Spittybug #2329 Peugeot 604 manifold custom conversion [Archive] - DMCTalk Forum

Spittybug

12-08-2011, 05:26 PM

I finally got off my ass and started optimizing my tables. They are pretty good I think, but the ignition advance is a bit of a black art I think. I did some more internet searching and I think I may have initially configured things a bit off. Here are my new understandings:

Static advance - how much a stock distributor is physically rotated to advance the spark. 13* BTDC. This is also the idle value to use since the idle micro switch prevents the vacuum advance from seing any vacuum at idle.
Vacuum advance - Here is where I think I was off a bit. Our book says 5"Hg is 3*, 10"Hg is 10* and 15"Hg is 20*. This translates into a range between 50kPa and 83kPa on the Megasquirt tables. So immediately after idle, the vacuum advance is at its maximum of 20* until MAP hits 50kPa, then it decays down to 3* by the time MAP hits 83kPa. I had it ramping up, then down, but not giving up that much.
Mechanical advance - Book says 1000 RPM no advance, 2000 RPM gets 10*, 3000 RPM gets 14* and 4000 RPM gets 20*. I've had some car savvy friends tell me that I can be more aggressive on this and get it "all in" by 2000, gaining performance. Any opinion on that?
Maximum advance - Not listed in our book anywhere that I can see, but again using some online calculator tools I think we max out at about 37* based on our valves, bore and era of the engine. Too much to potentially lose with too much advance, so if I'm leaving some performance on the table being conservative, so be it!

So, in the "idle zone", 13* looks to be the number.

As you move "up" the kPa scale on the table (Y axis), it starts at a high value (13*static + 20*vac) right after idle and decreases to only 13*static by the time you get to 83kPa. Of course by then you've increased your RPMs and you never really hit these bins on the table.

As your RPMs are increasing at WOT, the base 13*static gets mechanical advance added to it up to a max of (13*static + 20*mech). As you ease off the gas and the load goes down (decreasing kPA), the vacuum advance starts coming back in (13*static + 20*mech + upto 20*vac) all subject to the maximum of about 37*.

At cruise, or light throttle with your kPA sitting mid-low and RPMs at ~2000, the advance is (13*static + 20*vac + 10*mech) subject to maximum of about 37*.

The biggest change this will mean for me is in the shape of the vacuum decay. For some reason I have 17* in my idle area too that I will change. Nice thing is I can do this with just a few keystrokes and then burn it to the ECU no problem. If I make a mistake the roll back feature is nice. After the new timing table is in I'll run around with the VE tuner running and let it recalculate VE levels to get the target AFR of 14.1 I have set across the board. I still may tweak that table to go lean in cruise or decel and go rich on acceleration, but I want the timing dialed in first.

If this ends up being an improvement I'll post the final table.

For the non EFI folks reading this, don't let it deter you if you are on the fence! It's actually fascinating to see how much control you actually have over the operation of your car and how much performance you can actually get. Once over the hurdle of fixing everything to get ready for EFI and the learning curve (which hopefully these forums are helping to reduce), this part is really rewarding and fun. Better yet, you don't even have to open the engine compartment; just use your laptop!

Bitsyncmaster

12-08-2011, 07:20 PM

I had to add a little hysteresis the vacuum advance (turn off 100 RPM lower than the turn on) to prevent a jerking if your in gear at that RPM. So you would not want to do sudden advance with your ECU.

I still want to put a vacuum gage on my engine and see what values it sees at higher RPMs

opethmike

12-08-2011, 09:04 PM

I would advise against keeping 14.1 for every cell. You'll be going rich when you don't need to (cruise), and going lean when you don't want to (WOT).

Spittybug

12-08-2011, 09:57 PM

I would advise against keeping 14.1 for every cell. You'll be going rich when you don't need to (cruise), and going lean when you don't want to (WOT).

I shouldn't; the VE autotuner adjusts the amount of fuel based on the desired AFR. In other words, it should give me less fuel at cruise and more fuel at WOT, right? While I do indeed intend to alter the AFR table, since it is the target for the VE optimization, in theory I'd be close to stoicheometric under all circumstances as is.

opethmike

12-08-2011, 10:23 PM

Yeah, but you don't always want stoich. Stoich is good for cruise, but under acceleration you are going to want more fueling.

See the MS doc:

http://www.msextra.com/doc/ms2extra/MS2-Extra_Tuning_Manual.html#afr

Getting the fueling right in the areas that require more fuel will make your accel enrichment much easier to get right.

Byrne H&A

12-09-2011, 08:25 AM

Hello,
Set the distributor via vacuum guage (advance slowly) at idle: expect about 18". At 2000-2500 RPM the vacuum guage should read 24-25". Byrne

Spittybug

12-09-2011, 10:04 AM

Hello,
Set the distributor via vacuum guage (advance slowly) at idle: expect about 18". At 2000-2500 RPM the vacuum guage should read 24-25". Byrne

:headscratch:
I'm not really understanding that Byrne. Manifold vacuum is at its greatest when the car is at idle (but we block it from advancing the dizzy). The mechanics of the unit inside the distributor control how much it can advance the plate, and our book says 20* is the max, when the vacuum is greater than 15" of Hg "pull". As the throttle is opened, manifold vacuum drops as a result of being exposed to atmosphere (inches of "pull" goes down), so I guess I'm confused as to why you say the vacuum gauge would show more vacuum?

Maybe the difference is in manifold vacuum and throttle body port vacuum? With air rushing faster past the holes in the throttle body it would increase the venturi effect and lower the pressure for sure. The manifold vacuum isn't subject to that same effect because of the larger volume and slower movement of air.

In a stock Delorean the vacuum hose to the distributor comes from the thermal vacuum control valve and the vacuum source for that valve is the manifold itself, not the throttle body, right?

Or perhaps you were suggesting setting the static timing via vacuum gauge reading rather than timing light? Sorry, I'm confused as to your point.

Spittybug

12-09-2011, 05:16 PM

Using an Excel file I found that some MG owner had put together, I modified it to our stock specifications. It is pretty neat to visualize and see mathematically how it all comes together. Allows for "what-if" on getting the mechanical advance in quicker too. I guess we can't upload Excel files, so here is a PDF picture.

I'm going on the presumption that a lot of work went into developing the stock parameters (amount of advance and the specific springs used in the distributor) and they are therefore as good a starting point as possible.

I'm going to burn this to my ECU and see how it runs. The X and Y axes are configured so that I have a lot of detail down in the areas where the engine spends most of its time. It's ok that the table isn't linear as Megasquirt algorithms are calculating the appropriate values on the fly, not actually picking numbers off the table. The table is to establish the points from which to extrapolate the actual values.

jangell

12-09-2011, 05:24 PM

I guess we can't upload Excel files, so here is a PDF picture

Can you zip it and upload the zip? Just curious if that would work. That does look like a pretty cool visualization of the map.

