The L98 "two bank myth" debunked.
 

I've always wanted to do a rigorous step-by-step proof that all the injectors in a L98 are connected in parallel and NOT in two banks (which I call the "two bank myth")...

I believe the myth was started by Grumpy's (Bruce Plecan -- R.I.P.) "L-98 Engine Start Sequence". I believe he first posted this on his own web site (now gone) and now it's posted all over the Internet. He gets several things wrong in the "ECM Reaction" section. First, the ECM does not have a way to read oil pressure. Second, there is only one injector driver. Thirdly, the two banks do not fire alternately (this is where I believe the myth was started):

http://www.corvetteactioncenter.com/...0/016/article/

From the Factory Service Manual, page 6E3-A-24:

http://www.misterpeachy.com/VettePic...24_Diagram.pdf

As you can clearly see, the two banks are connected together inside the ECM at pins D15 and D16. Also note that there is only one "INJ DRIVER".

Here's one page of the ECM schematic, courtesy of Ludis Langens. Q1 (right side) is the (ONE!) injector driver. Its output is INJ#, which goes to another page (next schematic):

http://home.earthlink.net/~cliff_har...7165sheet4.gif

Here is the connector page of the schematic. Signal INJ# is shown in the lower right corner, where you can see that D15 and D16 are connected together:

http://home.earthlink.net/~cliff_har...7165sheet5.gif

Since I just happen to have a spare ECM sitting right here... Here's a picture of the PC board. As you can plainly see, pins D15 and D16 are connected together (lower right corner of the connector pins). It really surprised me when I first saw this because I had heard the "two bank myth" for so many years:

http://www.misterpeachy.com/VettePic..._D15_&_D16.JPG

Here's a picture of the (ONE!) injector driver transistor. It's a MOSFET in a TO-220 package, insulated from the heat sink because the heat sink is grounded:

http://www.misterpeachy.com/VettePic...tor_Driver.JPG

Here's an overview of the heat sink, with the white plastic piece that clamps the various parts to the heat sink. It has the (left to right) voltage regulator, two quad drivers, a couple of other parts and the (ONE!) injector driver on the far right:

http://www.misterpeachy.com/VettePics/ECM_Heat_Sink.JPG

:iagree: Nice work Myth Buster!

Then why would the designers choose to use two fuses?
I don't get it.

:cheers:

well, my thought on this is, and i'm not doubting cliff's theory, in fact, my initial thought is :iagree:. i think the 2-fuse, right and left bank thing, is the fuses are protecting the injectors themselves, at the component level. the ecm controls the timing, firing, and monitoring of the system. we might be getting the application of "apples and oranges" a little confused - ???

This is going to upset alot of people that thought they knew how the system worked... Me included....:ack:...WW

:iagree:

This would make a lot of sense.

If it were not for the two separate fuses, a short on one bank could take out both banks.

It would also help to isolate which side has the bad injector.

BTW: Nice write up Cliff.

This will help those handy with a soldering iron to trouble shoot their ECM if they should lose the
injector pulse.

:thumbs:

Bruce had the unfortunate malaise of working with as many EFI systems as he could and being a guru in them all. While I'm sure the "myth" you talk about was carried on, it was not something that Bruce himself perpetuated purposely. Many of the early CFI systems used dual bank injectors, and the early TPI source code (Which Bruce concentrated on) definitely had single and double fire procedures in them. The SyTy hardware was dual-banked and I know Bruce spend much of his time in that code as well as the turbo buick systems.

I remember reading write-ups like this from Bruce back in the early '90s, way before the 2007 posting date of the referenced article. If any of us working in the area back in those days were asked the same questions given the available info, we would give the same answers.

I'm not sure how many people here can say they ever met Bruce, but I did and I also worked and corresponded with him for many years until his passing. He actually purchased my neighbors '87 Grand National.

Great info !!!
Btw ...is the same grumpy am i thinking of ? from what i know he is alive
His site is still up
http://forum.grumpysperformance.com/index.php
maybe im confusing them

I've been saying the same thing for years.....ONE inj circuit split in two thats divided into 4 each side.

