One of the common problems on these engines, especially ones that are ported, is a stumble as the throttle opens. The Power FC has two main tables for the "accelerator pump" function (like on a carb) which is supposed to prevent this. "Accelerate injector" is a max amount of injector pulsewidth that can be added at a given rpm, and presumably it interpolates between the rpm breakpoints.

At one point I doubted this, but I now believe that the PFC delivers one major shot of fuel as the throttle opens based on this table. It is most likely "asynchronous," where the primary injectors fire at the same time and the normal injector firing order is ignored. Normally the firing order would go F primary, R primary, essentially synchronized with the top dead center signal ("G" signal) from the crank angle sensor. The tip-in fuel ignores the crank angle sensor. That's pretty typical on sequential fuel injection. Just about every modern port injected gas engine does it that way.

In the PFC the max allowable pulsewidth at a given rpm is mostly determined by the "Amount" setting under accelerate injector. From what I've seen the "decay" setting has to do with how the fuel tapers off after that initial shot. The setting isn't very important and is best left alone. The Inj vs Accel TPS1 table determines what % of this max value (Accelerate injector "amount") will actually be delivered.



The 2.128 update of FC-Edit redefined this table and it makes more sense now. I just did some tinkering with it:



On the left side are the values displayed in the new format, where the left is in V/T (probably voltage change over time) and the right appears to be a multiple of the "Amount" from the Accelerate Injector table. On the bottom are the default values, on the top are my values for a large street port series 4 FC3S.

The throttle response had been smooth under moderate throttle movement, but a hard stab would still cause sputtering, as would a very slight movement of the pedal. Now that I understood the table better I decided to change the top row completely. I increased the amount under "V/T" so that this row would be controlling tip-in fuel under very hard throttle. Then I increased the amount from 1.00 (which probably means 100% of the "Amount" in Accelerate injector") to 1.10 . That tells the PFC to spray more fuel, but only under very hard throttle.

It's a more precise way of fixing the problem rather than just jacking up the "Accelerate injector" table. That can cause an excessively rich mixture sometimes, especially when you are starting from a dead stop. I also lowered the "V/T" in the bottom row so that I could have a little bit of fuel kick in earlier. Now when I drop the hammer it never stumbles. A very tiny throttle movement (a transition from coasting down a hill) may cause a little sputter still but that is improving, plus I do have a big streetport so it may never be perfect.

So in conclusion, with this new update you can use the Inj vs Accel TPS1 table to more carefully target your tip-in fuel for specific amounts of throttle movement. One last thing: I made a thread a while back about using "INJ vs TPS" in the Datalogit to improve tip-in response. I have mostly zero'd this table out now.

 

I was going to try your last solution this weekend. I'll give this one a spin and see if it improves my less-than-smooth tip in response.

 

Thanks for the info, I've been fighting tip-in stubles for a long time now and would love to remedy this.

I just looked at the maxes allowed in the inj vs accel TPS1 table and I'm all confused again. You have 0-255 allowed for the input side, which makes sense since it's based on change/time. But the setting max goes to 65535. At first I was thinking 255 was 100% of the amount in accelerate injector table.....but that doesn't make sense.

Any ideas guys?

 

Update to the new 2.128 version. The old numbering is confusing. PM me if you need it.

 

It appears the left column is volts / millisecond which is a rate of throttle change so I would assume that between the three rows (three different rate changes) the PFC will interpolate the change in the three rates(smooth the transition in throttle rate change). In your first entry it appears that 0.752 vs 0.503 will tollerate a faster rate of throttle movement (0.752 V/ms) for the same listed fuel enrichment (1.10). If you left 0.503 V/ms and enrichment of 1.10 the system would add the same but at a slower throttle rate change - right?

Also I am kinda unclear as to the basis of enrichment. Is it percent delta to the current injector duty?

I wish these guys writing the screens would get a little more technically correct on the units but it seems much better than before.

 

Some DL screen explanations are hazy because the DL people do not know exactly what some of the functions due. After all it was reversed engineered.

 

Very good point yet things like "V/t" and "function" is still unclear especially if the change was due to additional knowledge of the actual function like V/ms or V/ms^2 and %DeltaInjDuty

 

I think you understand the basic idea of the left column. Unfortunately this is a very low-resolution table. I increased the number on the right initially (from 1.00 to 1.10) and then increased the number on the left so that it would take a faster throttle movement to actually hit that number. It's a workaround for a low resolution table and I'm still experimenting with all the tip-in stuff. I think idle and tip-in are the most frustrating and limiting things about tuning driveability with this computer.

My current working hypothesis is that the number on the right column of Accel vs TPS1 is a multiplier of the amount values from the Accelerate injector table:



The amount value is the max allowable pulsewidth for an additional asynchronous spray of fuel during tip-in. By asynchronous I mean both primary injectors fire at the same time. This is outside of the normal firing order which is usually synchronized with the "G" signal from the crank angle sensor. On the FC and FD the "G" signal indicates top dead center and functions a lot like the TDC and cam position signals on a piston engine.

Asynchronous tip-in fuel is used on a lot (all?) of cars with sequential fuel injection, including the factory series 5 ECU:



So here is a log from a PFC of tip-in being applied:



My analysis: As the throttle opens quickly (top red line), the PFC calculates the total amount of fuel (blue red line) it could inject based on rpm (Accelerate Injector table). Then it uses the INJ vs Accel TPS1 table to figure out what % of that maximum fuel to use based on the rate of change of TPS voltage. Finally, the "decay" value is used to taper off the fuel after the initial shot.

 

Hello Arghx, have you played with this function any more recently? I finally have found some time to play with my tune and I am going to start out with your settings.. I am going pretty lean on intial tip in so your number seem as good a place as any to start

Thanks for the information!

