MyE28.com DEV

I tuned the pots on my ms2 for a more solid vr signal.
read here
http://www.megamanual.com/ms2/vradjust.htm
good luck

Thanks Marc, that's helpful. I decided to change the lag factors in the megasquirt to try to isolate which signal is the culprit. Upon closer inspection I think my AFR could have the most noise. So I will be messing around with it in the next week.

One of the biggest differences with this engine is that the oil pressure is higher than before. Where as the old engine(s) had hot idle of about 7-10psi, this engine sits closer to 15 and that makes me feel a bit better. I don't know if this oil pressure could have caused a piston ring issues but I'm sure that was the problem with the other engine that I pulled.

I decided after talking to my brother that it may be worth it to buy the full version of TunerStudio. It was $10 off and I decided to go for it. I had a fair base tune on the old engine and it drove ok but right away I started with the Tune Analyze auto tuning. It is interesting to see the suggestions of the new fuel values.

I went for a short drive and immediately saw the difference in optimized AFR's. After a short 30 minute drive up and down the highway, I slowly opened up the car as I saw the target
AFR's were more in reach. I have to say, I am very impressed with the program and would recommend it to anybody who has Megasquirt. The performance is night and day compared to just a day earlier. I don't say this kind of thing very often. Boost comes on smooth and I am running about 6.5-7psi.


This morning I did a cold start and ran the analyzer to find that at cold start, idle, and then through the revs after the engine warmed up that the whole tune was too rich. It wants to take out fuel everywhere. Does anybody else have this problem? Here's the adjusted map after what I thought was a close base tune yesterday:



Here's a short data log pull from the same drive this morning.



I am going to try it again in the warm afternoon temps and see if there is a difference.

With additional smoothing on AFR signal, here's another one:

Looks great. I like seeing the engine swap pictures even though they sort of dropped off at the end.

I'm not surprised youre seeing so much fuel being pulled out, you only got like 20-21 mpg after the cruise to d4c which is pretty bad when you account for the downhills that should have helped the fuel economy. You can go into the 'Advanced Settings' in the autotune and adjust your 02 sensor feedback delay based on your datalog. I would look at when you let off the throttle after your pull and see how much time passes from when the TPS starts to go to zero and when the oxygen sensor AFR reading goes to 18+ to signify it is in fuel cut or remove fuel. That will ensure the oxygen sensor is reading the actual time it takes the sensor to read the inputs from your throttle and dial in the fueling to be measured correctly based on how far the sensor is down the exhaust system.

Another thing I would do is change your base fuel (Req_fuel) in the basic fuel settings. If you adjust it correctly, it will provide less fueling and you can increase the values on your fuel VE table to be increased. In effect you are increasing the resolution of the fueling table so you can fine tune each zone of the map. For example, at idle when you are (currently) adding 23 fuel, an additional fuel value of 1 point, to go to 24 is (1/23) or 4.3% more fuel. If your Req_fuel is currently set at 10.0 and you reduce it to say, 7.0 (a 30% reduction), your base fueling table will need to increase 30% to provide the same fueling across the entire table. Now that 23 will be 29.9 (23 x 1.3=29.9), or 30. Now when you idle, if you add one point of fuel (31) you are adding only 3.3% more fuel (3/4 as much fuel as before) which will help you idle closer to your target and dial in the entire table to a closer to optimal value.

The same is true at 120 kpa and 2900 RPM where you are at 78 currently, a 30% increase of Req_fuel puts you at 101.4. and now each additional adjustment point adds less than 1% more fuel. This allows very fine adjustments of the fueling. I have targeted about 100 for my maximum when I am at 6 psi boost, but there is no reason your couldnt specify a lower Req_fuel and get better resolution as your VE values increase. You will have to redo the entire map which is why I didn't recommend it before you had Autotune, but this should get your fuel really close without any EGO control. Also make sure that is not enabled while you are tuning.

