[Gmecm] Example: Simulating the "power valve" in the ECU for PE

Tue Sep 13 17:45:36 UTC 2011

David, et al., just for reference/conversation, I?ll show the changes that I made to my stock 1998 L31 5.7L (Suburban) ECU?s ?Power Enrichment Enable % TPS Vs RPM? table.

In the table, the values were formerly 89.8 everywhere.   Based on data I logged from the ECU/engine while operating in a variety of scenarios, and subsequent post-examination of those logs, I modified the TPS entries at and above 1600RPM to enable PE at about 27?Hg MAP (I?m in Illinois so my MAP peaks around 30?Hg).   Then, after programming the ECU with this change, I validated that PE was indeed being effected at approximately the target MAP.

I of course made other, typical changes to the ECU?s parameters to enable PE without delay; I?m sure most of you know I mean.

Clearly these values could not be used as-is on another platform other than a stock L31.  I post these only for illustration.  Some might find these values to be much smaller (e.g., 43% TPS at 1600RPM) than they would have expected.

I can discuss the method of obtaining the TPS values I used in table, i.e., via post-examination of the logs, if someone?s interested; it wasn?t rocket science.

I?ll note that the ECU does not enter PE often, even with these new values (based on logging I?ve done while I?m driving, without a trailer or an atypical load on the vehicle), but now PE occurs at a much more relevant, and opportune, throttle position.

I?ve also made similar changes to my 1995 4.3L CPFI OBDI Blazer, with the same benefit.

David, any concerns you noted in your earlier posts on this topic regarding emissions, etc., are still valid, of course.

Jim

From: ext Command Performance Engineering [mailto:c.p.e at comcast.net] 
Sent: Monday, September 12, 2011 6:33 PM
To: Butler James-CJB006; gmecm at diy-efi.org
Subject: RE: [Gmecm] Simulating the "power valve" in the ECU for PE

Hi Jim,

Your words were well put.  I do agree with you.  As you well know, EFI is supposed to mimic carburation, but do it better in many ways, including a very important one to compensate for elevation changes.  Anyway, a point well taken is that GM never incorporated a Power Valve table into the caldron on tunables in the ECM.  To deal with transient throttle increases we have the pump shot tables to work with and there is the one table that does reference the MAP signal to provide a tunable enrichment control that is vaguely related to power valve functionality even though it is brief.  Fortunately for us, we have a wide degree of control over the amount of fuel enrichment during PE mode tuning that the carbureted guys will never have.

I guess that ever since I started working with the GM EFI long ago, I have learned to tune the PE Thresholds as they are, based on RPM not vacuum ? tuned together with the normal PE Fuel Enrichment table.  Of course, it is important to tune all of your Open Loop and Closed Loop scenarios first before tuning the PE portion, due to the fact that when the engine crosses over the threshold into PE mode and not at 100% WOT, the calibration program always does a lookup to reference the C/L VE Fuel Control table(s) to use as a way to factor an appropriate amount of fuel to deliver for any given RPM, referenced to MAP in the VE table(s) together with the PE table for that same RPM.

Hopefully I explained things better this time.  Thanks for your comments.

David Johnson

From: Butler James-CJB006 [mailto:jim.butler at nsn.com] 
Sent: Monday, September 12, 2011 4:09 PM
To: c.p.e at comcast.net; gmecm at diy-efi.org
Subject: RE: [Gmecm] Simulating the "power valve" in the ECU for PE

By the way, I?m not ?pontificating?, below, but rather I?m testing my own understanding and challenging others to ?set me straight? where I run amiss!

Jim

From: Butler James-CJB006 
Sent: Monday, September 12, 2011 4:51 PM
To: c.p.e at comcast.net; gmecm at diy-efi.org
Cc: Butler James-CJB006
Subject: RE: [Gmecm] Simulating the "power valve" in the ECU for PE

David, I?m looking at this now.

I?m a tad confused.   I?ll set the baseline of my understanding, below.

Transient fueling needs to mitigate temporary lean conditions which occur during changes in throttle position, e.g., at least,

?         to accommodate wet-flow fuel transit latency (as might be more prevalent in a TBI setup) relative to throttle tip-in, or 

?         to fuel any brief, non-steady-state inrush of air into the manifold volume 

are, as I understand them, accommodated in the carburetor by the accelerator pump.   I?ve assumed transient fueling is accommodated in the GM OBDI FI systems through the ?Pump shot? tables (if, of course, we believe that those tables are properly labeled according to their function!)

