-
I think it's more like the factory ECU is programmed to supply fuel enrichment on tip in. The stock program senses how far and how fast one is pressing the accelerator (along with other operating conditions) and supplies the necessary enrichment for N/A airflow. When boost comes on suddenly, the enrichment isn't quick enough. The airflow increases much faster than the ECU expects. Running higher pressure throughout the N/A range gives the system a head start. Closed loop operation compensates at idle and cruise, but when you hit the throttle, the programmed amount of enrichment is augmented by the higher N/A fuel pressure.
[This message has been edited by rjaffey (edited January 15, 2000).]
-
JR Civic,
That is one of the best run downs I've seen on fuel pressure. Everyone appreciates researched knowledge, it helps everyone out - and for every five people replying, there are probably fifty reading and using what you've said to make their kits work better.
The JR regulator produces a 1:1 ratio with lb/in to inHg (are you sure it?s not 1:1 lb/in to lb/in)? That would mean fuel pressure is about 5 lb/in higher at 20 inHg than factory ? about 7% rich. It is just a strange way to deal with throttle enrichment at throttle tip in, because it will have its highest effect at the highest vacuum, not at 0 vacuum - where throttle tip appears. If this is indeed producing an over rich condition at high vacuum and idle ? yes the ECU can easily take this out with closed loop.
I agree with rjaffey regarding the sudden rush of air from the SC ? it seems to be related to the opening of the bypass valve itself. This does not occur on turbo cars that I have seen. I would also add another factor that can produce leaning at ?throttle tip in or boost tip in?? the factory ECU (being NA) stays in closed loop until nearly WOT, and since boost can be produced out of WOT, this is causing a 14.7:1 boost run at anything under 90% throttle.
-
When you back out the adjustment screw you can lower the baseline fuel pressure (no vacuum, no boost). If you drop it much there will be a moment of hesitation on tip in. On my car that was about 55psi, on another one running 8psi it was a little higher. Once the transition had passed the cars took off, but there was a moment of hesitation during the rapid transition from a steady state to a load state.
The motor is effectively seeing boost the moment the bypass closes even before there's positive pressure in the manifold. This condition demands higher fuel delivery even while there's still some vaccuum. So running the elevated N/A pressure is a means to handle the effect of boost between the point where the bypass closes and the point where the JR regulator starts its 5:1 rise in response to *positive* pressure.
As to the parasitic drag, if what's posted on Eaton's site is remotely accurate the HP requirement at 5500 blower RPM (that's about 4000 engine RPM with 4" blower pulley and around 6psi of boost pressure) is between 6 and 7 HP. It peaks at perhaps 17 or 18 HP at redline and I just see that as additional load, like there's an extra few hundred pounds in the car. If the fuel pressure's too low, the hesitation is there even when changing from steady to load state while the car's in gear (cruise to accelerate). How is this different from the drag created by the blower?
I can't comment on a turbo setup other than to say that I don't believe the transition is anywhere remotely as abrupt as the onset of boost with the roots blower. With the blower the boost is virtually instantaneous in direct response to the throttle. Changing the throttle position results in more rapid, greater changes in airflow than the stock ECU is programmed to expect N/A, even when some vaccuum is present.
My air fuel gauge hunts back and forth when the ECU is operating in closed loop mode. What I believe I'm seeing is the effect of the feedback and response between the ECU and the O2 sensor. The ECU ups the fuel until it tips on the rich side, then drops the fuel until it's lean, then back, and so on... All the while hovering around stoich where the O2 sensor is fairly accurate. Once I'm on the throttle, the gauge reads pretty consistantly in the 12.5:1 range. I'd say it's within +/- .5 in that area of the O2 sensor's operating range and accounting for the granularity of the A/F display itself.
When I suddenly mash the throttle, transitioning from cruise to accelerate, the display immediately jumps to the 12.5:1 range without going lean and there's no hesitation. As I drop the baseline fuel pressure the A/F display starts to indicate a very momentary lean condition and if I keep dropping the pressure the momentary lean reading and hesitation becomes more pronounced.