Some more info on Req_fuel equations. http://www.megamanual.com/v22manual/mfuel.htm
and the equations used: http://www.megamanual.com/v22manual/ve_tuner.pdf

What does your Target AFR table look like and your ignition timing table look like now?
It may may sense to verify base timing with a timing light since the m30 info varies between sources when I have looked.
I also have been looking at doing dual maps to increase ignition and fueling resolution across the map. Here is an example from the r3vlimited timing thread. Nando's 325ix stroker engine running Megasquirt. The dual tables allow for much more higher resolution of your timing and fueling. I have yet to do this but may try it soon once my tune is dialed in some more on the 16x16 table.

That huge table is ridiculous

The increased resolution does make sense, but I hadn't heard of it before. Did you feel like this was necessary, or did you notice improved target AFR's?

Here's my current tables which are slightly different from before. Useful for later reference:

VE:


IGN:


Target AFR:


In colder 40F weather yesterday, the AFR's all dropped to 11-3:1 rich in normal driving. I am not sure why it did this, but I suspect MAT or barometric corrections. I am going to mess with the MAT correction (to default) and barometric settings to see if I can work out when I hit 11.0 AFR's in colder temps.

One thing to note. I mentioned I had higher oil pressure with the new engine. This was true, but with my 15W50 oil, I was getting clouds of smoke after shifting and even after idling in 40F weather. I thought it was attributed to the higher pressure and flow of oil through the turbo (that my drain couldn't accommodate). So I ended up reinstalling the .090" turbo oil restrictor fitting to see what would happen. So far I haven't noticed any more smoke but will continue to monitor it.

I also replaced the passenger side headlight that got ruined at D4C 2016:


I also installed some new Recaro seats from CL.



My door hinge pin broke so I need to replace that soon!


After additional testing I have been stuck at the same basic problem.

I go out and drive the car and tune it with auto tune in the afternoon 70F weather only to find that the next morning the entire curve is too rich and autotune wants to remove fuel in significant amounts. Some say injector dead time may be off, so I found a procedure and ran the analysis in excel to find my deadtime should be .889ms with these injectors-- very close to where I already was.

I did some reading and have found it to be a problem from a lot of people using MS2. Simply put, they argue that air is not an ideal gas and the default curve is way off from what is actually needed for fueling corrections. Instead of removing fuel when the engine intake temps warm up, the MAT curve needs to be corrected to match the needs of the specific engine setup.



https://www.miataturbo.net/megasquirt-1 ... 176/page3/

Basically this suggested curve is the inverse of the ideal gas law and default correction. Many suggest that fuel should be added or removed (based on temp) as needed to get the engine to run as close to target values as possible.

One way to do this I suspect is to start with a target autotune map and record ambient temp during tuning with autotune. Get the target map happy everywhere. Then enable EGO control and drive normally in cold start situations. Take note of temperatures in the log files and also look at the EGO correction. Simply adjust the MAT correction chart up or down using these parameters until the target AFR's are close and EGO control is only working about 5%.

I am going to try this and report back.

Im interested in your results. I'm also have some issues with hot idle consistency when the motor and intercooler are heat soaked.

Still messing with MAT and don't have anything new to report since I started a new job this week.

But had some time so sit down and do some research which I wanted to document regarding turbo sizing. I hadn't really had a handle on what the specs mean with wheel sizes, AR and compressor vs turbine, inducer vs exducer on each side, but I think I do now. Here are some of the findings relevant to my current and past turbo.

Understanding what different manufacturers reference for their part number in relation to either exducer or inducer helped my understanding a lot:


With some research in the past I had written down the original Rotomaster turbo compressor specs as measured by another mye28 member as 48mm inducer, but not sure about turbine exducer. In cross referencing flow charts of other turbos, it was sized for max air flow of about 45 lb/min.
I already knew it was undersized for M30B34 but in running at ~11psi I was at roughly 35lb/min airflow and according to this chart for what I think is a close compressor map, I was getting past the ideal operating conditions and efficiency island. Not terrible, but I imagine the old technology parts didn't help and a rebuild would have made it a reasonable choice.

Estimated originally with 385 CFM and a Pr of ~1.8.


