Software & Tuning

[Spittybug]

Content to come.

Here is my .msq file that runs well. I use Tuner Studio rather than Megatune and think it is much better. Suck it up and pay the $50 for the full version; it gives you the autotune feature which is really helpful. You can drive and let the software optimize the fueling table in real time based on your desired air/fuel ratio. AWESOME.

[Spittybug]

The degree to which you jump in is up to you; simple fuel control or full spark control as well. With success comes the desire to push even further! Megasquirt software for example can be used to control other features (like relays) based on operating parameters. Say for example you wanted an alarm to sound if the coolant temperature exceeded a certain level - no problem. There are even ways to make compound conditions like; if throttle position is closed and RPMs drop below 750 with A/C kicking on, then open idle valve slightly to bump up idle. It is this kind of flexibility that makes programming one of these things fun after the initial start up curve!

Getting going can be tough when done alone. Let's make sure we share our collective knowledge. Tuning the 3 tables (Fueling, Spark Advance and Air/Fuel ratio) is an art that can give tremendous flexibility for more power, fuel savings or emissions depending on how you set them up. Switching to new tables is a mouse click away. Having a library of tables that had been tried and tested by someone running the same motor would be great!

A PDF document detailing the software and tuning of the Megasquirt system will be posted shortly.

This link will take you to the Megasquirt boards where TONS of information (not always organized in the best manner) can be found: http://www.msextra.com/doc/index.html#ms2 They have their own forums too, look at the left menu, three quarters the way down on this link: http://www.megamanual.com/index.html

[opethmike]

Here is my MSQ, which runs very well, for the following set up:

MS2Extra, version 3.57 board, firmware 3.1.1
Jeep Cherokee IAC valve
Renault Z7U intake manifold/throttle body/fuel rails
24 lb/hr Accel (150624) high impedance injectors
Innovate Motorsports MTX-L wideband O2 sensor

This tune should work as a good base for similar EFI installations

[opethmike]

Documentation on my conversion can be found here:

http://dmctalk.org/showthread.php?2445-Megasquirt-EFI-conversion

[Spittybug]

I spent a couple of hours working on my acceleration enhancement this afternoon. I have it working pretty well I think. One might almost believe this is a sports car!

I have it 99% controlled by the MAPdot rather than TPSdot. This is the recommendation of the MSextra write-up I read. Rather than regurge all the data, I've attached a screen shot.

I still have my AFR set at 14.1 across the board which isn't optimized and I want to revisit my timing table a bit, but I'm running very nicely! So if anyone is looking for a decent starting point for the acceleration enhancement, try this. I'll also post the .msq and my latest log.

[Spittybug]

I've been playing around with my tables again and I learn a little more every time I do. The VE analyzer of Tuner Studio is amazing and the newest version of Megalog Viewer has some neat new tools too. Well worth the money for both.

I used an online calculator I found somewhere along with our manual to determine my spark advance profile. I have it maxed out at 38*, but I'm wondering if I've calculated that wrong. The question I have is just where does our stock setup max out? According to the manual:

13* static (simple mechanical rotation of whole distributor)
Up to 20* for vacuum advance (purely a function of the load on the engine as this provides the suction to rotate the guts of the distributor. Highest amount under lightest load)
Up to 20* mechanical advance (a function of engine RPM. Weights and springs control the amount of rotation of the guts of the distributor)

Here's where I've forgotten... are the vacuum advance and mechanical advance additive (each capable of adding up to 20* independent of each other) or union (either one or both together can add up to 20*)? This depends on the mechanics of the distributor.

This makes a big difference; maximum of 33* or 53* of advance. I get no sensation of knock & ping (predetonation) at 38* using 87 octane. I do however feel like I'm not getting all the power I could be getting at the higher RPMs. I'd like to advance more but wish to understand the limits before I do.

To those who are new to this concept, it is really pretty simple. A flame front going through your cylinder at a given air/fuel ratio only goes so fast. There is an optimal point in the compression stroke (a little after TDC) when you achieve the most power from igniting the mixture. Well, as your engine speed increases and the flame front moves at that constant rate, you ignite your cylinder charge later and later, losing lots of power. The solution is to start the igniting earlier so that it ends at the right time at any RPM. Hence the mechanical (RPM controlled) advance.