-
Apparently, I have some wheels turning here... many of you have made some suggestions as to what can be causing the temporary "Lean" conditions present at throttle tip-in. I believe that all of you are right, although I have no concrete data to support these theories... now, I disagree as to the extent of these effects. The ECU is designed for a NORMALLY ASPIRATED engine, and thus is not capable of making the fast, drastic fuel changes necessary to compensate completely for a blower. The biggest problem is parasitic drag, or "LOAD" on the motor. Turbocharged engines do not have tip-in problems for this reason. The supercharger, even when NOT in boost, creates drag on the motor. It requires about 25 hp just to spin it at about 4000 rpm... that is like having three (3) Air Conditioning compressors running at once. The ECU cannot account for this load requirement, and thus does not deliver enough fuel enrichment until you get to 1psi boost (four with the older enrichment switch) and then the fuel catches up.
As for the Air/Fuel gauge, although they are helpful in giving you a general picture of where you are with your fuel delivery, they are inherently inaccurate and cannot deliver proper information quickly. This is why you see the lights bouncing all over the place. Being lean for a split second at throttle tip-in is not particularly dangerous... you aren't lean for any longer of a period than say the "lean time" while cruising when the ECU bounces from lean to rich to lean to rich !!! Don't sweat this. Here is where you worry: Watch you A/F gauge as you accelerate at WOT in fourth or fifth gear from 6500 rpm to redline... how far down does the ratio drop... and this took place over MANY seconds... and THAT can be damaging.
------------------
'93 Civic Si H/B with B18C1, JRSC @ 8psi
Best 1320: 12.88 at 110 mph
-
If the problem is that even with the new JR boost switch that fuel enrichment doesn't start until 1 psi. Is there any vacume triggered solenoid made by any manufacturer that's calibrated to the same vacume the JR bypass begins to operate at. If so it could be used to send the enrichment signal like the relay but at the proper time before 1 psi?
-
I don't know how true this is, but I heard that at tip in the ECU retards the timing just for a split second during that period where the engine runs lean. Not sure though.
-
I have installed a Hypertech Shogun in MrBurns' car, and during my tuning of said car, I saw NO evidence that the ECU momentarily retarded the timing at throttle tip-in. In addition, if there was some retard thrown in, the detonation/pinging would NOT occur.
Detonation is occuring in the midrange on JRSC cars because of load... nothing more, nothing less !!! Remember that all operating systems on any given car, given the same input, DO the same thing. This being the case, if the ECU was the problem, then it would be the problem at 2000 rpm... at 6500 rpm... etc. The reason that this problem only exists between 3000 and 5000 rpm is simple: the load or parasitic drag that the blower generates is too high in relation to the amount of HP created... and thus, there is not enough fuel enrichment to compensate for this extra load.
Cure = more fuel+less timing+larger injectors.
Timing will be my next post... stay tuned.
------------------
'93 Civic Si H/B with B18C1, JRSC @ 8psi
Best 1320: 12.88 at 110 mph
-
> rjaffey: My air fuel gauge hunts back and forth when the ECU is operating in closed loop mode. What I believe I'm seeing is the effect of the feedback and response between the ECU and the O2 sensor. The ECU ups the fuel until it tips on the rich side, then drops the fuel until it's lean, then back, and so on... All the while hovering around stoich where the O2 sensor is fairly accurate.
Yes, in fact the very definition of a "closed loop system": Input, processing, output, feedback
See http://me.mit.edu/lectures/rlocus/2.5-rationale.html for a diagram
> rjaffey: When I suddenly mash the throttle, transitioning from cruise to accelerate...
> rjaffey: I think it's more like the factory ECU is programmed to supply fuel enrichment on tip in....
"Tip in" is a term I've heard a lot of recently. I feel the term doesn't fully explain the problem. There are several things going on which are easily confused by using a single term.
1. Throttle has a transition speed, and acceleration. (this is 'acceleration pump' fueling as described below')
2. bypass valve is closing. (this is a specific fueling issue to the SC kit)
3. Fueling is trying to catch up in hardware (rising rate regulator - all kits but FMAX)
4. ECU may still be in closed loop (see my previous post above - problem with 'slow tip in')
Adding to confusion, And the term 'tip in' seems to be used for many dynamic situations:
1. Mash the throttle from cruise to accel (open loop boosting)
If you have detonations in open loop then it is a fault of the timing and fuel curves. These issues will not likely easily be solved with a mechanical solution, especially at boost pressures over 6psi, where the associated fuel rail pressure will be high enough to blow factory injectors, or be pushed past any actual fueling return at any pressure.
