angles and higher RPM, there are much more turbulence and air speed to effectively vaporize the fuel into a higher quality mixture than at lower speed cruise/idle.

 

Another reason to cool the fuel down is to lower the pump temperature, which lowers the output fuel temps as well as pumping in or above pump specs, and prolongs the life of the pump.

 

Joel

[conquestpa]

not even to mention the fact that buying 2 brand new injectors for a TBI starquest will cost you almost what a MPI converion cost among the other things listed here.

[heefner]

This ruins the mixture quality (don't read as lean/rich, but the actually quality of a homogenous mixture),

Joel

 

Damn language filter.  You can't even type h0m0genous !>

 

So basically its like an intercooler for your fuel.  

[Joel]

Oooh kay... Languauge filter is a database is a little to large IMO. Yikes.

 

Joel

[Boosted_One]

homogenous homogenous homogenous homogenous homogenous homogenous homogenous homogenous homogenous homogenous homogenous homogenous

 

 

If you see things like this with the filters please report the post to a moderator.

[shift1313]

as conquestpa said, buying starquest injectors is outrageous.  I built a DIY FI comp for under $100.  it has the option for batch fire or ping pong fire, which is half and half.  timing is hit or miss with a distributor type ignitions.  you can time the cycles better if you are running a crank trigger.  also as was stated the hp gains from either way is minimal.  the main reason big companies do this(and very few do) is for mpg, its much more efficient fuel wise if you can time it right.  i would rather puddle the fuel then risk it, imo.

[chiplee1]

timing isn't "hit or miss" with my distributor.  With a weights and vacuum advance type distributor it may be.  The hawk however, provides full digital control of timing.  If anyone has a tutorial I could read about electronically controlled timing via distributor I'd love to read it.  I've only sort of wrapped my brain around how the Hawk manages timing, but I do know it's not hit or miss.

 

I was told sequential fire systems are used by factories as much for throttle transition performance as it is for fuel economy.  In other words.  Batch fired systems are a bit unruly in the transition from off throttle to part throttle.  They tend to sputter and perform better under accel than part throttle cruise.  This makes perfect sense.  My car even does it, and it's batch fired.  Almost all of that can be tuned out by a pro, but that's not me.  WOT performance is virtually identical from batch to sequential fire systems.

[Joel]

timing isn't "hit or miss" with my distributor.  With a weights and vacuum advance type distributor it may be.  The hawk however, provides full digital control of timing.  If anyone has a tutorial I could read about electronically controlled timing via distributor I'd love to read it.  I've only sort of wrapped my brain around how the Hawk manages timing, but I do know it's not hit or miss.

 

I'm not talking about ignition timing. I'm talking about injector phasing. It's hit or miss on any batch fired or multipoint system... unless the hawk support sequential injection with individual injector trim a phase settings, you are at the same disadvantage, and the injectors hit or miss the 'exact' event.

 

I was told sequential fire systems are used by factories as much for throttle transition performance as it is for fuel economy.  In other words.  Batch fired systems are a bit unruly in the transition from off throttle to part throttle.  They tend to sputter and perform better under accel than part throttle cruise.  This makes perfect sense.  My car even does it, and it's batch fired.  Almost all of that can be tuned out by a pro, but that's not me.  WOT performance is virtually identical from batch to sequential fire systems.

 

Thats basically what I said in my last post... although many have seen so-called slight improvements in power just switching to sequential. Who knows if thats really the case, as each run on a dyno can be off by a certain error margin. In any case, part throttle to full throttle stumbles can be solved with a throttle pump setting. Usually there will be both a injection increase and a sustain setting. Try raising one and then the other in stages until it no longer misses on throttle movement.

 

Joel

[eeaston]

timing isn't "hit or miss" with my distributor.  With a weights and vacuum advance type distributor it may be.  The hawk however, provides full digital control of timing.  If anyone has a tutorial I could read about electronically controlled timing via distributor I'd love to read it.  I've only sort of wrapped my brain around how the Hawk manages timing, but I do know it's not hit or miss.

 

Chip...  Electronically controlled timing is fairly simple really.  I don't know if the Hawk uses magnets in the crank, a cam sensor, or a toothed wheel, but the function is the same.  The magnet type is slightly more calculation intensive, but still very simple.

 

The ECU recieves a pulse or two from the crank sensor per revolution(two for a cam sensor).  It knows exactly where in the crank's rotation that pulse is produced.  When the signal is recieved, the ECU starts a timer.  It knows the exact speed of the crank from timing the last revolution, and it knows whether the crank is accelerating or losing revs by looking at the last few rotations.  It extrapolates a very close estimate of the crank's speed on the current rev from this data.  It calculates a time delay corresponding to X degrees of advance/retard(from your ignition map) and waits that amount of time then sends an ignition signal to the ignitor(ignition module).  The ignitor relays the signal to the coil(adding in some dwell) and the distributor simply sends the spark to the appropriate plug.  The distributor timing doesn't have to be right on, but it has to be in the ballpark, because the rotor is lined up with each contact for maybe 10 or 20 degrees of cam rotation.  As long as the ignition pulse from the ignitor falls within that time frame, you're good.

 

The system for milti-toothed wheels is very similar, but the ECU has slightly more up-to-date info about the crank's speed and position because there's a signal every 10 or so degrees instead of 180 or 360.

 

I'm sure this is all clear as mud, but hopefully it helps.

 

Ford's EDIS ignition works differently and is very clever, for info on it check out http://www.bgsoflex.com/mjl/mjl_edis_summary.html  This is what I'll be running with my Megasquirt ECU once the car's back in commission.  Can't wait!