Power enrichment, e.g., to achieve the maximum torque output from the engine by increasing the AFR, is accommodated in the carburetor by a different mechanism which increases the AFR in the steady-state air flow.   Holley, for one, uses the ?power valve?; another, Edelbrock (with their Carter ?AFB? design), uses stepped metering rods and a set of pistons which respond to the pressure differential between the atmosphere and the manifold (aka ?vacuum?).  In FI systems I?ve assumed this is accommodated by the various ?PE? tables.

So ?pump shot? applies to transient events and ?power enrichment? applies to stead-steady events.  

Are we aligned on this?  A simple yes/no will suffice.  I?ll follow with more comments, later.

<thanks!>

Jim

From: ext Command Performance Engineering [mailto:c.p.e at comcast.net] 
Sent: Monday, September 12, 2011 9:48 AM
To: Butler James-CJB006; gmecm at diy-efi.org
Subject: RE: [Gmecm] Simulating the "power valve" in the ECU for PE

I?ve tuned more of these ECM?s than I can count in the past 15+ years.  While I do bump the TPS Thresholds down somewhat to enable PE as various RPM levels, there are two tables in these computers that deal with Acceleration Enrichment (AE), simulating the accelerator pump and the power valve in carbureted engine configurations.

The accelerator pump function is found in the Pump Shot Vs. TPS table, while the power valve function is called the Pump Shot Vs. MAP table.  Both tables work hand-in-hand and deal with one area of AE more than another, but they do overlap somewhat and contribute to the whole picture of what is needed when you are accelerating.  These tables add a measured amount of extra fuel (by uSec).

The Pump Shot Vs. TPS table should be tuned while accelerating in first gear while you are pushing into the gas pedal, usually at a moderate rate.  Watch your o2 voltage reading along with your TPS position percentage.  If the o2 sensor voltage plummets at one or more point(s) in the acceleration phase, just bump up the pump shot in that area.  You are looking to keep the o2 sensor reading in the 600 ? 800 mV range.  Less than this will be felt as a hesitation or bog, while more than 800 mV will cause the exhaust to go too rich resulting in failed emissions.  When I adjust the table values up or down to get the right feel in the acceleration (while watching the o2 readings) it is by small amounts, since this is a trial and error effort.   Make too much of a change and you end up chasing your tail, leaning it out again

The Pump Shot Vs. MAP table (power valve) should be tuned similar to what I stated above, but while accelerating in usually a high (4th) gear, say 40 MPH and higher.  There are tuning breakpoints at 0, 20, 40, 60, & 80 kPa.  It is quite common to be cruising in the 40 ? 50 kPa range.  I rarely need to adjust the 0 ? 20 kPa areas, sometimes a little at 40 kPa (like 14 -28%), but major improvements can be made in the 60 and 80 kPa areas.  The last calibration I made last week for an ?87 Chevy ? ton 4x4 Truck with a HT383 engine and a stock TBI running 16 PSI of fuel pressure (to get 20% more fuel) ? I needed to increase the 40 kPa point by multiplying the stock value by 1.28, the 60 kPa point multiplying the stock value by 2.299, and the 80 kPa point by multiplying the stock value by 3.335.  That may seem like a lot, but in reality it amounts to only an additional 793 uSec of pump shot to the existing Base Pulse Width (BPW) at the 60 kPa level and an additional 1709 uSec (or 1.7 mS) of pump shot to the existing BPW at the 80 kPa level.  Pump shots do not last very long and are indeed considered momentary to cover transitions to increasing load levels, just as those components did on a carburetor.  I know that a power valve could actually stay open longer in high load situations, but the intent of a 4-barrel was to run the primaries lean and the secondaries rich.  So the secondaries were the main thing that you would tune for PE at WOT, coordinated with the right power valve vacuum value.

