The Mega/Microsquirt Mega Thread, There's enough of us, time to share info Options

[GregAmy]

Have we decided on which is our official Mega/MicroSquirt thread yet? If so then merge this into it.

I recently took the Microsquirt 914 on a 5-hour each way journey from Connecticut to New Jersey Motorsports Park. I was a bit hesitant to substitute my comfy radio-A/C-and-DSG-equipped GTI for a 45-yr-old ride through the wilds (and traffic) of NYC, but the weather was forecast to be a nice Fall day so I went for it.

Loved it. Car ran flawlessly and the engine was a smooooooooooth as I've ever experienced it. No stumbles, no faults, ran cool on CHT and oil, just an absolute pleasure. 27.5 mpg average on 89 pump gas.

But of course, with no radio (by choice) I had a lot of time to think about my tuning as I was tootling down the parkways and interstates, watching the Android tablet display of engine parameters...it runs good, and it sounds good, and it feels good, but...is it actually good?

One of the biggest issues I faced/am facing with tuning is not the mechanics of the process; I quickly figured out how to use TunerStudio and Megalog, and I'm logging to either laptop or an Android tablet (ShadowTuner) via Bluetooth. And the VE Analyze function of Megalog makes it quite easy to review the AFR outputs in the logs and make changes to meet my AFR goals.

No, that's the easy part. The hard part is knowing where our engines want to be, for example in terms of AFR and ignition timing.

"Take it to the dyno!" you'll say. And you'd absolutely be right. But there's two problems with that.

First, dyno time ain't cheap if you want to do it right. It's not just a matter of showing up and tuning the VE tables to meet the AFR goals (and I can do that on the street). No, to do it right you need to spend an afternoon, or even a day, with a skilled tuner who can run the engine through a series of tests and tickle the edges safely to discover things.

The second factor relates directly to the first: finding a skilled tuner that not only understands the concepts of dyno tuning but also understands the needs and desires of a 75-year-old air-cooled engine design. Sure, the basics are the same as other engines, but I suggest the edges are much much fuzzier.

If I had someone with that talent in my area I'd absolutely leverage them. But I don't. I have some local guys that are good with their dynos but they're Honda guys, or Chevy guys; When I bring my 914 to them (and I have) they're glad to take my couple hours' money but other than tuning to my settings they're not really qualified to tickle (and if they're smart, they're not comfortable trying to tickle the unknown edges on a customer's car).

So any dyno work I've done was to my AFR and ignition settings. And I really don't know what I'm doing either.

I don't know what air-cooled dyno tuning experience there may be on this board, but I'd love to begin sharing some info on where we're going with these Micro/Mega installations. Where do these engines want to be? How much can (or should) we push them? What symptom(s) are we looking for to decide if we need more or less of anything? What's the sweet - and safe - spots?

I'm atttaching my AFR target and ignition tables, let me know what you think. Am I in the ballpark?

I'm not asking you for any secrets. I'm just looking for some feedback to understand if I'm setting myself up for problems in the future. This is not my race car, I'm not looking for that nth hp advance; this is my street car, one that gets driven only 3000 miles a year (if I'm lucky, especially lately). I want to enjoy it and not worry about it. And maybe I'll learn something in the process.

And...discuss! - GA

[bbrock]

Thanks for starting this thread! It's almost like you read my mind. As some of you know, I'm planning to megasquirt over the winter and think I have a pretty good plan but am in a bit of a holding pattern to finish up a few house projects and also for Mario to reopen his online store so I can buy some sensors. In the meantime, I've been doing the most dangerous thing which is watching youtube vids about tuning and reading this book by the DIYAutotune folks.

(IMG:http://www.914world.com/bbs2/uploads_offsite/images-na.ssl-images-amazon.com-20845-1636250753.1.jpg)

I've come out with the same thoughts as Greg that I don't see a dyno being an option. I have only found one car dyno in the area (there are a few motorcycle dynos) and this is Montana so you can probably guess what kind of cars (hint - trucks) they are versed at tuning. Even the area Euro car specialist mechanic has stopped working on aircooled cars even though they are Porsche guys and one of them has a 914. Also, while that book outlines what seems like a pretty safe procedure for tickling the edges with a dyno, it's all geared toward water cooled engines so not very helpful.

Also, I've read through this thread about AFR several times learning about what is safe for an air cooled Type 4. http://www.914world.com/bbs2/index.php?showtopic=304320. In particular, I'll be taking @Mark Henry 's advice to stay on the fat side to keep CHT down.

