[punisher454]

Started working on a 2 stage external injector driver board. 8 channels of input and one extra input to select stage A or B (or A+B). I started laying out the board, not quite finished, need to adjust some trace spacing and via's.
It uses a blue pill to process the inputs and outputs with a little of logic to keep any open injector(s) in the open state while switching stages to avoid lean fueling events. I might program it to optionally hold the first stage open one more complete cycle when switching for manifolds with secondary injectors mounted a lot farther upstream than the primaries(like on a tunnel ram setup).
The inputs are 8 logic level high/low signals, plus one channel to select A or B(primary or secondary stage). A jumper can be shorted to change the staging strategy from "or" to "and".
For my own use I want staged injectors to switch to really high octane race gas when I hit around 90kpa on a twin turbo LS that has 11:1 compression. This is for Drag & Drive type races so I can run on pump gas during the drive and even staging. E85 isnt really available in my home area, and Im not gonna street drive on alky or race gas all the time, so a dual fuel setup seems like a good solution.
Anyhow , probably in a month or two I might have a batch of boards made if anyone is interested I'll order some extras. its all thru-hole (except the Fets, which are easy) so if you can build a speeduino v4xx this will be easy. Its pretty small just under 3"x 4".

BTW, before I get flamed for the overkill of a Bluepill to do simple switching, I just happen to have a bunch of extras and they seem more cost effective than even simple discreet components, and I have a future possibility of SPI control or whatever.

[PSIG]

Interesting approach. I presume the existing staged injector function is desired, but with 4+4 channels instead of 2+2 on a Mega? Have there been issues with the many boosted LSs on 2-channel (e.g., MicroSquirt), or is this just in general principle?

2nd thought is that most engines will induct far more than the intake volume, so delayed/modded transition fueling for top-of-runner vs bottom shouldn't be necessary? Again, just knee-jerk thinking. Is this an assumption or is there data for this? I don't care which; I just want to know the dynamics and if they are established or not, for how to diagnose for the solutions.

[punisher454]

I’m not using a 4-channel speeduino. My application(s) is fully sequential V8. Currently I have an STM Pro (8x8) but I’m going to move that to another project car and I’ll be using a custom board for this one.
As far as the extra squirt during transition, that is just an option that will be disabled unless for some reason I want it active (probably won’t need it). My reasoning was that I was thinking about wall wetting and more distant injector placement. I have a slight paranoia about hiccups under boost, which is why I’m switching to the secondary fuel system at around 85-90kpa instead of 100.
Last time I looked, staged mode is disabled for 8cyl sequential, understandable since the ecu does not have 16 injector channels. So I’m planning to setup the secondary fuel and spark to trigger by map sensor and activate a single output pin at the same time.

[PSIG]

Got it. I assume from the Wiki and playing with settings, that staged is for paired setups, not sequential or more cylinders than channels. This makes sense, as the number of channels (8) would be cut in-half to allow primary (4) and secondary (4) sets of output channels. If you selected your Engine Constants to Paired, the Staged option would become available (8 Primaries on channels 1-4 and Secondaries on 5-8). Sequential option then grays it out.

With an 8-in 16-out board, the primary full-sequential benefit of ICFT would only be effective on either primary or secondary channels, possibly pooching the other. Hmm. Not trying to rain on your cool concepts at all, but trying to grasp the benefits you're after without losing others. What benefits are you after with sequential, and what benefits are you after with staged? I am tempted to assume, but those project requirements and answers are yours.

[punisher454]

TWICE in a row I wrote a long detailed description of what I'm trying to do, my reasons, and what I still dont yet know.
Took too long writing, cuz when I hit send i have to re-login and my message was lost. Should have copied it BEFORE sending. Very annoying!!
I'll respond again tomorrow

[punisher454]

OK shorter answers...
Part 1 ICFT:
I have a few engines with tunnel ram intakes that have front feed plenums. This arrangement tends to cause a rich condition in the front two cylinders and a lean in the back two. Its not a fuel problem, its airflow. Tunnel Rams are originally designed to be down-flow with a pair of Holley carbs on top. the throttle-bores all had a nice line of sight straight down to the runner entrances, all fairly symmetrical and worked well. Sometimes more advanced tuners would stagger the jetting when plug readings showed variations.
When we add plenum tops with a front feed the air shoots over the front runners quickly and they starve a little, the middle 4 are fairly even and the back two get more air deflected down into the runners causing them to run a little lean. its not dramatic, and is usually unnoticed by most tuners. It can also be compensated for with porting. You can also adjust with ICFT. To tune this you'll need to monitor the AFR's individually for each cylinder, overkill for a street car or most any stock engine. Worthwhile on a race motor that will run a lot closer to the edge.

Some example pics.
The Small block chevy Holley Stealth Ram and Accel super Ram for big blocks are stock pics from the internet. I have one of each in the garage. The Offy big block tunnel ram that has the plenum cut open is an unfinished past project from over a decade ago. I'm going to change it back to a down-feed configuration with a pair of large DBW throttlebodies and weld in the 16 injector bungs I need. My small block HSR is getting a plenum spacer and 8 more injectors. the big block Super Ram is an unmolested beautiful piece of art, but i think ill try to find the original style edelbrock "TR"-something(cant remember the original model) plenum top for it and keep the nice polished Accel top in a box.

But anyhow, I have a real use for icft, and its airflow related so i think it may still work (maybe) during primary and secondary fueling modes.
More to follow soon...