This kind of detailed tuning will make huge improvements in drivability in Closed Loop mode.  They will also help you tune the WOT (Wide Open Throttle) phase better, because the accelerator pump and power valve functionality are now working properly for you.  Tune the PE phase mainly at WOT (100% TPS) and watch your o2 readings to generally run in the 890 ? 900 mV range for TBI engines.  885 mV = 12.5:1 AFR.  Port EFI engines can run well at WOT in the 870 ? 880 mV range, closer to 13.0:1 AFR, depending on the severity of the load dynamics and non-boosted.

Once you have tuned the WOT phase, you can play around with the threshold levels at the various RPMs to obtain the desired transition from C/L operation into PE.  Just remember, if you adjust the threshold values down too far you will screw up your fuel economy and emissions.  It is normal to have higher threshold values at lower RPM?s and lower values in the upper RPM range.

I hope this information helps.

David Johnson

From: gmecm-bounces at diy-efi.org [mailto:gmecm-bounces at diy-efi.org] On Behalf Of Aaron
Sent: Monday, September 12, 2011 6:44 AM
To: 'Butler James-CJB006'; gmecm at diy-efi.org
Subject: Re: [Gmecm] Simulating the "power valve" in the ECU for PE

Interesting concept.

I?ve never like the power valves in the Holley?s for the same reason. 

Cam choice, throttle body size, boost, driving style, all play havoc with vacuum.

The TPS vs. RPM table always seemed a better choice for increasing the fuel outside the normal driving.

So when I watch the live data on the TPS vs. RPM table, it?s easy to recognize the area or zone were you spend most of your time. Idling thru the city being in the lower left, cruising down the road freeway in the middle and rocketing down that unused road being in the upper right. That would give me the upper left for jack rabbit starts, the upper middle for passing the left lane lounger freeway and the extreme right for ?let see what breaks first?. This also gives me a chance to take out fuel in the lower right and middle.

For the power enrichment, I?ve always saved for towing, driving in hilly country and getting up the boat ramp.

I should also mention, I am by NO means an expert. 

From: gmecm-bounces at diy-efi.org [mailto:gmecm-bounces at diy-efi.org] On Behalf Of Butler James-CJB006
Sent: Thursday, September 08, 2011 6:29 PM
To: gmecm at diy-efi.org
Subject: [Gmecm] Simulating the "power valve" in the ECU for PE

Some time ago I begin experimenting with the power enrichment function on my 1995 S10 to simulate the action of the ?power valve?.

I?m wondering if any others have tried the approach I?m about to describe.  It?s not something I?ve seen anyone write about, e.g., on Monodax.

Most (all?) carburetors enable PE when the manifold vacuum drops below a threshold (at least that how it works on my Edelbrock and the Holley ?power valve? appears to serve a similar function).  However, every GM ECU that I?ve seen (which is only about a dozen) enables PE based upon, at least, a table of TPS and RPM, and not MAP.  I assume other vendors? ECUs are similar.  I?m sure most of you know all of this.

I can see the rationale for not using MAP as a PE threshold in modern systems, and I?ll comment on that later.

In my case, I always liked the carburetor?s way of handling PE, i.e., when the engine approaches maximum load (e.g., 6? of vacuum or whatever one cares to define as ?maximum?), enter PE.

What I?ve done, in my S10 and some other vehicles since, is adjust the PE ?TPS threshold vs. RPM table? to enable PE at some target MAP, e.g., 26? or 28? of manifold pressure.  Of course, I had to have some idea of which TPS values to use for each RPM band in the table, so I simply logged a bunch of data from the ECU and sorted it to get an estimate of the throttle position/TPS values that would yield, approximately, a particular MAP for each RPM range in the table.

Upon applying these new TPS values to update the table, the drivability ?fun factor? was notably improved, which is why I?ve applied these changes to a number of my vehicles.  Configuring them for around 27? of MAP is usually to my liking.  I always validate the changes by logging across the RPM band and confirming that PE is entered at the target pressure.

In the real world a change like can cause at least one problem, i.e., when towing, or driving in hilly country, PE can be entered much more easily, and for long periods of time, with the typical consequences.  Apart from that downside, I find it?s a nice tweak.

Jim

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.diy-efi.org/pipermail/gmecm/attachments/20110913/0f9bc1c8/attachment-0001.html 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: image/png
Size: 2689 bytes
Desc: image001.png
Url : http://lists.diy-efi.org/pipermail/gmecm/attachments/20110913/0f9bc1c8/attachment-0001.png 