A caveat to the above was that the flame front speed is air/fuel ratio dependent. Richer mixtures (what you typically try to have for wide open throttle rather than cruising) allow the flame front to travel more quickly since the molecules are packed in tighter. Leaner mixtures cause the flame front to move through more slowly. Hence the concept of vacuum advance. Wen the car is under minimal load and the mix is leaner, we need to plan on the flame front taking more time, so we ignite earlier than we do when the mix is richer. Distributors do this by using the amount of engine vacuum to advance the spark. High vacuum at idle, minimal at WOT. To prevent the amount of advance from being excessive at idle, our cars electrically block off the advance at idle, but immediately upon coming off idle, the vacuum advance goes up to a full 20* (plus the static of 13*).

I think the answer is that since both the load and RPM advances are for separate reasons, they should be additive. At high RPM, low load (cruising at 70 mph for example) the spark needs to go early for the RPMs and is going through a lean mixture, so it should want to go really early, like 53*.............................

I can change it in 10 seconds with Megasquirt, but I don't want to blow up my engine if I'm wrong!! What is stock maximum??????????? Thanks.

[Spittybug]

I think I just found my answer.....

Plug the vacuum hose back into the distributor. You will notice the timing mark shift to well over 50 degrees. It's cool, vacuum advance has just been introduced into the system and its function is to provide extra advance when a vacuum signal is present to improve fuel economy (or so I've been told). Because this is in addition to your initial and mechanical advance (aka your total advance) you notice the degree spike when it's plugged in. The amount of advance can be adjusted by swapping out different vacuum canisters or by putting on an aftermarket adjustable canister. Unless you are a hard core racer, most of the time this is not worth the trouble.Found this at home.roadrunner.com a website for Mopar guys.


The vacuum advance control unit on the distributor is intended to advance the ignition timing above and beyond the limits of the mechanical advance (mechanical advance consists of the initial timing plus the centrifugal advance that the distributor adds as rpm comes up) under light to medium throttle settings. When the load on the engine is light or moderate, the timing can be advanced to improve fuel economy and throttle response. Once the engine load increases, this "over-advance" condition must be eliminated to produce peak power and to eliminate the possibility of detonation ("engine knock"). A control unit that responds to engine vacuum performs this job remarkably well. Most GM V8 engines (not including "fast-burn" style heads), and specifically Chevys, will produce peak torque and power at wide open throttle with a total timing advance of 36 degrees (some will take 38 ). Also, a GM V8 engine, under light load and steady-state cruise, will accept a maximum timing advance of about 52 degrees. Some will take up to 54 degrees advance under these conditions. Once you advance the timing beyond this, the engine/car will start to "chug" or "jerk" at cruise due to the over-advanced timing condition. Anything less than 52 degrees produces less than optimum fuel economy at cruise speed.
Per Lars Grimsrud, SVE Automotive Restoration
Musclecar, Collector & Exotic Auto Repair & Restoration

I'd still love to know what our max is for our engines.......but it sure sounds like 13* + 20* + 20* = 53*

[cineman]

Hello spittybug, maybe this is interesting for you.
This are the maps of the original eprom bin of a renix ecu injection for a Renault PRV 2849cc, even firing version, about 155hp. Map of timing and VE.
Like you can see timing goes to about 46deg in this oem factory maps, but like i told you there are sport maps wich goes even to 60deg... for modified engine.
Interesting to see also this line of where the stock engine has the torque ( around 2700rpm ) with greater VE numbers... ;)

8736

[Spittybug]

I tore open a distributor today and made it ready for the next EFI project. I took pictures and documented the whole process here: http://dmctalk.org/showthread.php?2056-Hardware-amp-Integration&p=45242&viewfull=1#post45242.

I also answered my own question. The spring/weight combo advances the rotor shaft inside the distributor up to 20*. The vacuum advance unit, when under vacuum, rotates the whole impulse coil & plate assembly up to 20* in the opposite direction, so it is independently advancing (moving towards the approaching rotor). So, if you are at high RPM with the weights flying outward advancing the rotor and at decel or light cruising load (high engine vacuum) then the advance could indeed be at or close to 20* + 20* + static ~13* = 53*.

The switch on the stock throttle assembly activates a solenoid to block the vacuum from reaching the advance unit at idle. The second you come off idle (where the vacuum in the manifold is at its maximum), the solenoid opens up and allows the vacuum to reach the distributor advance. This means that instantly when you put your foot down to accelerate off idle you go from 13* static advance to full vacuum advance plus static, or 33* !!!!! WOW, big jump. As your RPMs go up, you start merging in mechanical advance too, but by now your manifold vacuum has dropped as you've let air in through the butterflies. Interesting as hell to model and build a 12x12 table of values for. I'm pretty good in all areas except right off of idle where I'm bogging. Time for some more trial and error. No wrench required.