-- Joe

Byrne H&A

12-09-2011, 05:32 PM

Measure the vacuum at idle and slowly advance the distrubtor to maximum vacuum while idling. If the distribitor is advanceed too far the vacuum will drop. At peak it may be 18 plus and if higher that is better. Tighten the hold down distributor nut and increase the RPM, the guage will advance to 23" or more at 2000-2500 RPM. Byrne

Spittybug

12-09-2011, 06:16 PM

Can you zip it and upload the zip? Just curious if that would work. That does look like a pretty cool visualization of the map.

-- Joe

Let's try that..... you may have to change the .zip to .7z for it to work... I use a program called 7-zip for zipping. Dunno.

Bitsyncmaster

12-09-2011, 06:18 PM

I just checked my vacuum. At 850 RPM idle I get 17.5". Between 2000 to 3000 RPM I get 19.5". When I punch it it drops to 5" or under depending on how fast I punch it. When I quickly let off the gas it can climb up to 25" also how fast I let off.

Farrar

12-09-2011, 07:25 PM

Dave, off topic for this thread I know but it looks to me like you may have a small vacuum leak. When my engine is idling at 900 RPM I have 20-21"Hg reading.

Farrar

opethmike

12-10-2011, 12:52 AM

As I am planning on going with ignition control sometime over this winter, I am keeping a very close eye on this thread.

I've already decided that the winter project for 2012-2013 will be cams/heads/exhaust. I'm 100% convinced that those will benefit greatly from the full MS control, and will make for a very productive dyno tuning session. Better than stage II, here I come! (....in a year....)

Spittybug

12-10-2011, 05:50 PM

I burned and tested today with a little modification from what I last posted. I removed vacuum advance from the high RPM, low load area of the table. In the stock setup, with the foot off the gas (deceleration) the micro switch would engage, cutting off vacuum advance, duh. Not shown on this table I also bumped the 40 MAP, 700 and 900 RPM bins to 17 or so to help reduce the oscillation at idle. Helps a bit, but the bigger culprit is difference in VE table. Notice I increased the rate of getting the mechanical advance all in. Now comes in between 1000 and 3000 RPM. After these changes I tooled around with VE autotuner running and things improved even more. Basically dropped some of the higher VE values I had and made for less "jitter" when letting the clutch out.

The test drive felt good and strong. Better acceleration, no signs of pre-detonation. When cold, the idle hunts just a bit just like stock, but not once warmed up.

C'mon you other EFI guys, together we can nail this down tight! Take the plunge!

opethmike

12-10-2011, 05:59 PM

I'm planning on taking mine to the dyno in the spring so I can get the ignition tuned properly. I'll be sure to put in information at that point.

dmc6960

12-10-2011, 06:04 PM

My radically different cams and compression will likely make my ideal values a bit different from yours. But thanks a TON for that table, it'll be a perfect starting point! I plan to use the accelerometer in my phone as a poor-man's dyno.

AdmiralSenn

12-10-2011, 07:19 PM

Once I finish my finals + current work craziness (retail + Christmas = FUN) I will be spending a few days doing my idle controller, then ignition.

Right now my driving days are few and far between and I haven't done more than 15 minutes with VE Analyze live on - too busy going to and fro.

Spittybug

01-03-2012, 04:17 PM

Odd event today.... After sitting for about 1/2 hour at the dentist's office, I came out to start the car and it took a few cranking attempts. I thought I had "gone back in time" and had a hot start issue! Very similar symptoms, but then she fired up. That got me thinking; I need to remove the fuel accumulator now that I'm EFI. What I suspect is happening is that the fuel system is losing its pressure via the accumulator and the long cranking is necessary to repressurize it all, including the accumulator itself. It certainly isn't needed, but I didn't think there was any harm in keeping it. I have seen that the gauge on the pressure regulator drops after about 15 minutes, so I know the system isn't holding pressure. I have no visible gasoline leaks anywhere, nor odor, so I think this helps confirm the accumulator as the culprit. It's going to be removed, lowering weight yet again :biggrin:

Even under normal conditions (both before and after EFI), my car has always cranked for about 4-5 sec. before catching. Battery is strong and all grounds good, so I suspect that I just have a relatively weak starter which is believe is original. Any suggestions on an alternative?

So here's the question; what is the generally accepted way to bypass the accumulator? Common the two hoses that go into the one end and blank off the single one that connects to the other end? I suppose it could be the fuel pump check valve too, is there a way to test that?

Bitsyncmaster

01-03-2012, 06:15 PM

You would need to pull connections and put a gauge in different locations. Probably pump outlet would be first. Do you have a pressure regulator in the system?

Spittybug

01-03-2012, 07:27 PM

You would need to pull connections and put a gauge in different locations. Probably pump outlet would be first. Do you have a pressure regulator in the system?

Pressure regulator is post both fuel rails, right before return line.

Bitsyncmaster

01-03-2012, 07:35 PM

Pressure regulator is post both fuel rails, right before return line.

If it's a bypass type, you could pull the return line and see if it leaks after shutdown.

AdmiralSenn

01-04-2012, 05:47 PM

Christine does the EXACT SAME THING now. Accumulator was a new (not NOS) install less than 1,000 miles ago. It sucks because she used to fire up the instant the key was bumped - it was seriously impressive.

My next "engine" part is going to be an AN-6 check valve and the pieces to bypass the accumulator. That or buy an internally valved regulator and sell the one I have.

For the starter, you will absolutely not regret an Eagle Premier unit. It's amazing. Smaller, lighter, cheaper, spins faster... it does draw a tad more current so naturally clean and/or upgraded grounds are important.

Bitsyncmaster

01-04-2012, 06:41 PM

If I were doing an EFI conversion, I would want to leave the accumulator to smooth the pump pulses and hold pressure. I've seen NOS accumulators fail quickly but a new one should last 10 years. Your even running less pressure than K-jet. I bet your leak is elsewhere. Do you keep the pump running a few seconds after the engine has stopped?

Spittybug

01-04-2012, 07:38 PM

Dave, pump primes before starting but shuts down immediately at engine kill. I replaced my accumulator less than 4 years ago with a DMCH unit. NOS I'm sure. I think my game plan is to remove it, test, and if necessary, then replace fuel pump or put in a check valve of some sort.