 

I've got a streetported GT40 FD I've been tuning recently. I have used the Inj vs Accel TPS1 table and it has helped (values similar to what I have posted above). There is still a little bit of stumble on really slight throttle sometimes, and on really hard WOT stabs. I am unsure how much more it can be improved on this car. Adding to Accelerate Injector (the rpm based table, accessible in the Commander) didn't do much for that slight WOT stumble.

For that slight low throttle I may try the INJ vs TPS table next https://www.rx7club.com/power-fc-forum-47/success-settings-2-inj-vs-tps-tip-improvement-799079/ . I haven't gotten a chance to mess with the car in a while. The shop that is contracting out to me is fixing a boost leak on it. I do have access to a loading dyno so that gives me some flexibility to experiment.

Just to clarify--this particular car doesn't exactly run poorly. It doesn't have horrible bucking or anything. It just doesn't drive 100% stock in terms of throttle response. With a ported engine and aftermarket ECU it may not be possible to achieve that, but I am still trying to make marginal improvements.

 

You sir have a PM about newer program. Thanks for this tip!

 

Latest combination of settings I have been using. This is on a street ported FD with a GT4088R and 550/1680

 

Thanks for the update Arghx... I started with your original numbers and I was still going a little lean on tip in with mine, so I made a few small tweaks... still lean.

So for fun I hit it pretty hard to get it deliberatly go rich on tip in to get an idea of how far to go... I will now judge where the middle ground is ... If that doesn't work out for me I may try your new numbers.

I added a little on the low end of the inj vs accel tps as I am having trouble mostly with slower tip in pedal crowding.. I also lowered the cut point a bit. but it may be a lost cause with 850 primarys?

I also added a little to the accel inj table and left the pedal position table alone for now...

thanks again for all of your useful information...

 

In my experience at least, the accelerate injector table is pretty useless for really slight throttle movements. You can crank up the values but it will still stumble lean and then go excessively rich. If you get stuck, start with the numbers I just posted (use all of them from all three tables, they are meant to work together) and modify from there. There is a chance it may then require some adjustment of the INJ/base maps.

 

That is proably what will happen, I will abandon the accelerate inj table, it does seem to behave the way you mentioned. I don't get a chance to drive my car as much as I would like (I often have to drive test vehicles from work). This is as much as anything else an exercise for me to learn how the PFC logic works as anything else... I sure would like access to their strategy book, woud make things a lot more straight forward if we knew exactlly how all of the numbers are used... Like I have seen you say before unfortunately this ECU is kinda dumb... but at least we have the datalogit...

 

I made the adjustments suggested in post 12, and it seemed to help things
Thanks arghx

 

Bump for good info!

I also have this problem with a S5 streetported TII, tip is is quite ok, but as most seem to be having, the transition from overrun to cruise causes a lean spot if the throttle is pressed between really slow and really fast.

a possible reason for this is that the absolute voltage change from the tps is too small and gets partially filtered out by the ADC signal conditioning in the ecu so the resulting V/t value is smaller then the actual throttle acceleration. a strategy that might combat this is to fill the bottom row of the Inj vs Accel TPS1 with a really low V/t value and a quite large Setting value. this would normally cause a rich spot. but the clue is maybe if the decay setting is altered so the extra pulsewidth is only there for a very small time, it will provide just enough fuel to avoid the stumble, but not for such a long time you get an excessive afr undershoot.

this is just food for tought, I will try it out in 3 weeks.

Adjusting the Inj vs TPS changes the fuel everywhere, so this will require adjusting the base or INJ map to keep steady state AFR unchanged. what this does is create a 4D fuel map (rpm/pim/TPS1) instead of the 3D (rpm/pim), so speed density mixed with alpha-n

 

If you use the settings listed above, before you tune your main INJ/Base map, generally speaking there will be no issues. I have tried a lot of different tricks with changing the Inj vs Accel TPS1. There are limits to what you can do with such a low resolution table. You're welcome to experiment. You'd be hard-pressed to find much improvement over the settings shown above, especially given the tedious and time consuming nature of such tuning.

Part of the issue is that there are no settings to control tip-in fuel injection timing. That is, there's no way to apportion the tip-in fuel between synchronous (with the normal sequential firing order) and asychronous (extra injection pulse). A Haltech for example can do that.

 

Even with synchronous accelerator pump injection it is still very hard to get it perfect (stumble and stink free) with the typical stand alone ECU resolution and testing/tuning process.

I have managed to get it smooth with synchronous injection and only 4 shot amounts for varying rates of throttle rate of change. What I found was smooth in all cases = excessive enrichment in others, and excessive enrichment in others = un-burnt fuel = stink and smoke.

Just think about what you are capable of vs. 3 experienced automotive electronics engineers with a budget to map the ECU of $100 000 and the dyno and gas analyser and the training to use all of the equipment.

 

Great post, thanks for sharing this! I know I'm dragging this back from the dead but I've been focused on tuning for driveability lately and this info was just the ticket for my ported motor's tip-in problems.

 

Glad to hear the settings are working for you

 

This is a great write up(thread revival sorry) it's nice to have a explanation for what settings actually mean as a lot of us have learnt through adjusting settings then monitoring change.

Just my 2 cent(pence) but I found that if you have a stockish motor with a big TB it cN be easier to adjust the inj v rpm amount settings in the 1k up to 3k brackets to achieve a good throttle response as this won't affect the settings higher up the rev range,obviously with a larger throttle body the big change is at low rpm so that's where I've found you have to add quite a bit.
I've spent a lot of time on a customers car with a large TB but I don't believe you can completely iron out the light flatspots(compared to stock) as the air flow almost stops so won't carry the fuel.

Hope I haven't runied your thread

 

another bump because i have read this thread 10x over the last 2 days