[AdmiralSenn]

All just in time for me to buy a spare distributor to go full spark by Spring Break.

[opethmike]

I expect that in another week or two that my VE table should be close to dialed in, at which point I will share it. It should be good for most Spec 2 or Stage 2 engines.

[Spittybug]

Hello spittybug, maybe this is interesting for you.
This are the maps of the original eprom bin of a renix ecu injection for a Renault PRV 2849cc, even firing version, about 155hp. Map of timing and VE.
Like you can see timing goes to about 46deg in this oem factory maps, but like i told you there are sport maps wich goes even to 60deg... for modified engine.
Interesting to see also this line of where the stock engine has the torque ( around 2700rpm ) with greater VE numbers... ;)

8736

Looking at your advance table, I don't see a lot of load based (vacuum) advance. In the idle area you have +/- 5* changes, doesn't that give you hunting? VE numbers change a lot in the idle area too. When you hit the throttle and go atmospheric, our stock distributor adds in 20* of advance, to the mid 30s. Your's only goes to the mid 20s. I'm wondering why? Generally speaking, I think your advance table looks a lot like what I'd expect, just lower on the vacuum. How does it behave? Well? Was it determined on a dyno or by feel? Do you have torque values to go along with it?

[cineman]

Looking at your advance table, I don't see a lot of load based (vacuum) advance. In the idle area you have +/- 5* changes, doesn't that give you hunting? VE numbers change a lot in the idle area too. When you hit the throttle and go atmospheric, our stock distributor adds in 20* of advance, to the mid 30s. Your's only goes to the mid 20s. I'm wondering why? Generally speaking, I think your advance table looks a lot like what I'd expect, just lower on the vacuum. How does it behave? Well? Was it determined on a dyno or by feel? Do you have torque values to go along with it?

That is not MY table, is a direct reading of a factory Renault OEM 2.8 PRV eprom injection, so it's surely roadly and thinkered ;) and with miles and testing behind it.
I did not tested that maps directly on a delorean engine, but i drove a car with that PRV and maps and it was normal and good...

[opethmike]

Attached is my tune for my Spec 2 EFI build, after 300 miles of driving and auto-tuning. Idle bounces between 52 and 56 Kpa; expected considering the large cams.

This tune should be good for an engine build similar to the following:

Ported & polished heads
Aggressive camshafts
Spec 1/stage 1 exhaust

The spec 2 package is more aggressive than the stage 2, so it may not be an optimal tune for stage 2.

[Spittybug]

Mike, I notice your acceleration enhancement goes from 1ms to 6ms. Do you find this to work well? No jerking, bogging or otherwise negative performance? I find this the hardest thing to dial in. I'm using MAP not TPS since the TPS is noisy and give unintentional acceleration boost.

[opethmike]

Mike, I notice your acceleration enhancement goes from 1ms to 6ms. Do you find this to work well? No jerking, bogging or otherwise negative performance? I find this the hardest thing to dial in. I'm using MAP not TPS since the TPS is noisy and give unintentional acceleration boost.

I haven't noticed any undesirable effects from my AE choices. I do tend to drive my car pretty aggressively.

I can't really use MAP very well, as the bigger cams don't cooperate as well there as using TPS based does.

The most noticeable difference with my new engine build (as far as tuning goes) is the idle. The MAP sensor reads a higher and more varied KPa. Have to run a little richer at idle too, to keep the cams happy.

[opethmike]

I have had to do a pretty significant re-work of my cranking pulsewidths, afterstart enrichment, and warm-up enrichment in light of my significantly souped up engine.

With the values in the attached tune, the car starts nicely after very little cranking.

Then, ASE remains on for ~20-30 seconds, and the AFR runs at 11-12 during this time. Once the ASE switches off, and the WUE is still going, the AFR will run at 12-13.5, until fully warm. Once the car is fully warm, the AFR will fluctuate between 13.5 and 14.2.

With the idle speed set at ~750, the KPa will flux between 52 and 60 at idle. This about right for cams of this size.

This tune should be pretty much spot on for a DPI Spec 2 build, or a good starting point for a DMC-H Stage 2. The stage 2 is noticeably less aggressive, so my values are likely too rich for it, but would be a good starting point.

9479

[Spittybug]

Thanks Mike; getting everyone figuring out the nuances of tuning will really help.