DARCOM

01-09-2012, 03:19 PM

i removed my accumulator, and did a bypass. all i have left is the stock fuel pump and have been debating changing it to one with out a check ball to get rid of the buzz when hot.
my system has never really held pressure except wile running. it leaks past my fuel pressure regulator. but when i bump the starter it fires right up if i have been driving it. now if i have let the car set for a week usually just turn the key to the on position let the pump prime then start the car. when i used to remove my rails they were always full of gas just no pressure, and i assume that the second i turn the key and being that the rails are hot ,fuel will make some pressure wile it evaporates i think i got my megasquirt priming settings just right to work this way.

Spittybug

03-07-2012, 06:09 PM

Weird one today. The only possibility is vacuum leak somewhere, but I'll be damned if I can find it. I'd been meaning to clean up my vacuum lines with my new manifold and throttle body setup. The necessary lines are:

MAP sensor line to MS unit - on its own nipple on manifold
Brake booster/vacuum reservoir - dedicated nipple on maniflod
Charcoal canister purge line (thicker) & crank ventilation (oil tube) - T'd together to 1 nipple on manifold
Charcoal canister signal line (smaller line) - to throttle body (ported vaccum)

I fired up car and the RPMs went above 2000 before settling around 2000 as the idle air controller ratcheted down. Before moving lines around I was idling fine. Of course the first thing I did was to make sure I had all of the lines well connected. When that didn't do anything I removed all of them except the Megasquirt MAP sensor and capped all of the nipples tightly. No change. I can't remove the Megasquirt one, but I know it must be sealed well because the reading on the computer is nice and low, indicating a strong vacuum. So where the hell is the air coming in???? Since it is the volume of incoming air that dictates idle speed (fuel is added based on the amount of air, but it is the job of the butterfly plates and/or idle air bypass to set the idle speed. My idle air stepper gauge shows me that it is shut and when I used a rag and my finger to block the idle air hole, indeed, nothing changed at all, so it wasn't sucking air through it. That leaves only my butterflies. Of course I can't find my T15 wrench to adjust it, but to my eyes the throttle butterflies look closed. No decel springs in them like Delorean ones either. I didn't touch a single mounting bolt or unscrew the air temp sensor or anything!! I'm really at a loss. It's running great and the throttle response is impressive, but I'm 1000 RPM too high, lololol..... I'm going to let it cool down and then in the morning I think I'll take off the manifold and make sure I haven't got a crack or a loose nipple or something odd that I can't detect with it on the car.

Before someone asks, I had the charcoal purge line and the crankcase vent both going to the throttle body before the change. No big deal, it was doing the trick, but it was getting a little oily in the throttle body.

This one has got me stumped, but there is no ambiguity about it.... too much air coming in from somewhere! I'll find it, but yet another reason why I hate relying on mechanical/vacuum/gravity/pressure systems rather than digital!

Farrar

03-07-2012, 07:53 PM

Is it possible that the little J-shaped vacuum line at the brake booster itself is leaking? I think folks easily overlook it; apologies if you have already checked, but I thought I'd mention it.

Bitsyncmaster

03-07-2012, 08:00 PM

A 2000 RPM does sound like a vacuum leak. I've found the ignition advance to yield about 500 RPM change. By any chance is your new idle valve bolts holding the plates from closing?

Spittybug

03-07-2012, 11:08 PM

Is it possible that the little J-shaped vacuum line at the brake booster itself is leaking? I think folks easily overlook it; apologies if you have already checked, but I thought I'd mention it.
I disconnected all hoses and capped the nipples, so moot point.

By any chance is your new idle valve bolts holding the plates from closing?
The idle valve/housing is an integral member of the throttle body and in no way interferes with the butterflies.

I'm going to sleep (or not) on it and try again in the morning. Thanks for the suggestions.

Farrar

03-07-2012, 11:19 PM

I disconnected all hoses and capped the nipples, so moot point.

DOH! I forgot about that. :coffee:

lazabby

03-07-2012, 11:24 PM

Could you try spraying ether around the engine? Wouldn't the engine RPM jump when it gets sucked in? That would help locate the area of the leak.

Spittybug

03-08-2012, 04:39 PM

Ok, so you've all read that thread about stupid things we've done right?:banghead:

In what could only be considered a senior moment I forgot that I had unscrewed one of the nipples from the back of the manifold. It was sitting in the fiberglass "lip" of the pontoon the whole time. That's why I couldn't see a hose-less nipple. DUH. FYI, should anyone care, a ~10mm hole will draw in enough air to idle our cars at ~2000 RPM!

I'm kinda glad this happened, it got me off my ass and I removed the manifold (easy). Now I can replace that heater valve and hose and generally route my wires the way I want to clean things up. I never really did that post EFI install. I think I want to change a fitting or two on my fuel rail setup as well. Now's the time to do it.

Dangermouse

03-08-2012, 04:43 PM

From Post #1;
In this thread I'll post pictures and data files about the custom conversion of my car.

:sleep:

Spittybug

03-08-2012, 08:21 PM

From Post #1;

:sleep:

Fair enough. Pictures to follow. And I'm sure you will share with us your progress?

Spittybug

03-09-2012, 06:10 PM

88648863 This is the Peugeot 604 manifold I found at an online parts guy from California. Note that it was for dual carbs, one was double barrel one was single barrel. The single port gets a blanker plate. I haven't plumbed it for coolant since warming the manifold is not necessary here in Houston. The backside has vacuum ports to which I connect the Megasquirt manifold air pressure line (MAP), the car's vacuum reservoir (for brakes and climate control), the paired line for charcoal canister and crank case ventilation, and one blanked off. The 90* aluminum adapter to allow for mounting the throttle body horizontally rather than vertically was a custom build by Strine Racing (shameless plug for Al and Nic). The manifold was modified by welding in the fuel injector bungs.

8858 This is the VOD before mounting the manifold assembly. The stock harness has been removed and rebuilt using new wires and polyester looming material. New wires for Megasquirt include: fuel injectors, throttle position sensor (TPS), coolant temp sensor (CTS - attached to our coolant sensor in the Y pipe), idle air controller (IAC) and intake air temp sensor (IAT). You see these laying in the valley waiting to be covered by the manifold.

8859 Close up of the throttle body I used (LT1 from a '95 Camaro I believe). You also see the TPS and two throttle body vacuum hoses (one blanked, one for carbon canister signal. The black thing on the bottom is the IAC valve that sits in the throttle body's housing and controls the amount of air that bypasses the butterflies at idle.

8862 Mounted on engine, no fuel rails yet.

8861 Pretty simple fuel rail setup: line from filter comes in from left, three injector holes drilled on underside of rail, cross over hose to passenger side, three more holes on underside, gauge mounted in rail, pressure regulator set to ~45psi, and off to return line.