2. Slow throttle transition from cruise to accel (closed loop boosting)
The condition of a cruising, slow throttle transition is where the JR/Moss SC seem to really show its bad side. Take a look at a J&S knock installed in such a car, and my previous notes on the ECU being in closed loop until nearly at WOT, and you will see the difference. I don't care what kind of piggy back or regulator you are using here. There is no way to do this correctly without reprogramming the ECU's closed loop - or going with an aftermarket 3D EFI.
3. SC specific problems including the bypass
JR/Moss's location of the MAP sensor itself may be contributing to the problem. It may be seeing a quick blast of erroneous data to the MAP sensor, either hi or low voltage due to air pressure and turbulence changes as the closing of the bypass occurs. Locating the MAP to the boosted side of the SC seems to help the problem.
Another trick we can do with the SECU (which for those that don't know is a factory ECU with boost recognition and 3D real-time adjustability - but this technique can translate to most 3D EFIs) is to increase fueling in the 1" to 0" vacuum zone - this helps fatten up fueling in an effort to 'predict' boost.
------
From my post to Endyn BBS...
Regarding: ?stock program's less than desirable transient response when used with the supercharger.? ... There are two issues. The first is that if you are using a typical rising rate regulator, you should note that it doesn?t perform any likeness of a ?rising rate?, it works like an on/off switch and the so called rate is only specifying the on time ramp. Anyone with an FMU and a dash mounted fuel pressure gauge call tell you that. The issue is that the fuel pressure is only increasing at 1:1 until about 2-3 lbs. This gives a lean situation right as boost crosses 0BAR. A typical and common problem from FI kit owners I have heard over and over, especially once their TA Sensor trick has been ?learned out ? by the ECU. Obviously the SECU can apply more fuel pulse width anywhere in this ?danger zone?. The second issue is the ECU?s accelerator pump feature. ...familiar with carbs...? The ECU watches the TPS delta speed, and if fast enough quickly sends a fuel pulse blast to all cylinders. Again something to be tweaked for ?transient response?.
------
> JRCivic: "The ECU is designed for a NORMALLY ASPIRATED engine, and thus is not capable of making the fast, drastic fuel changes necessary to compensate completely for a blower."
The factory ECU is very fast, and certainly CAN compensate completely. The issue is defining what the problem is exactly and then configuring it to do so (reprogramming).
> JRCivic: The biggest problem is parasitic drag, or "LOAD" on the motor.
Magnuson recently invited ZDYNE over for a tour of their facilities in Ventura, CA. For those that don't know, they distribute, assemble, and repair the Eatons for JR/Moss, and many factory 'Detroit' vehicles. We discussed the drag issue, and according to Technical Director Chris Weisberg, the typical drag of the chargers while the bypass is open is less than 1hp. I do not think the "parasitic drag" of the SC contributes to 'tip in' issues - as far as non-boost running.
> JRCivic: It requires about 25 hp just to spin it at about 4000 rpm...
Note, any additional "load" on the engine due to the SC drag, could be compensated by figuring in by the ECU, as you hinted with the AC comment, it already compensates for the load of the A/C and alternator(ELD).
> JRCivic: As for the Air/Fuel gauge, although they are helpful in giving you a general picture of where you are with your fuel delivery, they are inherently inaccurate and cannot deliver proper information quickly. This is why you see the lights bouncing all over the place.
The factory heated O2 sensor although very temperature dependent anywhere but stoich (see http://www.zdyne.com/images/O2curve.jpg) , is very fast. It has a very steep voltage response on either side of stoich, a bit like trying to keep a marble on a mountain sharp as a pencil. Its response is quite fast, it only takes a few milliseconds for a stable reading (unlike a gas analyzer that could take up to 30seconds). The lights move quickly because the sensor is fast, this is especially noticed with ECU's reaction is fast in closed loop. We have a wide band Bosch LSU sensor, its response is under 5milliseconds - so we can do a 'dynojet' type dyno run and get accurate AFRs.
-
WOW!
This board is rocking!
-
It's taken forever to get the people "in the know" to speak up...now we can start to learn the "whys" and start working on the "hows"!
JeffM, how damn long are we going to have to wait for a "plug and play" ECU for our blowers? I'd like to lose all the fuel pressure(easier on the pump) and use bigger injectors and a properly programed ECU.
[This message has been edited by Zeus (edited January 17, 2000).]
