DILYSI Dave
DILYSI Dave SuperDork
4/11/11 3:30 p.m.

Honda 4 cylinder. Firing order is 1-3-4-2. Megasquirt will support 2 separate injector channels, meaning I need to inject the fuel for 2 cylinders at once.

Option 1 - Group 1-4 and 2-3. This is the common way to do it. Reason being I suppose that that's the way the crank groups them. But those pairings are 180* opposed in firing order, so if you are timed perfect for one injector, you're completely imperfect for the other.

Option 2 - Group 1-3 and 4-2. Follow the firing order. It seems like you could more accurately emulate sequential injection this way. This way, if you're timed perfect for one, you're only 90* off for the other.

In reality, I realize it doesn't make a damn bit of difference when we're talking about duty cycles of 80% and 8000 RPM, but still...

alfadriver
alfadriver SuperDork
4/11/11 4:50 p.m.

Option 1 is better.

Have you checked the injection sequence for your computer? I've not found a batch fire that doesn't fire once per revolution- so for each cycle, it injects twice- once "correct" and once 360deg out. (nit pick correction- 1-4 and 2-3 are 360 deg apart in firing order....)

One other note- mixed results from open valve injection. In theory, it will allow you to run more spark, but you need a dyno to check it. The trade off is on a cold engine, you'll inject liquid fuel into the combustion chamber (and possibly on the walls....). For safety, I would start on a timing that does closed valve injection. It works fine.

fifty
fifty Reader
4/11/11 4:52 p.m.

I went with option one on my 4 cyl VW engine (same firing order) - worked fine.

Of course, with MS2 you can have true sequential. You need a toothed wheel on the crank, and then something for cam timing - the dizzy would work fine for this. http://jbperf.com/sequential/index.html#ms2mods

DILYSI Dave
DILYSI Dave SuperDork
4/11/11 5:00 p.m.
alfadriver wrote: Option 1 is better. Have you checked the injection sequence for your computer? I've not found a batch fire that doesn't fire once per revolution- so for each cycle, it injects twice- once "correct" and once 360deg out. (nit pick correction- 1-4 and 2-3 are 360 deg apart in firing order....) One other note- mixed results from open valve injection. In theory, it will allow you to run more spark, but you need a dyno to check it. The trade off is on a cold engine, you'll inject liquid fuel into the combustion chamber (and possibly on the walls....). For safety, I would start on a timing that does closed valve injection. It works fine.

Thanks!

DILYSI Dave
DILYSI Dave SuperDork
4/11/11 11:18 p.m.

And done.

MadScientistMatt
MadScientistMatt Dork
4/12/11 8:33 a.m.
alfadriver wrote: Option 1 is better. Have you checked the injection sequence for your computer? I've not found a batch fire that doesn't fire once per revolution- so for each cycle, it injects twice- once "correct" and once 360deg out. (nit pick correction- 1-4 and 2-3 are 360 deg apart in firing order....)

MS can be set for anything from once every other revolution to twice per revolution on a four cylinder.

alfadriver
alfadriver SuperDork
4/12/11 8:39 a.m.
MadScientistMatt wrote:
alfadriver wrote: Option 1 is better. Have you checked the injection sequence for your computer? I've not found a batch fire that doesn't fire once per revolution- so for each cycle, it injects twice- once "correct" and once 360deg out. (nit pick correction- 1-4 and 2-3 are 360 deg apart in firing order....)
MS can be set for anything from once every other revolution to twice per revolution on a four cylinder.

For batched fired, the least compromise would be once per revolution. that way, the pair will get two injections, and one could schedule open valve injection.

The one note that I would make, do make sure that MS is capable of overlapping the injections on the two drivers. Last I heard, it was capable of this, but I ran into this problem on a TECII, which limited a pair batch to half the time it could have. If I put all of them together- as in twice per rev, it would have worked- but for whatever reason, the TEC was not able to schedule the injector drivers at the same time at different timings... odd. I'm pretty sure MS IS capable of that, but you should double check it. No reason to cut your injector capability in half just due to a driver issue....

MadScientistMatt
MadScientistMatt Dork
4/12/11 11:11 a.m.
alfadriver wrote: For batched fired, the least compromise would be once per revolution. that way, the pair will get two injections, and one *could* schedule open valve injection. The one note that I would make, do make sure that MS is capable of overlapping the injections on the two drivers. Last I heard, it was capable of this, but I ran into this problem on a TECII, which limited a pair batch to half the time it could have. If I put all of them together- as in twice per rev, it would have worked- but for whatever reason, the TEC was not able to schedule the injector drivers at the same time at different timings... odd. I'm pretty sure MS IS capable of that, but you should double check it. No reason to cut your injector capability in half just due to a driver issue....

Kinda surprised an Electromotive ECU would not do this. MS has done overlapping injection since the beginning.

alfadriver
alfadriver SuperDork
4/12/11 11:22 a.m.
MadScientistMatt wrote: Kinda surprised an Electromotive ECU would not do this. MS has done overlapping injection since the beginning.

I found out the hard way- my challenge car ran out of fuel at the top end... Thankfully, it didn't break anything. I paniced and cut the tips off the injectors, and then found out that I can batch fire the entire set later.... Oh, well.

Really surpised me, too!