8860 Whole thing mounted, wired and plumbed. I have a rubber boot to mount on throttle body face that connects to a hose to the passenger side pontoon within which I have a K&N air filter.
Look how FREAKIN clean that it! :rock_on:

I took all of these pictures today as I got ambitious and replaced the section of hose that includes the heater control assembly. Not counting fighting with hose clamps and trying to make things pretty, re installing the manifold, fuel rails and electrical connections to this setup takes about 5 minutes. 4 bolts hold down the manifold and the fuel rails.

Procuring one of these manifolds is getting progressively more difficult. Yes, the installation is nice and clean, but not everyone can find a competent welder to fabricate the necessary adapters and manifold modifications. That's what lead me to explore the stock conversion that is documented under its own thread.

I hope this motivates others to modernize from K-Jet.

Spittybug

03-27-2012, 06:39 PM

I'm a perfectionist. Is there an AA equivalent for us? I ripped out the wiring harness for my Megasquirt today and will be making it beautiful. I waited this long simply to ensure that it was a keeper, and the resounding answer is YES!

I've ordered a nice fuse box and relay sockets and will be doing the final install in the space behind the driver seat. Shortening wire runs that were deliberately left long and otherwise cleaning up the "functional but not pretty" setup.

I wonder if I should replace the wood panel and carpeting over it with a transparent cover....nah, overkill :hmm:

Spittybug

04-02-2012, 11:23 PM

Spent the day wiring it all back up. I'll post a picture tomorrow. I'm getting good at this stuff; fired up first time. The wires are now all trimmed to length, fused in a nice block instead of in-line and generally much nicer. I also got sick and tired of those 4 crazy brackets that hold the rear panel in place. I've always had a hell of a time getting the screws in. Pushing the panel tight against the wall, working upside down, crouching where the driver's seat should be, in the dark, trying to find the screw hole, glasses fogging up, etc..... So I got 4 long bolts, cut their heads off to make studs and screwed them into the holes. Then it was easy to push the brackets on and retain with a nut. I should have done this years ago.

My fuel pressure is a little different than when I started (backed off regulator to bleed down so I could remove manifold), so I will need to revisit the VE table again. That too is getting pretty quick and easy by using the VE tuner. Once the idle block of ~4 cells is identified and made equal to one another (stops hunting), then it's just a question of driving around and letting it do it's thing. The advance table and the AFR tables don't change from where they were. I'm going to spend a little time on the enhanced acceleration enrichment too like I mentioned earlier.

Spittybug

04-03-2012, 01:08 PM

Here is a shot of the driver side relay compartment as it now looks. I've tried to label the main pieces.

In a nutshell, the new FUEL relay gets power from a dedicated 12v line from the battery, through one of those fuses you see. It sends the 12v back to the RPM relay socket (W/P wire) which goes to the fuel pump. It is energized by an incoming 12v from the green wire in the RPM relay socket and this grounds via pin #37 on the Megasquirt unit (priming the system for a couple of seconds and shutting off).

The other relay gets its power from the brown wire in the RPM relay socket, through one of the fuses shown and distributes it to the near side of the fuse block, where each component, 2 banks of injectors, LC-1 and MS each get power, individually fused. This is also energized by the same 12v coming from the green wire in the RPM relay socket but is simply grounded via the RPM relay socket black wire.

I left room for a third relay if desired. As MS has programmable outputs that can drive a relay, I may wish to rig up something. For example, an audible warning if the coolant gets above a certain temperature.

There are a couple of connections between the wiring harness for MS and the LC-1 unit, but most run through the loomed bundle, through the firewall to the sensors and to a big ground location on the driver side head (not cover).

I think it's pretty clean. A schematic will get printed out and posted on the underside of the wood cover for easy tracing. I haven't yet done so, but at my leisure I will now unbundle and remove all in-cabin traces of K-Jet and Lambda wiring, cleaning up the fuse area.

Farrar

04-03-2012, 01:30 PM

I think it's pretty clean.

I think it's beautiful! Two thumbs up! :)

dmc6960

04-03-2012, 01:52 PM

Looks better than mine, great work!

Spittybug

04-17-2012, 03:13 PM

I got ambitious and turned on the enhanced acceleration enrichment ("sucked from walls" and "added to walls") and went for a spin. I'm pretty impressed. With just the default values in there I noticed improvement over my already pretty good tune. I also changed the MAP lag factor to 90 from the 50 I was at. That had the effect of making changes to the manifold vacuum (like stepping on gas from stand still) immediately correspond to RPM increase. I used to have a little hesitation or bog that would make me let the clutch out a bit slower than I normally would. I would get the sensation of inadequate power on launch. No longer. My car is definitely running the best it ever has, and I never had any real issues pre-EFI other than occasional hot start problems.

I'm going to dig through the log file a bit more and read the acceleration instructions again so as to make sure I've eeked out all the benefit. A quick look shows me for example that on sudden deceleration I go momentarily very lean. That shouldn't be a problem I don't think since it is brief and the engine is unloaded. However, it tells me that I need to decrease the sucked from walls or increase the the added to walls. Likewise, I go a little lean on acceleration, but not bad. All in all I like this better (its a transition compensation algorithm) than simple gas "pulses" that normal acceleration enrichment uses. I found those difficult to set the threshold for and equally difficult to get the pulsewidth correct for the rate of acceleration. I would sometimes get unintended little injections that felt like stutters when I was driving along.

I'm really getting the hang of this I think and once I've got this tune 100% as good as I think I can, I'm going to make sure to save the project and then load a new code. There's a guy on the Megasquirt 2e boards that has taken the most recent code variant, 3.30, and added some cool modifications to the idle control. Even though my idle is pretty solid right now, it sometimes will hunt just a bit when the a/c compressor kicks in or while still warming up. Since I'm all digital now I can get it to settle down to where it should be by just giving it a quick rev if need be, but this new code is closed loop and somehow predictive I understand. It's getting a ton of good comments. If I don't like it or can't get it to work, 5 minutes later I can have the original project re-burned and I'm done.

I know we are a patchwork of different engines and intakes, but if one of you other guys wants to load my code and give it a shot on your car and see how it feels, go right ahead! I'm a stock 5 speed with the only modification being the Peugeot manifold and my EFI (spark and fuel).

opethmike

04-17-2012, 03:36 PM

I think I may just give the EAE a shot after I get back from the spring social in NJ this weekend. Like you said, if I don't like it, I can just revert back.

Farrar

04-17-2012, 04:59 PM

Does the Peugeot 604 manifold make EFI conversion any easier? I may be able to get one... a first step in a long and expensive journey toward EFI, possibly.

Spittybug

04-17-2012, 07:40 PM

Does the Peugeot 604 manifold make EFI conversion any easier? I may be able to get one... a first step in a long and expensive journey toward EFI, possibly.

