When you click on links to various merchants on this site and make a purchase, this can result in this site earning a commission. Affiliate programs and affiliations include, but are not limited to, the eBay Partner Network.
Hi guys, looking for a little assistance with a DSC (safety) related to fuel pressure being out of range; P0191 error code to be specific. I have an Elite 1500 setup on a 13b rotary. I believe I have everything configured correctly, but possibly this issue is due to my own misunderstanding of how this is supposed to work.
Manifold referenced base fuel pressure set set at 46.6psi in the main fuel settings and I have the following configuration for the fuel pressure sensor:
Problem is, I seem to be triggering the fuel pressure safety at unexpected times. My understanding is that it's based on the injector differential pressure going above/below the base fuel pressure by more than the deadband value above (7psi).
In the log below it looks like the first ignition cut might be due to the drop in fuel pressure, but the 2nd cut immediately after doesn't seem to relate to a fuel pressure also. As a side note, what's the best way to see when a safety/code is throw in the log views? Currently I just look for ignition cut as that's how I have the safety configured.
In this 2nd log I can't figure out the reason for the safety trigger at all as there's no dip in differential fuel pressure:
It's my understanding that injector diff pressure is fuel pressure minus manifold pressure (based on a post on the haltech forum by a haltech employee), but that doesn't seem to be the case in the below image; it's a couple psi off at the marked location:
Based on a reading on the haltech forum, it seems like a discrepancy between differential fuel pressure and base pressure from manifold pressure is to be expected because the fuel pressure regulator doesn't adjust fast enough to handle quick changed in map, but going above 7psi seems like quite a bit of allowance. What settings have others gone with? I didn't have any false triggering issue of the fuel pressure safety with my Adaptronic Modular ECU used just prior to this.
This should be fun. I had the same issue when I used to use haltech. Would trigger during a transient throttle event cause the Fuel pressure dipped momentarily.
No hysteresis on Fuel pressure protection = fail IMO. I should also mention 7psi is a substantial fuel pressure drop. (although it might be necessary to make the haltech one work, although in your car that is still not enough)
On the adaptronic I run 3-4psi .
Quick update on this, spoke with haltech and they admitted the issue with not having a hysteresis value in the configuration. They said they'll be adding one in a future software update, but for now I should increase the deadband value. Agree with, rx72c, this is an oversight on their behalf. I'm surprised they haven't already fixed this as it sounds like most user who want to run a reasonable deadband value would run into this issue.
This should be fun. I had the same issue when I used to use haltech. Would trigger during a transient throttle event cause the Fuel pressure dipped momentarily.
No hysteresis on Fuel pressure protection = fail IMO. I should also mention 7psi is a substantial fuel pressure drop. (although it might be necessary to make the haltech one work, although in your car that is still not enough)
On the adaptronic I run 3-4psi .
Let's see what all the haltech fan boys say.
They'd probably inform you of how you were using it wrong or not understanding it correctly. If you don't look at the data and just throw darts...your opinion isn't worth much.
If you look at each of the ignition cut events where engine protection tripped the P0191 level 2 cut for exceeding the dead-band, the events were triggered in VACUUM conditions just after throttle had been rapidly decreased. In the log except below, you can see that you've hit vacuum (-7.6 inHg) and the injector differential pressure spikes as the gap between vacuum and the fuel pressure exceeds the dead band AND the throttle position is still above the 25% minimum TPS which then trips the pressure. In both cases adequate fuel pressure has been present, AFRs on target, and the cut was tripped by vacuum driving the injector differential to exceed the dead band where it is 'technically' triggering from the pressure being too high, not low.
This is PRECISELY why you have the ability to modify the preconditions for the engine protection. In your case, I'd raise the min TPS up to say 35-50% but also change it to add the condition that min MAP be somewhere just north of positive pressure. This will prevent you from tripping the conditions when vacuum pulls the differential pressure beyond the dead-band when fuel pressure is otherwise in check and where its supposed to be. Using these settings, you should be able to run a tighter dead-band in boost and still keep from tripping from the vacuum pull. It won't cause a change to the function when in boost, which is where you need it to be functioning anyways.
Thinking further on this, having a hysteresis on fuel pressure pressure to cover short time events where fuel pressure exceeds the dead band seems like a bad approach in general compared to properly setting up the control band for TPS and MAP. Assuming the hysteresis would be applied to the complete range, you'd be subject to it in boost as well and in the event the fuel pressure drops outside of the range I want the resulting ignition cut to be IMMEDIATE preventing firing on the next rotor face where adequate fuel may not be present.