So in my research and paperwork, my TCD turbo with P trim with a Garrett '60' compressor wheel. This translates into and a Turbine wheel/Exducer of ~65mm and Turbine A/R of .58.
And as quoted by Todd:

A Garrett journal bearing cartridge with an oil cooled and lubricated bearing.
A 60mm compressor wheel.
A Comp TurboT04e compressor cover and backing plate.
A Garrett .58 on center exhasut housing.

My brother shared the following information which shows the placement in relation to the standard offering, and Todd has stated his unit is a custom variant from Comp.

Product# Turbo Compressor Wheel/Inducer Turbine Wheel/Exducer HP Rating

325558 CT3S 55 58 570
325858 CT3S 58 58 600
326062 CT3S 60 62 660
326262 CT3S 62 62 710
326265 CT3S 62 65 710
326465 CT3S 64 65 800
326767 CT3S 67 67 930

He also created a great thread for turbo comparison and estimated max flow here: http://www.r3vlimited.com/board/showthread.php?t=399047

Which gave me a list of other comparable turbos (sorted by max flow):

Brand..................Turbo...............CmpInd.....CmpExd.....TrbInd.....TrbExd.....Flow(lb/min)

BorgWarner.......EFR 7163....................57..............71..............63......N/A.........60
Garrett............GT 3582R...................61.4............82...........68..........62.2.......60
Holset............HX40 - 6 blade...............56..............86...........76..........65..........60
MHI...............TD06H 60-1..................59.5............76...........67.........58.8.........61
Holset............HX40 - 8 blade...............60..............83...........76.........65...........62
Holset.............HX40 - 7 blade..............60..............86...........76..........65..........62
BorgWarner.......EFR 7670....................57.2...........76............70...........62..........64
Garrett............GTX 3076R..................58.............76.6..........60..........54.9........64
Precision..........PTE 5862....................58.............76.2..........71..........61.9........64
Blouch.............Dom 5.0 XT-R...............62..............84...........68..........62.2........65
Forced Perf.......FP Red......................61.21..........81.79.........67..........59..........65
Garrett............GTX 3576R..................58..............76.6.........68..........62.2........65
BorgWarner.......S257SXE....................57.15...........76.2.........70...........61..........65
CompT..TCD 60/P Trim......58......74.9......74.2.....64.6.....? *Garrett 60 wheel(442476-0017): 6 blades, tip height of 6.1mm.
Holset.............HE351CW-9cm..............60..............82...........68...........60..........67
Garrett............GT 4088R..................63.5.............88...........77.........70.4.........67.5
BorgWarner.......BW 8375-S360...............60..............83.........75.9..........68..........68

I wanted to get an idea of where I was at on the compressor map in terms of efficiency for later reference.

So plotting a few total guesses into Matchbot by Borgwarner and plotting the most similar sized Borgwarner I could find from the list, I am able to estimate the compressor map at 11psi. If I enter 6psi (like I am currently running), I barely touch the lower edge visible on the map. This is data point 2.

Slightly smaller - BW S257SXE


Slightly bigger - BW S360- 8375 (i think)


This is a good starting point, but I am wondering if I can get a compressor map from CompTurbo.

great info! I wish some of those with knowledge would share their thoughts on what's the best turbo for street.

Have you taken a look at your estimated boost level/ pressure ratio? M30B34 vs B35 will be different slightly.

Plotting out your intended Pr and airflow on individual compressor maps will give you a good idea of what makes sense in your application. Then you can determine if different compressors would be in surge or in high efficiency islands or not. From there compare pricing.

The list above would be a good starting point though.

Hey guys,

I have spent some time tuning the car, but my tuning laptop hard drive failed and my driver's door hinge pin broke (literally sheared) so I didn't get as far as expected. I got the laptop hard drive replaced and spent a little time tuning again this afternoon. Hope to replace the door hinge tomorrow or so-- it totally failed and can't really be opened.

I downloaded a program which has positive reviews called Virtual Dyno which supposedly lets you get an idea how close you to the real thing after defining specs on your car. So I entered in what I found to be my gear ratios, cD, estimated weight and others and plotted a quick graph just to see what it might show.