Farrar;

I think our work on Shannon and Ian's cars is evidence that you do not NEED to go with such a manifold. While it makes the car lighter, the engine bay look nicer and does indeed make working on things MUCH easier, it is not necessary to get the benefits of EFI. There is cost associated with a new manifold in that you must now also procure and mate a throttle body (and linkages). The whole reason for developing the stock alternative was to avoid such aggravation and cost.

"Expensive" is a relative term. Yes, the box costs ~$400 and the need for fuel rails, wide band O2, idle air block and injectors add to that, but is ~$1,000 for all of that "expensive" compared to K-Jet parts? You do have to be prepared to learn, be patient, experiment, learn more and then be sufficiently confident to show off your new found knowledge.

I for one would NEVER go back, but it can be frustrating to solve some of the challenges that the internal combustion Gods throw your way. :smile: The key is truly understanding your car......almost zen like.

Farrar

04-17-2012, 07:55 PM

"Expensive" is a relative term. Yes, the box costs ~$400 and the need for fuel rails, wide band O2, idle air block and injectors add to that, but is ~$1,000 for all of that "expensive" compared to K-Jet parts? You do have to be prepared to learn, be patient, experiment, learn more and then be sufficiently confident to show off your new found knowledge.

"Expensive" is indeed a relative term. In my case it simply points to time. The more expensive something is, the longer it takes me to get it. I do like the look of the Peugeot manifold, though, so I may spring for it and keep it as part of a future EFI kit for my car. Thanks!

Spittybug

05-21-2012, 06:05 PM

Wow, wow, wow! Over in this thread: http://dmctalk.org/showthread.php?2057-Software-amp-Tuning I made mention of the gslender version of the MS2e code that I loaded onto the controller. Well today I made the change to my ground wires from the sensors so that they all go to the MS unit's ground plane, which in turn is grounded to one location on the engine head. My coolant, throttle and incoming air readings are now smooooooth. I've also got the acceleration enhancement in pretty damned good shape. While I'm using 100% throttle based, I looked at MAPdot on the attached log and there is really no reason it wouldn't work just as well (for the guys without TP sensors). The key is to just keep the same relative amount of PW added as I have, just use a good kPa threshold value so that you don't get unintended spikes.

I've attached my tune (I've got one or two bins I want to change, but that's about it..) and a log file from a drive I took this afternoon to pick my 16 year old son up from high school. Yup, that makes his day when all his friends drool and he just plays it cool.......

For anyone lurking that wants to start learning more about all of the diagnostics that we discuss in these threads, you can download a free version of the log viewer to open these files. Once open you can make it graph a whole bunch of the variables that were recorded as the car was driven. You can even define your own scatter graphs to see correlations. Two screen shots are included for those that don't want to get their hands too dirty yet. Really neat stuff. Here is the link to EFIanalytics download page: http://www.tunerstudio.com/index.php/downloads I know if I had played more with this stuff before starting my tuning, I would have been done in a fraction of the time.

Oh, I don't know if it is related to having dialed in the tune or if it because I got rid of the long air intake hose that went to my K&N filter in the passenger pontoon (smallest diameter was only 2" and therefore restrictive) and just attached the filter after the rubber adapter to the throttle body, but I now have a pronounced throaty growl from my exhaust! :bigclap: Either that of I've got a muffler leak..... But damn, it sounds like it should!

Keep in mind that if you want to load this tune into your controller you will need to have the gslender code installed. Here is the download area: http://www.msextra.com/forums/viewtopic.php?f=91&t=44946#p320193 So far I HIGHLY recommend it. Looking at my log I was surprised how much the battery voltage fluctuates. I'll be digging into the smoothing feature so that it doesn't adversely affect the pulsewidths for those momentary changes. Stay tuned.

opethmike

05-22-2012, 03:34 PM

Speaking of battery oddities - my Tunerstudio never reports a voltage for a battery that matches what a multimeter will read.

For example, the other day after replacing my alternator belt, I checked the voltage from the jump lead and ground in the engine compartment with my multimeter. I read 14.4.

But when I looked at TunerStudio, it reported only 13.1.

Any idea what is up with that disconnect?

dmc6960

05-22-2012, 03:37 PM

Every single connection between the raw 12v and the GND introduces a slight voltage drop, as does the wires themselves. After its all added up, you have what your reading.

Bitsyncmaster

05-22-2012, 07:28 PM

opethmike

05-22-2012, 08:47 PM

You would need to measure the voltage at the ECU power pin and ECU ground pin. This may show your where you need to use larger gauge wires. Of course it may be a calibration fault of the ECU but 1.3 volts sounds excessive.

I'll do that. I'm not having any voltage issues (at least not that I notice!), but it still is curious. I'll update after take the measurement this weekend.

opethmike

05-28-2012, 07:06 PM

Spittybug

05-28-2012, 08:09 PM

Owen,

I noticed in your tune file that you are running rich AFRs across the board. Wondering what your thinking/logic was here? I've run a similar AFR table, and just ended up with terrible gas mileage. I've found that running around 14.5-15 for my cruising areas gives me about 20 city, 25 highway.

-Mike

Higher torque, more power, feels good! While 14.7 is stoicheometric, it isn't the optimum for power. The relatively few times in life I drive the D (compared to all other cars), I don't really worry much about the gas. I also don't want to get things too hot. Richer=cooler. 12.6 : 1 is power optimum for gasoline if I recall correctly.

That said, I'm a tweaker and am always willing to experiment with different values. Right now I'm running the gslender modifications to the MS2e code and learning to do the closed loop idle air and RPM targeted idle advance features. Powerful and can already see that it makes the idle very solid. You can hardly even tell it is odd fire!

opethmike

05-28-2012, 09:09 PM

Ah, gotcha. I keep my cruising areas at 15, and then I progress to 14, 13, and finally 12 for the acceleration band. I've found that the car performs beautifully like that.

Do keep in mind that I'm running very large cams (270 degrees, versus stock 192!!!), so the location of my powerband is very different. Pretty neat to see the differences between engines, isn't it?

The stock DeLorean cams peak their torque at 2,750, and drops off after 4,500. Mine builds almost linearly until the peak at 5,000, and then stays steady until red-line. Very different behavior.

opethmike

05-28-2012, 09:12 PM

I'll post my tune in a few days, so you can check out how I'm doing things with my set-up.

Another curiosity question/point - I have found that the Megalog's VE analyzer does a MUCH better job tuning my VE table than VE Analyze live. How about you?