Our proto car #2 has the DMCH stage 2 cams and he had a hell of a time getting it to start using my tune as a starting point. I haven't seen his tables yet, but he told me he had to cut down the fuel on the cranking and up it on the VE table. Ian claims it is now idling well and he has driven it (fixed distributor timing!). I have the two coil driver chips from DIYautotune to put into both of the MS units this week. At that point we'll have both of them full fuel & spark control.

[opethmike]

I would be curious to hear what is idle KPa is at. Should be a good judge of just how different the cuts of camshaft are.

[dmc6960]

My idle KPa is around 68. This is with the Z7U intake and throttle, DMCEurope cams, high compression pistons and a 1mm decked head.

[opethmike]

Since the cams would have the biggest effect on KPa, I would be very interested in knowing what the actual specs are on each of the different vendor cams. Drives me nuts that they don't publish those numbers.

Oh, I should mention that for the tunes I have uploaded, that on top of the heads/cams/exhaust, the timing has also been advanced to 18 degrees.

[Spittybug]

IIRC, Ian's is idling at about 60-65 kPa, but we may be able to lower that a bit once the sync loss issue is behind us. I've got mine down to around 34 and the MAP lag set to 90, so as soon as I hit the gas I get response. I run my idle in the ~850 or so range where it is nice and smoooooooohth.

[opethmike]

Mine fluctuates between 52 and 58, and my cams are much more aggressive than Houston's. While I had my base ignition timing set to 15 degrees, I was up a few points higher. I advanced my base timing to 19 degrees, and perfected my idle, and I am now down to 52-58.

Keep in mind that unless you bump up the idle quite a bit, big cams are going to be less vacuum and rougher idle. My idle kPA won't hit 40 unless I idle at ~1,300 RPM. No thanks! (Plus, then I lose the bad-ass lope!)

[Spittybug]

I shared this somewhere else, but it probably belongs here in this section. Under BASIC, General Lags, find the MAP averaging lag factor. The closer you get this to 100 the more responsive the RPM climb will be to opening the butterflies. Default is something like 50 which makes the car boggy when starting from a stop. I've got mine at 90 and it makes a world of difference. Just one of those things that isn't particularly well documented in the manuals I'm afraid.

[Spittybug]

I've been doing even more tinkering to get the most I can out of my car. It is pretty cool to just do it with a laptop and not even pop open the engine compartment. :headbang:

I got my VE (fueling table), AFR target and spark advance tables to the point where I was running really, really nice. Checking logs I would see that there were times (hard accel or decel) where the resulting AFRs (air/fuel ratios) would go rich or lean by a few % for very short periods. The optional EAE (enhanced acceleration enrichment) algorithms in MSIIe try to eliminate these transitional events by modeling the fuel stuck or or sucked from the walls of the manifold that results in an actual amount of fueling that is different from what has been commanded. Carburetors have similar issues. I tried using those curves and while I found I could get the AFR variation to virtually disappear in steady state and most transition states, I was having a hard time with a super lean spike when lifting off the gas (like gear shift or coming to a stop light). I tried and tried to modify the curves, but all I seemed to be doing was introducing oscillation (PW going up a bit, down a bit as the "puddle" equalized). I basically gave up, at least for a while.

Enter the code variant by gslender. He has added some nice features to the already impressive set. One of them is an TPS (throttle position sensor) smoothing window. It gets rid of a lot of sensor noise by oversampling. I loaded it up and have now gone to using the traditional AE (acceleration enhancement) based 100% on TPS. No herky jerky unintended AE kicking in because of TPSdot (rate of change of throttle position) noise. I have the lower threshold at 50 and a nice shallow curve adding PW (pulse width) from 50 to 300 %/sec rates. I think the most I add is only like 2ms of PW, but this has made the car run really, really nicely. Combined with the MAP (manifold air pressure) lag of 85 -90 or so and I have no hesitation off idle and when I put my foot down I get a nice little kick. Looking at my resulting log files shows me that while I sometimes run a little rich in those transitions (better for power), I don't go lean (knock or engine damage). His code also has some closed loop idle improvements I'm going to start playing with to make it even more solid than it already is. Once warm I'm rock solid at idle, but during warmup I do still get some oscillation as the VE values, idle air valve closing, temperature rising and timing all equalize to run state. Apparently he's got the means to wire your A/C compressor circuit into the MS unit so that it immediately signals that load is about to increase and the idle therefore changes at the same time; no lag. THAT would be cool. The code has improved battery voltage correction and other niceties that I'm exploring.