[punisher454]

Part 2, 2 injectors per cylinder:

Some people run two injector per cylinder strictly for fuel flow reasons. Smaller injectors for idle and cruise and then switching to, or adding (more common)a second set for more fuel as required.
I think the built in staged injection features in speeduino assume the same size injectors for both primary and secondary. But being able to specify a second set of injector parameters for the secondaries would sure be a nice feature when using the built-in staged injection feature (which I'm personally not able to use on my application).
Some example dual injector manifolds...

[punisher454]

Part 3 , Sequential.

Not a Must have, Just a want to have.
So far in my experience an engine will run fine a single channel batch fire. Chevy TPI used on camaro's and vettes was alternating bank batch fire (2CH) and the Cyclone and Typhoon used basically the same ecu, but ran a different program that was single channel x6 injection (V6), turbocharged and fast as hell.
My desire to run sequential is for improved idle control. On some of my builds I run some pretty nasty cams. Over 250degrees at .050" with LSA's around 106-107 degrees. Cant make them smooth, don't even want that, but ANYTHING that helps even a tiny bit to keep it under control is IMHO a worthwhile effort. In my Carburetor days I spent countless hours and days messing with jets, air bleeds, drilling throttle plates, power valve channel restrictions, power valves, wires in the metering block orifices to help restrict flow, drilling orifices to add more flow.... With a little tuning work you might get a somewhat stable 1100-1200 rpm idle, and that where a lot of racers stop. But with a LOT more work you might get that same angry beast down to 850 or so (in gear with like a 3500-4000 rpm converter). Idle vacuum might be only 4.5" but idling in gear without it dying or surging is nice.
With sequential injection I'm hoping to get MAYBE just a little more control of a nasty idling cam. I've gone to the extreme before (and will again I'm sure) of plumbing in the IAC system to its own manifold and running individual tubes into the ports like they did on the factory LT1 manifolds. Helps a little, not a lot, but ill take every little bit I can get.

[punisher454]

Part 4, My intended setup in Speeduino:

So what i think I'll do is setup the secondary fuel and spark tables for my upper(secondary)injectors on race gas. Set the tables to activate conditionally with something like 90kpa as the setting. I will also setup a digital output to trigger at 90kpa.
Or maybe setup the digital output to activate at 90kpa and have the secondary tables to activate triggered by THAT output (if it will work that way).
The secondary fuel table is tuned for ONLY the secondary fuel system to be supplying the engine, NOT both.
The external controller (what I'm building now) reads the 8 input channels (injector outputs from speeduino) it either just passes them through to the drivers on the "A" bank (1A-8A) OR it passes the signals to the drivers on the "B" bank (1B-8B). Depending on what the state of the switch selector input line is (high or low).
A jumper on the board can be shorted so that when it switches to secondary mode (B) it fires BOTH the A and B injectors(logic "AND" mode). For other applications than mine this is possibly the desired mode.
In the uC (a cheap but fast bluepill) when the primary/secondary selector input changes state, the controller will finish out any injection event on the primary that is in progress, while also activating the secondary side. This is to prevent a fueling cycle to get cut short while the secondary injector does not have a chance to open up yet.
I figure one cylinder firing rich is a whole lot safer that firing lean. As for the option to fire both banks simultaneously for one (or more) cycles at transition, it probably wont be necessary at all, but its easy to write it into the program and just not use it, than to someday need it and not have it. Probably disabled by default type thing.
I dont think Speeduino secondary tables have any hysteresis features, maybe I'm wrong. But it would be nice to be able to have different set points going up and down.
The way I'm building the board right now it can use direct outputs from the uC in speeduino, making speeduino's injector drivers un-nessesary. I do need to probably adjust the design to be able to read the output from the actual speeduino injector outputs as if the wires were going directly to the injectors instead of this board. I'll probably adjust the design a little.

[PSIG]

TWICE in a row I wrote a long detailed description of what I'm trying to do, my reasons, and what I still dont yet know.
Took too long writing, cuz when I hit send i have to re-login and my message was lost. Should have copied it BEFORE sending. Very annoying!!
I'll respond again tomorrow

If you haven't, try checking the log-in box to "Keep me connected" or some such wording. I used to lose lots of posts when I answered the phone or something. Gah!

Thanks for the extensive write-up. It will not only help us to get an idea of how you're approaching it, but also to document the project and follow-along. have you considered "auto" so the injector flows blend, and octane from the second fuel increases average with load? Is that an option? Having tried staged with switched injectors, they were a bit clumsy to transition, and the blended action makes the transition effectively seamless, in-spite of the non-linear big injector flows initially. Small primaries made the transition earlier and with excellent low-load control, and no funky fuel pressure manipulations, just a standard 1:1 FPR.

You might want to consider an airflow divider to turn the airflow fore and aft. It splits the flow top and bottom halves. We used one in a hard 90° turn into the manifold (under-hood clearance) where it was leaning the rear cylinders at top-end. The divider simply forced half the air into the front and the other half to the rear. Worked great. Crappy concept sketch:

flow_divider.jpg (8.47 KiB) Viewed 3710 times

I found interesting photos of a soon-to-be-released Venom manifold casting before machining. Another 16-injector V8 manifold :

Venom_16_inj_2.jpg (63.8 KiB) Viewed 3710 times

Venom_16_inj_1.jpg (71.5 KiB) Viewed 3710 times