Spittybug

05-28-2012, 11:46 PM

While both work pretty well, I think I agree with you that the Megalog one is a bit better. I also found that both leave the table fairly uneven. I like to put it on the big screen (not laptop in car) and look at the 3D view. I then smooth the whole thing out. It just doesn't make sense to have a pit or a peak in the middle of a trend. Then I drive and capture a log. I'm getting pretty good at looking at the log with the .msq open and seeing the correlations of the AFR values and the position on the VE table. You just have to remember that the AFR spikes/dips come a couple of time pulses after the VE table setting (lag). I then just tweak those couple of cells up or down to flatten the AFR trace for next time. Do this a few times and the table is smooth and your AFRs are pretty good. They do bounce though, no two ways about it.

The EAE and the theory of the wall puddle explains the bouncy AFR. Depending on air velocity in the manifold, temperature and the MAP, varying quantities of fuel get stuck to the manifold and intake port walls. That means something other than what you commanded on the VE table is actually making it into the combustion cylinder. As you drive and the "bouncy blue ball" goes up down and side to side, this puddle is always changing, thus too is your resulting AFR. The EAE really did help stabilize the AFR and made it hug the target AFR values much closer (smaller Beta for the financial people in the crowd). The problem I encountered with the EAE was that when I would let off the gas quick and the MAP value dropped hard, the AFR would peak off the chart - super lean. In other words, the curves were telling the unit that more was getting sucked off the walls than really was, or that less was being added to the walls than really was. These are the two curves. Problem was that I couldn't find the right changes to make without introducing an oscillation. You have to know how to offset the change made on one curve with the other. Tough without lots of bench testing. I quit trying.

By using the gslender code which has a TPS noise dampener in it, I got the TPSdot down to nothing..... no false readings of acceleration. This allowed me to go 100% TPS based acceleration enhancement. With a low threshold, a very low pulsewidth boost at that low threshold, and a shallow curve, I've been able to get rid of AFR spikes under quick acceleration and don't have them on decel either. Whether a slow and steady acceleration or a throttle stomp, my AFR stays pretty close to target and recovers quickly. It doesn't hug as tight as when using the EAE, but shit, who cares if the performance is good and I'm not damaging the engine by going lean. Moving the MAP lag up to ~85% - 90% made the response very good. Now that I know how the acceleration curve (pulsewidth values) looks against TPSdot, replicating that shape using MAPdot (for the stock conversion guys without TP sensors) isn't hard. Yes, MAPdot is a little noisier, but not much. May have to go with a slightly higher threshold.

So, good VE table based on AFR table built to get more power than economy and my best attempt at an advance table that is a bit more aggressive than stock. It's running better than ever. My focus now is on the idle and warm up features mentioned earlier. Goal is to get it like my Infiniti; just turn the key and after a short high RPM period have it settle down into a rock solid, no oscillation 800 RPM idle that self corrects when additional load like A/C or headlights are added. I'll get there shortly using that new code I'm sure. I'm going to see if I can get any good accelerometer data recorded with my Samsung Galaxy. Poor man's substitute for a dyno. Once I get things the best I can, maybe then I'll go find a dyno somewhere.

Then I can start with the fun stuff like seeing how much I could advance the timing if I add HHO or plain water injection......or maybe experiment with turbos..... we'll see! I'm getting a Ferrari 308 bug up my skirt to though, so I'm trying to learn more in case I get the opportunity. Timing belts is their Achilles heel, along with K-Jet! The latter I can lick. Early models were carbed, but I don't want to start that discussion!

I hope that all of my tinkering and sharing benefits anyone else looking to take the plunge...

opethmike

05-29-2012, 09:36 AM

Other than an area or two of the VE table at very load and RPM (Think creeping along slower than usual in a parking lot), I'm not going lean anywhere, and my car is driving very nicely.

Cruising along, my AFR hovers around 15 and drives very nicely. When I stomp on it (which isn't a rare thing! :evilgrin:), the car responds very quickly and nicely; between the quick burst of AE and how my AFR table is laid out.

My cold start idle goes a little high for a few seconds, then settles into a rock-steady 800 RPM. I get no RPM change when I turn on the A/C, or the fans come on. At idle, my AFR is ~14. When the A/C comes on, the AFR goes to ~13.5. I've found that by running a hair richer in the higher load cells that the A/C goes to that I get no change in idle RPM or off-idle response. So for my engine setup, I believe I have my MS set up pretty close to spot on.

It really is pretty interesting just how different the fueling requirements are for different engines, I think.

Spittybug

05-29-2012, 09:59 AM

Cool. Hurry up and post your .msq and a log file, I'd like to see them! Do you have an Android phone? There is an app called ACmeter that I'm going to use. I don't know if Apple has something comparable, but I'm sure it does. Comparing performance would be a good accompaniment to the tunes. Yes, you are Spec 2 and I'm stock other than my manifold, but it would still be interesting.

opethmike

05-29-2012, 10:05 AM

I'll have stuff posted this weekend, hopefully. As far as performance goes, I have clocked a 0 - 60 of 6.5 seconds.

opethmike

05-29-2012, 11:31 AM

EarlHickey

05-29-2012, 02:07 PM

Higher torque, more power, feels good! While 14.7 is stoicheometric, it isn't the optimum for power. The relatively few times in life I drive the D (compared to all other cars), I don't really worry much about the gas. I also don't want to get things too hot. Richer=cooler. 12.6 : 1 is power optimum for gasoline if I recall correctly.

That said, I'm a tweaker and am always willing to experiment with different values. Right now I'm running the gslender modifications to the MS2e code and learning to do the closed loop idle air and RPM targeted idle advance features. Powerful and can already see that it makes the idle very solid. You can hardly even tell it is odd fire!

13.3 has been found to be optimum for max power in many archaic engines having combustion chambers far superior than the prv.

Even an agriculturally seeming shovelhead pushrod hemi running 10 second quarters, if comparing appearances and refinements to the prv, is doing so with that mixture and 25 degrees spark advance, or less, and reasonably streetable.

I'm not surprised you are finding 12.6 makes max power in this engine.

I'm sure you know the more you advance the timing, the higher the chamber temps become while the exhaust temps drop.......... and vice versa.

You may find it beneficial to consider whittling in a multi-stage electronically actuated EGR valve, if a strictly operated vacuum unit doesn't suit you.

There is also a hysteresis window that bootstraps efficient igniting of the combustion event to mixture.

It would be interesting to know how rich the chamber and ignition will tolerate before eight stroking or fouling.

I have an engine which uses one slide/needle carburetor that will run without fouling a plug even with the needle jet, it's holder and the main jet removed.

It does get a tad lumpy and harsh under a load, but, can motivate down the road just the same.

It's chambers and ports are a bit unusual.