There is also some good new writeups that have pointed out that some of the wiring diagrams previously released by vendors like DIY can be improved upon to reduce sensor noise. I haven't yet done this to my install, but will be doing so immediately. Whereas it was previously recommended to ground the sensors to the same spot on the engine as MS grounds to, it has been found that doing so introduces noise to them. The intake air, coolant and TPS sensors should all have their grounds go back directly to the MS unit which in turn is grounded to the engine. Apparently the 5v reference voltage that drives the sensors varies up and down to be a constant 5v over the MS ground plane. If your sensors go straight to engine ground, that changing 5v is introducing noise! The way to check if you are correct or not is to unplug your 37 pin connector from MS. If your sensors are still common to engine ground, you have it wrong and are introducing noise on your sensor wires. I've seen before/after graphics of the noise and it is significant enough to reduce/eliminate the little PW changes that make the RPM and AFRs jiggle.

Hope the above is helpful to those of us already enjoying the benefits of EFI and to those thinking about it. We EFI guys sometimes get smacked for "look at all the problems they have" or "look how long it takes" or "they're always having to play with it"...... Let's set the record straight; we get TONS of diagnostic data from our systems that lets us optimize conditions, continuously improve performance, minimize emissions and most importantly, UNDERSTAND AND CONTROL exactly what our engines are doing and what they could be doing better. It's like chess; a minute to learn but a lifetime to master.

I don't recommend EFI if you are content with the status quo, not interested in improvements or are looking for a drop in solution of someone else's making. Drilling out broken studs or hunting down electrical problems sucks, but this stuff is fun!

[opethmike]

Agreed Owen. I'm often tinkering in some way with my EFI, but it is constantly improving, and I enjoy the process. It is definitely for the tinkering type, or the type that likes to learn. So glad I did my conversion!

[Spittybug]

NEWSFLASH....

Tuner Studio has a new version out. It has a new feature, WUE auto tuning! WOOOO HOOOOOO!!!!! So, just like you can tune the VE table by letting the VE auto tune do it, now too can you dial in your Warm Up Enrichment.

I've used it twice and in each case the base temp was only 70*, so I'm waiting for a cold morning to go further back up the curve, but already I'm seeing a big improvement. I've all but eliminated the dreaded idle hunt or surging that is common until warm. It's easy to use and fun to watch it dynamically play with your WUE values as the car warms up. The key seems to be in making sure your VE table is is good shape to begin with (not too hard, right?) and then on your first start up, try to keep the RPMs fairly steady by using the gas pedal. Then on the second warm up run you won't need to do that and he improves it even more. I'll update this thread with more passes.

Remember, the actual RPM level during idle is set by the idle control valve settings; how far open and how steep the curve to close it with increasing temperature. I think my IAC settings are pretty good, so with this WUE tuning I should end up with a nice, smooth warm up. At temperature mine is already rock solid.

I also ordered the bluetooth transmitter so that I won't need my cable to connect my laptop anymore. $59 from the EFIanalytics site. Yup, sit in the lounge chair and tune the car.... ah, technology.

[opethmike]

I just downloaded that beta last night, with my main interest being the same as yours - the WUE auto-tune. Mine is decent right now, but won't argue with getting it better. It is in the 40s up here right now, so I should be able to go through the whole range over the next few days.

[Spittybug]

Should have posted this earlier....I got the closed loop idle function working nicely and have also modified the timing to pull out the 20* of vacuum advance when the throttle is closed (like stock). When the A/C kicks on the idle recovers nearly instantly.

The *CODE* you have to break with Megasquirt is beginning to become clear;



AIR - you must get your idle air right first. This controls your idle speed. All else comes after. The goal is to be able to totally close the throttle plates at idle and let the valve do all of the air control at idle.
SPARK - get the timing close. The exact shape of the curve is negotiable, but this is second in importance only to air.
FUEL - the VE table can be autotuned, but then smooth it. Especially around the idle area, keeping the values constant or very close is key to a smooth idle. Let the AFR bounce a bit in idle area, but it will keep the RPMs more constant and you'll get a nicer idle. The autotune goal seeks to whatever you have entered in your AFR table, so make sure it is reasonable. A little rich at idle and WOT, a little lean at cruise.
WARM UP - once all of the above is working, use the new warm up tuner. It will get you really close.
AE - this is ripe for a new autotune feature. Right now it is trial and error to get the acceleration enhancement dialed it. No particular hints or shortcuts.

[dn010]

Is anyone running a B280F that could post an msq? I'd like to compare some values particularly VE table...