jcsbanks

Latency would apply to both sets of injectors though if they use the same signal. Latency is only really going to be a substantial part of the injector time at low fuel demand which is when the secondary set will not be operating. I can't see why anyone would need more than two sets of 1000cc injectors anymore than anyone would need more than 640K RAM

dan l

iTrader: (3)

I'm not so sure that timing of the switchover is extremely crucial, however I could be wrong. I think under worse case scenarios you would have a slight misfire on one cylinder which I doubt would be noticeable. I guess take a look at how AEM does it and if people have issues with secondary injectors. Also how much processor time would timing it soak up? I like to save the processor if possible from doing high speed checks.

I think with some searching we can find a low impedance driver boxes that take a 5 or 12 volt dc input. If not it should be easy to make. We would just have to make sure the darlington's are quick enough so that you aren't adding your own latency and/or hysteresis. Also, it could be 99% plug and play. Plug the primary injector plug into the box so that you can parallel off of it to grab the signal for the darlington pair. Then send it straight out of the box to its intended primary injector to clip on. At the output of the darlington put a plug on it so that you can simply clip into your secondary injector. You would only need to give the box power, ground, and the enable signal. We might want to do some testing to make sure we don't have to filter the darlingtons so that they don't accidentally trigger from time to time, however I don't *think* it will be an issue.

I think JohnBradley has experience with secondary injectors on the AEM. Can you comment on if we should/shouldn't worry about timing the phasing of bringing in the secondary sets on an enable?

jcsbanks

The timing is only an issue if you're suddenly halving the injector time and suddenly bringing on a second set of injectors (you need this to happen at exactly the same time), whereas with AEM you would gradually bring on the second set whilst leaving the first set near capacity.

I don't think processing time or power is the issue, it is coding it so that the switched output happens just after the last full length injection and just before the first half length injection. Otherwise I think it it should be fairly easy.

dan l

iTrader: (3)

Gotcha. I think that if the switchover point is chosen properly it won't be an issue as long as we time it correctly (I can now see a slight issue). IE: if the car achieved 360% load make the switchover point 280% load. Chose a load that the engine swings through really fast and that the driver doesn't spend much time throttling around. IE: it would be bad to have the switchover point at 320% load if the driver spends time around 320% load in first gear for example. The system would possibly be turning on/off.

Maybe slowly bringing in the injectors softly will be needed, but I sure as hell hope not. On a return style fuel system the secondary fuel rail should always be primed as fuel is constantly running through it. The injectors should be able to switch over and be perfect from the first pulse to turn on. Also I don't think I'm crazy about running the primaries at 80% and bringing in a set of secondaries which spend a significant amount of time at 5 or 10% duty cycle. I like the idea of splitting the duty cycle so as to not stress the primaries etc...

I think JohnBradley will have some good input.

jcsbanks

Agree they should be good to go electronically and physically, with thought it should be possible to arrange the switching so it is completely seamless.

Using TIP122 Darlington, we can switch the ground using the open collector, handle 5A (will handle injectors down to 2.5 ohms) and have a switching frequency up to about 1MHz.

JohnBradley

iTrader: (30)

As far as I know the AEM has the secondaries idle but ready to go and then will switch to "on". I am not 100% sure how fast that switch is other than the secondary table is another IPW based map (think normal "target map" in 0-255 values) and it will calculate IDC on the second set and record in the logs. I will try to get John Reed in on this conversation as he can probably give the details better than I.

03whitegsr

iTrader: (8)

Would it be possible to build a box that replaces the factory resistor box that also provided a peak-hold signal to the primaries?

Basically like two of those FJO controllers with a trigger on the secondary to turn it on and off as the code scales the pulse width at the switch over point. It seems like an easy way to tie into the factory wiring for the secondaries, and would also provide a drivability improvement by going over to a true peak and hold arrangement.

As long as the code eliminated the chance of halving the pulsewidth before the secondaries are turned on, I think you would be ok. A single engine cycle running ridiculously rich probably wouldn’t be noticeable, but a single cycle being very lean could potentially cause some damage.

jcsbanks

It looks like the injection cycle for any given cylinder begins at 5 degrees BTDC of the exhaust stroke, or 365 degrees BTDC of the combustion stroke and is fixed in the code. It would need an unfeasible rewrite to alter it.

The output to enable the secondary injectors would take effect immediately, whereas a change in injector size variable would take effect one engine revolution later to reduce the next calculated IPWs. Even if you did separate the changes by one engine revolution, a glitch would still occur since there would unlikely be a time when no injector was firing.

Also the accel enrichment settings would need to be altered at the switch since they are not relative to injector size variable.

The alternative of batching firing a second set isn't ideal either with rev limit and enrichment issues.

dan l

iTrader: (3)

To injector timing being fixed. Oh well.........

JohnBradley

iTrader: (30)

Well they have it for DSMlink (or will shortly)...come on guys if the 1G guys have it we dont have an excuse now lol

http://www.ecmtuning.com/wiki/v3firmwarechangelog

scroll down to dual bank injector

project_skyline

iTrader: (30)

This would awesome cause then we wouldn't have to worry about not being able to run 1600's when we could run 8 1000's or 8 1450's.

JohnBradley

iTrader: (30)

It looks like its setup to use the low octane maps and then has a something to control how fast it brings it in. I have noticed that is more or less like the AEM. It starts phasing the secondaries as the primaries achieve a preset limit (we have ours set at like 75% for instance). More stuff to think about anyway.