MLV:


Virtual Dyno:


I will continue experimenting with this program just for fun, but a lot people say it is quite accurate. It would be actual power after drivetrain and losses. I assume it uses sea level for the calculations and my inhg barometric pressure would really be ~24.5.


Regarding MAT correction I have started leaning against changing the values from close to the default ideal gas law. Some people say set everything to 100% and tune at the same temperature. Then adjust after you are happy with the tune and see differences in temps.

Here's a great thread to read discussing this problem: https://www.miataturbo.net/megasquirt-1 ... ike-89026/

What I found was if you tried to use the table to account for ambient temperature swings, the range below atmospheric would get fairly close, but then (if the setting for air density was above the default value) the calculation for fuel in boost would compound and the tune would then be too rich. You would have to go back and remove fuel in the boost regions because the compounded calculations became too rich. Since it is compounded, temperature swings will more greatly affect the required fuel values and you will be in the seam position as before.


I also don't like the idea of autotune having to mess with the boost region of the map if possible. So alternatively (in the default settings) the engine conditions below atmospheric pressure mean that mixtures will vary greatly with ambient temperature swings. In my mind, If EGO control can account for about +/-10% in this region, your routine driving should be pretty close to the target fuel mixture.


With that in mind, what are people's EGO control settings? Who uses it at all?
Right now I use:
Ignition events per step: 36
Step size: 1.0
Authority: 14
Below TPS: 85%

I got a door hinge last weekend and attempted to replace it yesterday. First tried an electric drill, but the three phillips head screws didn't budge. So i purchased a manual impact driver. Unfortunately that didn't work either. I tried air tools on as low of a setting as possible and the screws still weren't budging. As I slowly cranked up the air pressure, nothing happened until sunddenly the screw stripped. I tried all three and the same happened.

So yesterday I spent the evening drilling out the three.



To get the door hinge off, I had to almost pull the door off the lower hinge to be able to slide out the upper.





The offending hinge/pin.



And finally got it reassembled with new screws. Here's a pretty good photo of the Recaro seats


Trying to resume tuning but my usb isn't being recognized by my laptop anymore. It seems hit or miss when it works

The cars is looking good1 Maybe try reinstalling the usb driver, and make sure your on the right usb port. Good luck

I need to do my door hinge pin too. My door dosent align right, and shakes at highway speeds . Not looking forward to those screws

The hinge wasn't bad after I got the screws drilled out, but certainly more in depth than I originally thought!

Regarding the drivers, thanks for the suggestion Marc. I noticed I had issues after a windows update, so I reverted to the previous restore point which seemed to work. I do still have communication issues for ~5 mins after a cold start until suddenly the connection is recognized. I suspect it is related to internal temperature of the MS2 box, I'm not sure.

I created a custom dashboard based on one I found online. It's based on the TunerStudio dash shown on the load/open screen. Still trying to decide what is important but this was a good start for me to work on idling and to prevent stalling (Idle PWM%).



I also like comparing target AFR to actual measured. I am still getting noise in my harness, but I am not sure where to really start to try to iron out the issue. I think it is due to previous splicing into the harness before I bought the engine.

Also I took two different virtual dyno runs in different temperatures, humidity on different days. I thought the result was interesting. I really need to try one of these on the flat part of an empty highway to see if it changes at all in 4th gear. Next step when I am fairly happy with the tune will be to try these tests with my manual boost controller and see if there is a difference of how boost builds. Then I will turn boost up close to ~10psi/150kPa when I plan to leave it for continued track use.

Your dashboard is looking good. I made mine black so it would be easier to see and tune at night. I like to see what you've got there, IAT, CLT, Ignition Advance, engine map, baro, voltage, and now ego correction/ afr target. I also like to see what pwm my icv is at. I turned on my on/off display for fuel cut, need burn, WUE, ASE, and when accel enrichment is on. That is handy to know. I like the font on your as well as a couple of the gauges. Here is a view of mine. I was adjusting it so it doesnt look super finished. I like the afr log map so you can see what your afr was after a boost session.