Part of the benefit of using an ignition cut for engine protection is the immediate effect of the cut, which you would forfeit with the presence of a hysteresis on the fuel pressure dead-band in boost. Things happen so fast that you truly have to protect at the electrical speed level. Look at what happened to the Dahm 3 rotor when those fuel pumps hit their internal relief pressure causing the fuel pressure at the rail to instantly **** and the engine to detonate all to hell, not that he was running any engine protection at the time. It only takes one boner'ed up engine combustion cycle to blow seals out of our car, which can happen on the order of milliseconds at 7000+ rpms which could very well happen within a hysteresis period that was meant to band-aid the non critical condition between vacuum and boost differential.
I dunno, I think a that an intelligent implementation of hysteresis as a control is a completely reasonable approach to take. Id think (and hope) that they would apply it with a multi axis table (which you touched on).
I also feel like while you're completely right in that you have proper configurability now with settings available you quickly can start to make a mountain out of a mole hill by throwing more conditionals which, in my opinion, may be adding more complexity for complexities sake.
I dunno, I think a that an intelligent implementation of hysteresis as a control is a completely reasonable approach to take. Id think (and hope) that they would apply it with a multi axis table (which you touched on).
I also feel like while you're completely right in that you have proper configurability now with settings available you quickly can start to make a mountain out of a mole hill by throwing more conditionals which, in my opinion, may be adding more complexity for complexities sake.
You could also set it to be map based only so the fuel pressure cut based on the dead band is only active in boost. That would cover you in any situation where you need the protection but with only the single condition. To me a hysteresis control system that requires a multi axis table sounds way more unnecessarily complex than simply instead to only be active during boost conditions.
You could also set it to be map based only so the fuel pressure cut based on the dead band is only active in boost. That would cover you in any situation where you need the protection but with only the single condition. To me a hysteresis control system that requires a multi axis table sounds way more unnecessarily complex than simply instead to only be active during boost conditions.
Skeese
Thats fair. I try to live life by the 80/20 rule while also trying to plan around eeking out that last 20% at a later date and I suppose I feel that (what I see in my mind as its implementation) visualized hysteresis control would be a lower barrier for entry as far as getting that first 80% of protection while also giving flexibility and the ability for refinement. I also could be on crack and it'd end up in allowing people to get into hot water with hidden variables too easily.
edit: I've been re-reading/thinking about your suggested approach and I think its a way more elegant and simple solution than mine, I wonder if my suggestion is now just adding complexity for complexities sake :P
They'd probably inform you of how you were using it wrong or not understanding it correctly. If you don't look at the data and just throw darts...your opinion isn't worth much.
If you look at each of the ignition cut events where engine protection tripped the P0191 level 2 cut for exceeding the dead-band, the events were triggered in VACUUM conditions just after throttle had been rapidly decreased. In the log except below, you can see that you've hit vacuum (-7.6 inHg) and the injector differential pressure spikes as the gap between vacuum and the fuel pressure exceeds the dead band AND the throttle position is still above the 25% minimum TPS which then trips the pressure. In both cases adequate fuel pressure has been present, AFRs on target, and the cut was tripped by vacuum driving the injector differential to exceed the dead band where it is 'technically' triggering from the pressure being too high, not low.
This is PRECISELY why you have the ability to modify the preconditions for the engine protection. In your case, I'd raise the min TPS up to say 35-50% but also change it to add the condition that min MAP be somewhere just north of positive pressure. This will prevent you from tripping the conditions when vacuum pulls the differential pressure beyond the dead-band when fuel pressure is otherwise in check and where its supposed to be. Using these settings, you should be able to run a tighter dead-band in boost and still keep from tripping from the vacuum pull. It won't cause a change to the function when in boost, which is where you need it to be functioning anyways.
Think smarter, not harder.
Skeese
So to understand your reply? Your saying to fix the issue we should band aid the problem by putting in a large dead band like 7 psi and increase the TPS threshold to cover a poorly designed function in the first place? Are you really that thick?
So to understand your reply? Your saying to fix the issue we should band aid the problem by putting in a large dead band like 7 psi and increase the TPS threshold to cover a poorly designed function in the first place? Are you really that thick?
No. Not at all what I implied there, not sure how you even came to that. Try again.
Skeese, Thanks for taking a look at the log and providing your analysis. It hadn't occurred to me that the safety would also trigger when exceeding the deadband, but in hindsight that makes perfect sense. I'll probably try using just the min map precondition first and see how tight I can get the deadband w/o false triggering. I tend to agree that adding a hysteresis is better to be avoided if possible. As you said, you don't want any delay in applying a safety in a true failure situation. My hope was that a small hysteresis would allow you to set a very tight deadband value that would very quickly catch a drop in pressure and not false trigger, but that might possibly still lead to engine damage if you had a very sudden pressure drop due to, say, a major fuel leak.