Spittybug

05-29-2012, 04:26 PM

Oh, and regarding the 308 - don't worry about the timing belts. I've done that job on a 308; its easy, good access through the wheel well, and takes only 5 or so hours from start to finish. The K-Jet is NOT an achilles heel on the 308. However, the fusebox is! If you do get a 308, one of the very first things you should do is get Birdman's updated fuse box. Trust me on this one!

I watched a video on the belt replacement and I think you are spot on. Just getting the AC compressor out of the way and getting one of those little jigs to hold the cams from rotating seems to be the key. I've already been warned about the fuse box and I've even seen Birdman's replacement! So the K-Jet is somehow less troublesome on the 308 than it is on our little Ds?

Spittybug

05-29-2012, 04:30 PM

It does get a tad lumpy and harsh under a load

So do I Earl. What's the point of your post, if there was one?

opethmike

05-29-2012, 04:42 PM

I watched a video on the belt replacement and I think you are spot on. Just getting the AC compressor out of the way and getting one of those little jigs to hold the cams from rotating seems to be the key. I've already been warned about the fuse box and I've even seen Birdman's replacement! So the K-Jet is somehow less troublesome on the 308 than it is on our little Ds?

You don't have to have one of those jigs. There's nothing mystical about the timing belt setup; just like any other car as long as you don't rotate the cams you are fine. And you are VERY right, getting the A/C compressors out of the way is VERY annoying.

I couldn't tell you why K-Jet seems to behave better on the 308s, but I've driven and worked on 3 different 308s (all cars I've known for a few years), and I've never run into a fueling issue with them.

EarlHickey

05-29-2012, 06:12 PM

So do I Earl. What's the point of your post, if there was one?

I couldn't imagine how anyone could possibly interpret any of that as having a point or even any possible usefulness.

Mea Culpa.

If it doesn't bother you that your efi car apparently guzzles gas, it doesn't bother me.

Rock on.

82DMC12

05-29-2012, 07:57 PM

So do I Earl. What's the point of your post, if there was one?

He's just toying with you.

opethmike

05-29-2012, 07:59 PM

He's just toying with you.

+1 on the troll alert

EarlHickey

05-29-2012, 08:56 PM

+1 on the troll alert

Yeah, it may not be a good idea to attempt an EGR valve with this quantity of raw fuel being pissed out your exhausts.

Besides, how ever in hades could recirculating exhaust gas ever benefit an engine with very pronounced and obvious chamber woes?

It's not a rhetorical question and you either know it or you don't.

http://www.youtube.com/watch?v=Pgs4bpfAVYQ

Spittybug

05-30-2012, 10:59 AM

This is an AFR table, not a lambda table.
For anyone wishing to understand why I have set my AFR table up the way I have, and why I shoot for under the 14.64 "optimal" AFR;

The theoretical air fuel ratio for 100% ethanol is: 9.0078 : 1
So, for "gasoline" sold these days with 15% ethanol content, 15% x 9.0078 + 85% x 14.64 = 13.795
FYI, remember, the math for these is based on mass, not volume. Ethanol is a couple of percent denser than gasoline (which has a fairly broad range depending on time of year and blended components).

Gasoline must meet octane specifications. Blending ethanol (RON 108.6) does not require the addition of any other octane boosting component to offset it, (it in fact allows the use of lower octane components so that the resulting pool is at spec) so I've gone with the relatively safe assumption that "standard" gasoline's optimum-combustion AFR of 14.67 didn't materially change when ethanol started getting added in larger volumes. I run 87 octane in my car with no knocking issues where I have set the timing table.

I am neither a combustion engineer nor a professional auto tuner, but I do understand refining, gasoline composition and chemistry. As mentioned many times in these EFI threads, it's all about learning, trial and error and optimization. I welcome constructive advice, questions, suggestions or practical help (such as guidelines on exhaust temperatures perhaps?), but I don't welcome sarcastic, cryptic, boastful, arrogant or inflammatory posts.

As prefaced in the "lobby" of the EFI section, this IS a moderated forum and I DO intend to keep it that way with the previously granted blessings of the moderators. If you don't like it, don't post.

opethmike

05-30-2012, 11:43 AM

"So, for "gasoline" sold these days with 15% ethanol content, 15% x 9.0078 + 85% x 14.64 = 13.795"

You know, I hadn't even thought about that. That's a great point, and certainly something I will noodle on. I should have my VE table mostly smoothed out for my current AFR table this weekend, so it will be posted along with a log at that time.

Spittybug

05-30-2012, 12:18 PM

Despite what some might say, the internal combustion engine, at least of the era we are concerned with, is NOT a precise instrument. There will be many factors to make the numbers go up and down. We are talking about doing the best we can using stock or similar era engines within reasonable boundaries. If money, equipment and technology weren't an issue, sure these things could be much better. One of the reasons we are on a DELOREAN forum is because we want to keep them as close to old-time DELOREANS as possible while marginally modernizing them. If someone wanted to slap a Delorean shell over a modern or extensively modified drive train and then call it a Delorean, that is their business. That's not the purpose of this EFI forum.

I think the key is simply to get a base line that is as correct as possible most of the time and to avoid big excursions when accelerating/decelerating. By altering the VE table for different RPM and MAP conditions along with acceleration enhancement and AFR feedback, we are eliminating a lot of the variables that get introduced in a system that used MAP only (air plate deflection) and then a feedback loop based on a narrow band lambda sensor.

However, while Lambda in name, I'm not aware that the stock Lambda system adjusted for changes in the gasoline composition in any way. I understand the definition of Lambda to be the ratio of AFR to the stoicheometric AFR. Does the stock system somehow analyse the composition? Modern cars have sensors that recalculate the stoich AFR based on the incoming fuel (hence flexi fuel capability). Question to self; does this mean that all stock cars are now running ~1 AFR point leaner than they think they are due to the new ethanol volumes in our gasoline?

Adjusting the timing based on true cylinder needs/wants rather than simply based on mechanical feasibility also adds performance and more precise control. Ever wonder why the idle micro switch disengages the vacuum from advancing the distributor timing at idle? Stock Deloreans idle in the 13* range while many cars idle many degrees higher due to the "thinness" of the mixture and the extra time needed for flame propagation. The INSTANT you come off idle, the vacuum advance goes maximum, or another 10* of advance! Why not just set it there? Well, try cranking and starting with 23* of advance and you'll find it very hard. But it isn't bad once running...... How to accomplish that with simply vacuum? Hummmm.... I don't know why they didn't just disable the vacuum advance on crank and then let it come back in at idle.

If I'm wrong on the above items, I sincerely welcome civil and productive correction of such.

Bitsyncmaster

05-30-2012, 01:18 PM

But your O2 sensor (wideband) reads lambda. Then it converts the lambda reading with a formula. It does not matter what your burning, you want a lambda of 1.00 (for best burn). Unless you have changed the formula, you still want the reading of 14.7 AFR with any mixture of gas.