John Brown
John Brown GRM+ Memberand SuperDork
4/12/11 11:30 a.m.
alfadriver wrote:
MadScientistMatt wrote: Kinda surprised an Electromotive ECU would not do this. MS has done overlapping injection since the beginning.
I found out the hard way- my challenge car ran out of fuel at the top end... Thankfully, it didn't break anything. I paniced and cut the tips off the injectors, and then found out that I can batch fire the entire set later.... Oh, well. Really surpised me, too!

Ahh... the patented Eric Storhok 36lb injector mod!

alfadriver
alfadriver SuperDork
4/12/11 11:35 a.m.
John Brown wrote: Ahh... the patented Eric Storhok 36lb injector mod!

I can't take credit for it- someone told me about it. And it only works on some injectors- others- the tip is part of the pintle seal.

AWSX1686
AWSX1686 GRM+ Memberand Dork
11/2/17 9:15 a.m.

Resurrecting an old thread here since I will likely be doing a MS based build here soon, and I am not familiar with the whole batch injection thing. It seems like it is basically the fuel injection version of wasted spark, is that right?

 

Wasted spark would spark a pair of cylinders, one on the ignition stroke, the other one the exhaust which would be wasted. 

So batch injection would do the same? Injecting a pair of cylinders one when it is supposed to and one that would just go out the exhaust? Or the intake valve would be closed, so it would build up for the next injection effectively doubling the amount injected?

MadScientistMatt
MadScientistMatt PowerDork
11/2/17 10:52 a.m.

The firing can be similar to wasted spark, but the unused fuel generally sticks around in the intake port until the valve opens (some of it can be sucked down the wrong cylinder due to intake reversion effects). It's only going to blow out the exhaust if you have a massive amount of cam overlap - and in that case, sequential fueling is only going to be able to prevent a small amount of that.

alfadriver
alfadriver MegaDork
11/2/17 12:21 p.m.
MadScientistMatt said:

The firing can be similar to wasted spark, but the unused fuel generally sticks around in the intake port until the valve opens (some of it can be sucked down the wrong cylinder due to intake reversion effects). It's only going to blow out the exhaust if you have a massive amount of cam overlap - and in that case, sequential fueling is only going to be able to prevent a small amount of that.

And to follow up on that, for the most part, OEM level PFI calibrations inject on a closed valve.   So having it build up in the runners isn't that big of a deal.  Even doing SEFI injection, it's on a closed valve- so even on the very rare occasion of blow through- it just would.  But that's really rare.

AWSX1686
AWSX1686 GRM+ Memberand Dork
11/3/17 9:11 a.m.

Ok, thanks guys! I think I am starting to get an understanding of how this works!

(Which is great because it looks like I may have found a good source for a standalone computer for the Challenge car. devildevil)

xflowgolf
xflowgolf Dork
12/29/17 8:38 p.m.

Seeking GRM input here.  Trying to sort out my partially megasquirted NA 2.0 VW 16V 9a 4-cyl.  I untaped the injector harness and found the PO has tied cyl 1&2 on bank 1, and 3&4 together on bank 2.  

Should I bother redoing this so it’s 1&4 / 2&3 together?  

Fwiw it’s never run, so I’m just trying to get it going for the first time.  

Stefan
Stefan GRM+ Memberand MegaDork
12/29/17 9:34 p.m.

Set it up for two squirts per revolution and it won’t matter in the long run.

Look at the firing order extended a bit further and think about the injectors firing in pairs.

1-4-3-2-1-4-3-2-1-4-3-2

it doesn’t really matter what the pairing is since you’ll always be firing a pair at a closed valve and an open valve.

With a motor with more cylinders it would be a larger issue, obviously.

if you’re a perfectionist, or you want to confirm the wiring is correct, go for it, but I don’t know that it matters.

xflowgolf
xflowgolf Dork
12/29/17 9:37 p.m.
 

if you’re a perfectionist, or you want to confirm the wiring is correct, go for it, but I don’t know that it matters.

Ha.  After typing that I already cut it up and redid it 1/4 & 2/3.  yes

alfadriver
alfadriver MegaDork
12/30/17 7:56 a.m.
Stefan said:

Set it up for two squirts per revolution and it won’t matter in the long run.

Look at the firing order extended a bit further and think about the injectors firing in pairs.

1-4-3-2-1-4-3-2-1-4-3-2

it doesn’t really matter what the pairing is since you’ll always be firing a pair at a closed valve and an open valve.

With a motor with more cylinders it would be a larger issue, obviously.

if you’re a perfectionist, or you want to confirm the wiring is correct, go for it, but I don’t know that it matters.

 

As a side note, unless you are really confident in your injector aiming and the pattern you have, I would suggest not injecting on an open intake valve- to good of a chance at bore wash.  Up until direct injection, all OEM injection was on a closed intake valve- generally timed so that the end of the injection is just prior to the valve opening.

If one is worried about injection timing, especially with batch injection, I'd suggest plotting out a line of 720 deg and draw out all of the events there-that will make it clearer where to inject the fuel.

GameboyRMH
GameboyRMH GRM+ Memberand MegaDork
12/30/17 9:46 a.m.

^This. Without the excellent aiming and spray pattern of injectors on modern engines, open-valve injection runs like garbage.

This is the same injector timing table I'm using for my 4age (stretched to the right with a few more columns for higher revs), set semi-sequential, but also running 4 squirts per cylinder, so I'm not sure how much difference it makes:

You'll need to log in to post.

Our Preferred Partners
8CIsK6kgnnYdCjpX6BdbW5wKt09CyJMye36WdSQsMgAbXJAIC0AO2MJ90WxA5YAS