To that end, I thought I'd start with an ignition table pretty much like Rob's which looks to be close to a stock curve just to be safe. I still have a lot to learn about when it is safe to run more advance, and what you gain so am enjoying the discussion on that. I was also a bit surprised to see areas going to stoich and even leaner on both of your tables - again, thinking about CHT and being a total ignoramous on the subject myself.

I mentioned this on my megasquirt thread but it seems like a closed loop fattening the AFR or dialing back advance when CHT starts to climb beyone spec would be ideal, but the only out of the box solution I've found is using alternate AFR and Ignition tables triggered by CHT as Greg mentioned. A loop seems like it would be a good way to tickle the edge from the efficiency side at least by building finding the edge into the algorithm. I am not shy of diving in and writing some custom code, but not when a bug could cost me my engine.

As usual, I have only questions and no answers but am bookmarking this thread as it is surely going to be a great resource as I move through my project. Oh wait, I do have one bit of advice, Hey, Greg, you should take 3 out of row 2 column 11. (IMG:style_emoticons/default/lol-2.gif)

[BeatNavy]

To that end, I thought I'd start with an ignition table pretty much like Rob's which looks to be close to a stock curve just to be safe.

Well Brent, I just made mine a little more aggressive last night after reading this thread (IMG:style_emoticons/default/smile.gif) Reminded me that I solved my hard acceleration pre-detonation with fuel rather than timing but forgot to put it back the way it was. Haven't uploaded to the ECU yet, though.

I've got a CHT that's pretty prominent in my console, and my head temps are surprisingly cool, all the time. I don't get above 300 very often. It may be the ceramic coating on the combustion chambers, the general engine combo, or something else entirely. I guess I'm also aware that the CHT may not be telling me everything. Oil temps are a different story, however. They are manageable, mostly, but not necessarily ideal (IMG:style_emoticons/default/dry.gif) In hot weather it sometimes changes my driving style. For example, I may violate the cardinal rule of air cooled and cruise at lower RPM's, e.g., under 3K. Seems to keep oil temps down without bothering head temps too much.

Again, my tuning "pain point" other than cold starts is fuel efficiency -- I suspect I'm not running as efficiently as I could be based on MPG and a couple of other indicators. The nice thing about these systems, of course, is you can have multiple tunes. Want a max efficiency tune for a long drive? Load that version into the ECU. Want a really fat tune for performance at the next AX? Boom, load that one in. When you're done, put the efficiency tune back on for the ride home.

I must admit I'm not 100% sure of the relationship between the AFR and VE table. I probably should do some research on the forums, but here's what it says in Tuner Studio:

Including the AFR target with the speed-density, alpha-N, ITB or map/baro algorithms, this allows the VE table to be a 'real' VE table and the desired AFR to be specified in the AFR table.
With AFR not included, the AFR table is for reference only and the 'VE' table takes full control.

With the MAF algorithm, this MUST be enabled to factor the target AFR into the fuel equation.

Note that this function does not require or use feedback from an O2 sensor.

That last sentence has me a bit confused. Does that only apply to MAF? (For reference, I'm running speed density. I updated my first post above to reflect that, as that's kind of important to know.)

More info. Here's an annotated AFR table. I spend 99% of driving time in one of these circles (very roughly). I initially struggled with idle, as mine has a higher pressure (kPa) (lower vacuum) than almost anybody else I referenced, air or water cooled. I don't know if it's the Raby cam or something else. After a while I stopped fighting it and "gave the engine what it wanted." Any AFR numbers way outside of those circles, or when moving between them during driving, are basically irrelevant.

[JamesM]

I must admit I'm not 100% sure of the relationship between the AFR and VE table. I probably should do some research on the forums, but here's what it says in Tuner Studio:

Including the AFR target with the speed-density, alpha-N, ITB or map/baro algorithms, this allows the VE table to be a 'real' VE table and the desired AFR to be specified in the AFR table.
With AFR not included, the AFR table is for reference only and the 'VE' table takes full control.

With the MAF algorithm, this MUST be enabled to factor the target AFR into the fuel equation.

Note that this function does not require or use feedback from an O2 sensor.

That last sentence has me a bit confused. Does that only apply to MAF? (For reference, I'm running speed density. I updated my first post above to reflect that, as that's kind of important to know.)

So if you are talking about the setting I think you are talking about, the difference between the "incorporate AFR target" setting being enabled or disabled is best explained by showing exactly what it is doing

The basic fuel equation is:
PW = ReqFuel * VE(rpm,map) * MAP * GammaE

With incorporate AFR it becomes:
PW = ReqFuel * VE(rpm,map) * MAP * stoich / targetAFR(rpm,map) * GammaE

So... thats the math behind it, but why? Basically the idea behind it is that IF you had your VE tabled tuned so that they produced exactly the AFRs represented in your AFR table you could then enable the "incorporate AFR target" setting, at which point you could in theory just update the AFR values on the AFR table to change your tune.
...in theory

its a feature that requires a lot of prep work in that you have to have your VE table tuned to your AFR table perfectly before even enabling it for its function to make any sense. You absolutely dont want it enabled when initially tuning your VE table or you will cause yourself unnecessary headaches.