Spittybug

05-30-2012, 03:12 PM

Dave, from the LC-1 manual:

The LC-1 is a stand-alone Wideband Controller used to measure the Air/Fuel Ratio (AFR) or Lambda for an engine. For gasoline-driven engines, the theoretically optimal air fuel ratio is 14.7 pounds of air for every pound of fuel. At this ratio, theoretically, all available oxygen in the air combines with all available fuel. This ratio is called the stoichiometric ratio. Stoichiometric for different fuels are as follows:

Gasoline 14.7

LPG (Propane) 15.5

Methanol 6.4

Ethanol 9.0

CNG 17.2

Diesel 14.6

The measurement Lambda is the actual air fuel ratio over the stoichiometric ratio. A Lambda measurement of “1” equates to the air fuel ratio of 14.7 (for gasoline engines). When Lambda is less than 1 the engine runs “rich”, i.e., unburned fuel exists in the exhaust stream. If lambda is greater than 1 the engine runs lean, i.e., free oxygen (02) is present in the exhaust. Depending on the engine, maximum power is typically delivered when the engine runs slightly rich (for example at lambda values of 0.8 to 0.9 for most engines). This instrument provides a means to measure the actual air fuel ratio or lambda in the engine in operation directly from the exhaust. For this a special wide-band oxygen sensor is used to measure the lambda value derived from the oxygen content (or lack thereof) of the exhaust gases.

So, you've answered my question about Lambda; the composition of the incoming fuel is irrelevant to the lambda probe, it is reporting the ratio of air/fuel no matter what is coming in, normalizing that around a value of 1. In our AFR table case with EFI, we are taking advantage of the fact that the LC-1 can report the output as AFR for a given fuel or a Lambda value. I have gone and manually translated a table to AFR for a fuel that it a mix of ethanol/gasoline, whereas my life would be simpler to just set up the table for Lambda vs. RPM and not worry about the math. Thanks, I'm going to dig into that in the software! I was under the mistaken belief that the probe measured AFR directly. I don't think your last statement is correct though; No matter what you are burning you do want a Lambda of 1, but what that translates to in AFR depends on the composition of the fuel. A Lambda of 1 for ethanol is 9.0 AFR, right?

Dave, you NEED to come over to the EFI side....... :woot: You'd have so much fun with the capabilities!

opethmike

05-30-2012, 03:14 PM

Dave, from the LC-1 manual:

So, you've answered my question about Lambda; the composition of the incoming fuel is irrelevant to the lambda probe, it is reporting the ratio of air/fuel no matter what is coming in, normalizing that around a value of 1. In our AFR table case with EFI, we are taking advantage of the fact that the LC-1 can report the output as AFR for a given fuel or a Lambda value. I have gone and manually translated a table to AFR for a fuel that it a mix of ethanol/gasoline, whereas my life would be simpler to just set up the table for Lambda vs. RPM and not worry about the math. Thanks, I'm going to dig into that in the software! I was under the mistaken belief that the probe measured AFR directly. Dave, you NEED to come over to the EFI side....... :woot: You'd have so much fun with the capabilities!

Not to mention what a great boon Dave would be for this small community of ours!

Bitsyncmaster

05-30-2012, 04:49 PM

Lambda 1.00 will burn all the fuel. I'm not saying you want to stay at that value. I've found the engine will run much better running a little richer. But you will sacrifice a little MPG running richer.

Your better off just doing all your mixture adjustment using lambda values. Then things don't change when they start playing with the ethanol content again. They still use 10% in my location but the farmers are pushing for 15%.

Spittybug

05-30-2012, 06:29 PM

I just did some more research and chatted with Philippe @ Innovate. The unit simply puts out a voltage between 0 and 5 V which Megasquirt can interpret as either AFR or Lambda depending on how you initially set it up. I created a new project (so as not to muck up my good one) and configured it for Lambda, including the dashboard gauges. I imported my tune and MS automatically changed the AFR table values to Lambda values using 14.7 ratio. So, in this project I now have a table of values under 1.0. Since I was assuming a 13.7:1 ratio, does this mean that if I were to re-tune (automatically) to these values that I would immediately get leaner by 1-(13.7/14.7) = 7%?

This is all squirrely in my head right now, so what I'm going to do is simply re-tune the VE table shooting for a Lambda of 1.0 in the cruise areas, richer in the idle and richest in the WOT area. For the varying composition of gasolines around the country and the world, this should provide for a more standard table/tune for others to benefit from/compare.

I need another martini. :relative:

opethmike

06-02-2012, 06:26 PM

Owen, I've been using your AE table/values for a few days now. Very nice!

Spittybug

06-02-2012, 07:46 PM

Cool. That's the whole reason for sharing.

I got motivated this afternoon and tried the gslender code for closed loop IAC control again. Not 100% sure, but I think I've got it pretty close. Like most MS things, the trickyness is in the fact that it so variable. You have to carefully define your RPMs with the IAC fully open and fully closed (set the extremes). You also have to carefully define the number of steps for fully open and fully closed. This is tougher than it would appear to be!

After that, you set the P, I and D (usally 0%) gains by trial and error. These values are part of an algorithm that returns the idle to the point you set for any given coolant temp (a curve). You are optimizing how quickly it gets there, the amount of overshoot and the absolute accuracy. This takes a lot of patience and trial and error.

In addition, the code has an adjustment for VE at idle and an adjustment for timing at idle. Once all are enacted, the car performs like a modern one with respect to "settling down" and adjusting for changes of load at idle. I need to go road test after my changes.

Mike, the code you loaded, was that before or after I went to Lambda instead of AFR? Because Lambda is fuel indifferent and is measuring stoicheometric optimum no matter what fuel, it makes more sense to use. You have to make that definition at the project level, not on the menus. Once done however, it will automatically change your AFR table to a Lambda table by dividing by 14.7. Since it doesn't change the VE table, the car drives the same. I tried leaning my table out (5% IIRC) and I didn't notice a performance change, so I've adopted it for some fuel savings.

On a separate note, we had our monthly "Cars and Coffee" meet this morning. WOW. This thing has gotten huge! If you aren't there by 7:00 AM, you don't get a spot! Ferrari had a semi there with 3 race cars. Some BIG $$$ vehicles there. Bugatti, Maserati, Lamborghini, Porsche, Ferrari, Fisker, you name it. On the parade of cars to exit the show at 9 AM or so, everyone lines up to take pictures. Who do you think they take the most pictures of and shout out the most praise for?????? YUP, the Delorean with the doors open, Disturbed blaring on the stereo and the digital dashboard on the laptop!! Good times.