I may take some elements from yours and make my own. But after riding in it last weekend, it seems much smoother. The last time I rode in it was either at the track or at D4C so take it with a grain of salt, but the new engine runs good and aside from a little boost lag, it seems to be really good. Hopefully the adjustment to the autotune makes cells change less readily and the smoother fuel map makes it run better. Once you get the noise reduced from the afr gauge it should be close to dialed in. I'll have to try a log of vitual dyno. I was going to do it the other night on the way home but couldn't get the dang driver to work. But 250/300 is pretty good, is that at the wheel or at the crank?

That's allot of shit to be looking at.

I have continued with tuning and have still had issues with the fuel map leaning out in hot weather like today. In 60F morning temps, everything was rich. In the afternoon at 90F ambient, the map was lean. I do think it is related to MAT correction and the map trying to run based on ~85% fuel or so. Will continue messing with it and report back.

Also, I visually updated my dashboard and I am pretty happy with it. Like I said, it is sort of like the TunerStudio entry screen





For tuning return to idle-- my closed loop idle settings still occasionally cause a stall and I think it is due to my initial value map. So I am tuning that map. Here is where I am currently at. Full closed is about 32-34%.




I used some of George's settings for enrichment and that was an improvement from what I had. Here's where I am currently at (which has helped with throttle blips and eliminating the lean spike in throttle tip-in).



Adjusting my warmup enrichment has also helped the cold start tendancy of the car to want to stall.



Lastly I did some reading about ideal fuel mixtures under boost. I thought this was interesting in relation to AFR under boost.

Jeff Hartman's book http://www.amazon.com/Modify-Management ... 538&sr=1-2. He's a chapter on Air to Fuel Ratio (AFR) theory that speaks directly to boosted engine that are knock-limited.

Lets have a discussion as to what is the best AFR for our engines when we want to make the most power without blowing up. Lets limit ourselves to one fuel: gasoline. 91/93 octane. Since that's what the vast majority of us use.

First some Ratio's:
14.68:1 (~14.7:1) Stoichiometric mix
12.8:1 - Lean best torque
12.2:1 - Mean best torque
11.76:1 - Rich best torque
11.01:1 - Fastest flame speed in cylinder

All five ratios are directly from Hartman's book.

He argues that for boosted engine that are knock-limited, "at a very minimum you are aiming for mean best torque" or 12.2 AFR, "if not rich best torque" at 11.76 AFR. 11.76 AFR has a side benefit of higher cylinder flame speed.

Torque starts dropping off the moment you start to go richer than Mean Best Torque and begins to drop off dramatically after Rich Best Torque. Counter-intuitively the Fastest Flame Speed at 11.01:1 AFR does not produce the most torque, but there is a cliff-like drop off in torque when you start to go richer than that point.

To sum up, Hartman states that 11.76 AFR is the richest one should shoot for. You may have to tune back the boost to get there but you'll get the same torque with better fuel usage than if you turned the boost up and went crazy rich (e.g. 10.8:1).

Going richer than target may depend on boost level, fuel quality/octane, static compression ratio, air temperature fluctuations and things like the condition of the fuel pump. With that said I aim for 12.2 and slightly lower as boost level increases.

Good info. I see mega log viewer use 12.5 for it's ve analyzer at boost. I am still whittling down my ve from an earlier rich map, after my latest hg job. The rich(like 10:1) just blubbers, but I get zero knock from my knocksense light, so I'm going slow.

The EAE in the old tune in sent you was pretty rough. I've since dialed it back a little bit. Im still a little too rich on a big throttle spike. More tuning required on my end. Are you running 3.4.2?

Also, are you running fuel cut on decel? For me the transition back to fuel can cause a hiccup if the RPM reduce down to the transition point for idle. Still a work in progress.

Yeah, 3.4.2

I need to continue to revise the accel enrichments, but increasing values closer to yours seemed to help. If you do have settings you are more happy with, I'd certainly like to see them. I still get a lean spike when hitting the gas. Not as bad as before, but I think slightly rich is better than lean!