I just leave "incorporate AFR targets" disabled all the time.

That is not the only function of the AFR target table though as it is also used for the closed loop wideband O2 correction(if enabled) and with tunerstudios autotune.

[BeatNavy]

So if you are talking about the setting I think you are talking about, the difference between the "incorporate AFR target" setting being enabled or disabled is best explained by showing exactly what it is doing

The basic fuel equation is:
PW = ReqFuel * VE(rpm,map) * MAP * GammaE

With incorporate AFR it becomes:
PW = ReqFuel * VE(rpm,map) * MAP * stoich / targetAFR(rpm,map) * GammaE

So... thats the math behind it, but why? Basically the idea behind it is that IF you had your VE tabled tuned so that they produced exactly the AFRs represented in your AFR table you could then enable the "incorporate AFR target" setting, at which point you could in theory just update the AFR values on the AFR table to change your tune.
...in theory

I just leave "incorporate AFR targets" disabled all the time.

That is not the only function of the AFR target table though as it is also used for the closed loop wideband O2 correction(if enabled) and with tunerstudios autotune.

James, thanks for helping out and sharing so much experience with us. It's amazing how deep this can get, and it does seem one can spend years trying to learn it, and many minutes hours trying to explain it. Uh, I'm also glad to see, based on your response, that at least I didn't completely step in it (IMG:style_emoticons/default/biggrin.gif)

That equation makes perfect sense. I had to look up "GammaE," but I guess it's the combination of any other enrichment factors currently enabled, correct (e.g., warmup, accel enrichment, etc.)? So really ALL the AFR table is doing is changing the pulse width by the factor of dividing stoich (14.7 for pump gas) by the target AFR (e.g., 13.9). So in this example pulse width would be modified by a factor of 14.7/13.9, or roughly 1.05, all other enrichments being the same.

Curious why you wouldn't use it? Is it because you are more comfortable with manipulating the VE table directly at this point?

Greg, I incorporated fuel cutoff similar to your specs and took her out for a spin yesterday. I like the change. We'll see if it increases efficiency, but I'm going to run with it for a while. I do recall that D-Jet incorporated a fuel cutoff logic on one or more model years (maybe '73?), but then discontinued it. IIRC the thinking was that it somehow impacted emissions negatively by introducing more variation into piston temps. Or something.

I also bumped my total advance back up toward 33 degrees, but guess what, I started getting pre-ignition under hard acceleration again. So I backed it down to 30, and it seems better there.

Here's the actual Megalog histogram (that I've never really used before) of my drive. Darker green is where I spend most of the time, I guess.

[JamesM]

That equation makes perfect sense. I had to look up "GammaE," but I guess it's the combination of any other enrichment factors currently enabled, correct (e.g., warmup, accel enrichment, etc.)? So really ALL the AFR table is doing is changing the pulse width by the factor of dividing stoich (14.7 for pump gas) by the target AFR (e.g., 13.9). So in this example pulse width would be modified by a factor of 14.7/13.9, or roughly 1.05, all other enrichments being the same.

Curious why you wouldn't use it? Is it because you are more comfortable with manipulating the VE table directly at this point?

A few reasons im not really interested in it at the moment

In my mind its just a feature that adds some simplicity by adding complexity and an additional factor to the fuel calculation. Also for it to make ANY sense your VE map has to be a dead on match to your target AFRs BEFORE enabling the feature, and for various reasons im not 100% sure I trust my O2 readings in all cases.

I think some of the bigger issues for me is that once enabled your VE table is no longer an accurate representation of the fuel going into your motor as it has a factoring table then calculated on top of it. It prohibits you from editing you VE table directly without royally screwing things up, which then means you can no longer run autotune either. I dont necessarily have anything against the feature, but its a feature you only turn on once you have everything dialed in as far as hitting your AFR targets and then its only real use is experimenting with changes to AFRs. I already know what I want my target AFRs to be when im dialing in the VE table to hit them and even then, if I want to adjust them after the fact its not a big deal, so yeah, just not worth the effort to me. VE table is easy enough to use, you want more fuel in a specific area, increase the number, want less, decrease it, then take some datalogs and see where you are at. I could never trust a factor is going to get my fueling exactly where I want it so im going to be checking the logs either way.