Makes NO sense...ZERO, NONE for there to be an alternate bank fire when the ign firing order flops back & forth alternating which side gets a spark..
The TRUE theory behind the TPI injection bank ()batch fire is more accurate) is to have them ALL fire so there was a fuel charge WAITING in the port as the intake valve opened.
Some proof of this comes from the minute hp increase from sequencial multiport fuel injection...a much much faster inj driver set, with 8 seperate channels that allowed each inj to fire as needed, not as a batch. This was very cool and a big atta-boy for the tech geeks at GM, but it yielded less than 20 hp for all that it did. True that big power comes from small steps and SMP EFI (multiport) did help, but it also proves that batch fire was 90% as efficient and a bunch less complicated. Remember, as the hp numbers went UP the MPG stayed where it was as these inovations appeared and gave balance to the new system.
The big advantage with multiport EFI is that now there was some lattitude in the cam timing and ign timing with the fuel timing now able to be customized to the cylinders so when cams or ign changed so could the fuel program...more small gains that eventually added up to big hp. Before, there was fuel being sprayed whenever...some cyl got wet fuel while others got mist...it found some balance in rpm at some point obviously, it worked, but it was not something that could be dialed in for precision.

Single circuit, split in 2 then by four. 2 fuses to cut the load on each side on the first split. Less likely to short with only 4 inj to protect. So, 2 fuses.They cut the odds in half that way.
This is also why our L98s suffered and were so sensitive to crappy injectors...not being efficient in the first place, anything less than 100% was going to cause a problem pretty quickly. Besides, batch fire was behind many a catastrophic failure....I'd MUCH rather have a single inj short out and run on 7 cyl than for the whole bank to flood...due to one shorted inj. (as me how I know).

Are you sure you have new info?

http://forums.corvetteforum.com/c4-t...bank-fire.html





BTW: Don't actually POST in that other thread as we all know that's a no-no.

Hey Cliff, thanks for giving solid info that clearly defines and qualifies your findings.

As a L98 owner its nice to have this thread printed and inserted in my FSM.

:cheers:

As Mike85TPI pointed out, Bruce worked on a lot of systems, especially the Grand National engines. I think he got it right that Bruce just got confused when he said the banks fire alternately (TBI systems do that, which Bruce spent most of his time with). Bruce also said in the article I referenced that the ECM looks at oil pressure, which is also wrong -- there is no oil pressure input.

My photographs and the schematic are for the 1227165 ECM, which was used from '86 through '89. I don't know how the other years are configured.

I looked at the diagrams in the linked thread posted by GREGGPENN. Wow! engle1147 sure went to a lot of work to put all that together. I did see that some of the diagrams are wrong, though. As I said, the 1227165 ECM was used from '86 - '89, but the diagram for 1986 shows the even bank connected to pin D1, which is ground. It also shows two injector drivers, not just one. The diagram for 1987 shows pins D16 and C1 connected together. C1 is the fan control output. Also two injector drivers. The 1988 and 1989 diagrams again show D16 & C1 connected together, but only one injector driver this time. Supposedly all these diagrams came from Corvette FSMs, so I can only conclude that the FSMs were wrong. I found a bunch of problems in mine...

I found that my FSM is marked up (by me) on the injector page (8A-20-6). I also have corrections on pages 8A-20-0 & 8A-20-1, which have the wrong connections for the TPS and MAF.

Ever since I learned to, and programmed my 1227165 ECM (in 2010), I'd decided that it's a batch fire system where the injectors fire once/revolution (compression and combustion). That means the injectors act twice their size.

It was the only thing that explained some very high HP applications running with 24lb SVOs.

Jon Banner, 87IrocZ, and Engle's post helped with that realization.

It was weird to me how the system works when I first started looking into it. I had a tough time with the fact that the injectors fire twice before the intake valve opens.

I was having a problem with my car and hooked up LEDs to both banks of injectors. I noticed that both LEDs were flashing at the same time, which seemed strange since I had heard the "bank fire myth" forever. Then I discovered the connection on the ECM PC board between D15 and D16.

One really surprising thing to me was when I was on the freeway and when I let off on the gas both LEDs went OUT. Then I found out about DFCO -- Deceleration Fuel Cutoff. The ECM really does shut off the injectors when you decelerate.