I have 100% decel fuel, so no cut at the moment. I'm not sure quite sure what I should set this to either. With the Motronic setup I had noticed rich surges during shifts and such. I just looked at a datalog and see the same thing happening. To a point this can be good when you are in boost, but in general I think you want to make sure the AFR stays close to the prior value (before lifting the throttle at all). I am going to try 90% and see how that works out.

Hoping to report some more MAT info later this week. I have started watching Fuel correction% and overall correction in my TS dashboard. In the morning it would be at 110% and 94% in heat of the day, causing my map to be either too rich or too lean in the temperature variations.

A nice long drive on an empty freeway can do wonders for accel enrichments.

Have you tried 0.2ms for accel time and 0 for end pw and taper? Mine seems to respond the best with a momentary rich dip then right back to stoich. I tuned for no change in the target afr and still fought an intermittent stumble.

Interesting take on the mat correction. I read the algorithm isn't as effective at higher rpm and/or engine loads. I haven't seen the issue like you describe with boost regions running richer. I still have hot restart lean conditions, but it's not severe like it used to be. I have found that as my fuel temperature creeps up, the afr starts to lean out. I'm thinking all the e85 has done some damage somewhere or one of my pumps is dying. Either way, I'm completely revamping my fuel system.

I am wondering about the corrections-- maybe they are more inaccurate due to my altitude. Like possibly that air density due to ambient temp doesn't fluctuate quite as much at 5000ft as at seal level, so the default MS2 ends up being too aggressive for me.

Mike, I'll have to try your accel enrich suggestions. After seeing a datalog I can already see my TPSdot values should be set closer to 50-60mS for a pulse of fuel with smaller throttle adjustments.

Regarding your lean hot restarts, why not simply add more fuel to the Warm up enrichment table at that spot? So if coolant is 175F, it adds more fuel to the taper process... It could also be failing hardware too like you mention.

What I did was add more temperature than needed to my tables for ase and idle DC. (I'm back to open loop iac) So now instead of continuing to add on to the "at operating" temp range, I have a new point thats 30 degrees higher for hot restarts. I haven't turned ego on, but I'm pretty sure when I do, it's going to handle any temp swings thrown at it.

Are you using just the MAT air density or are you also using MAT/CLT? I seem to find a lot of conflicting info on the MAT CLT.

ASE, that's what I was thinking. That's how I would try to resolve the issue. Interesting you went back to open loop, however....


George, yes. I suspect MAT air density is better to use. If you do CLT/MAT you can run into a whole broad range of ambient temps for the same coolant temperature. Tuning with that table doesn't make sense to me.

In driving around a fair amount lately, I seem to be getting a very noisy signal into my unit. With the engine off, all sensors report very smooth signals. However, once I start the car, the CLT, TPS and AFR signals all become very noisy. I was hoping to try to reolve this today but have gotten nowhere yet...

The wideband is grounded through the MS2 adapter wiring. I have tried powering and grounding the AFR signal directly to the battery for no change in behavior. Idle AFR's jump from target idle of 13 up to 15 and down to 12. Ignition advance and Ve table values are constant in this closed loop idle situation.

Regarding variable TPS wiring,I do not have a way to shield these wires in the stock wiring harness. Is this something I need to do? Connecting the ground reference wire directly to the battery made no difference in the noise I am seeing.

Do you have any ideas for me to troubleshoot? I don't know if my VR adapter cable is shielded or not, all I know is it connects up properly to the stock harness through the reference sensor.

I have checked engine ground and all seem to be working well...

Can you post a log?

It sounds like a bad or intermittent ground, possibly fretting buts its difficult to diagnose without seeing a log file.

These files are good for 30 days:

https://uploadfiles.io/68zpa

https://uploadfiles.io/5veh4

I believe these settings show the inputs after lag factors (Which right now are set at 15-20)

What are the lag settings on your wideband? Can you slow it down to get a better average reading?

Most of the noise I've seen/read about with MS is engine harnesses routed to closely to secondary ign wires. Do you have the coil wire or plug wires running near any other wires?