Another MS conversion, Progress, at last. Options

[MarioVelotta]

Injectors are ususlly rated at lb/hr or cc/min. Most pumps are rated in gph or lpm at free flow.

Are the injectors rated 23lb/hr at 36.5psi? If so that would be ~240cc/min. At 43psi fuel pressure they would be ~260cc/min or 25lb/hr.

My ignition table was optimized to the T1 platform, my gearing etc. I like to use the scatter plots in Tunerstudio to find where I drive the most and make changes.

Here is a trip I made. You can clearly see where I spend most of my RPM/MAP and that is where I made the most changes in my spark map.

(IMG:http://www.914world.com/bbs2/uploads_offsite/sphotos-a.xx.fbcdn.net-15980-1370919314.1.jpg)

So before you can make big changes in yours, I would get an idea of where you spend the most time.

[Mike Bellis]

If it is running decent at the current psi, don't change it.

[JimN73]

Mario, I've spent most of my time trying to sort out specific issues. I know where I drive most, but final tuning could give me different values than I have now.

then, I can work on your suggestion.

[wndsrfr]

The injectors he has a spec'd out at 36psi. I think we bumped it up to that.
The regulator is not manifold referenced.
And the setup Jim is running is a collection of parts that has been proven on multiple engines.

Mark,
With the ITB's and Jakes cam on my 2316 I never could get a quiet MAP signal for the SDS to use. Finally gave up & went to TPS sensing only and never looked back. Maybe give that a try to see if it provides any clues......

[JimN73]

Time for an update.

I felt that,if the injectors were to work as planned, they should have the proper pressure. I raised the pressure to 36 psi, plus or minus.

At the same time, I used the 'Initial spark advance table calculator' and massaged the axes to match what we had been using for RPM and MAP. The car seemed more responsive, but was less tunable.

I was going to redo it, but realized that the 'stock' settings for the calculator are not too different from we were using. So I used them instead. After tuning, there was no bucking and very little hesitation.

I did some logging and tried to tweak the VE table to get rid of the problem. Worked a bit and then got worse again. Since the spark advance table got me this close, maybe the final answer is hidden there. When I learn a bit more about the spark table, I'll tackle it again.

Smells lean at idle and surges, but changes to the VE table at idle values hasn't helped.

For now, though, I want to restart AE and warmup enhancements. But not for a few days.

Here's a tune and log from today. I welcome your comments.  June_15.zip ( 462.59k ) Number of downloads: 73

[JamesM]

Time for an update.

I felt that,if the injectors were to work as planned, they should have the proper pressure. I raised the pressure to 36 psi, plus or minus.

At the same time, I used the 'Initial spark advance table calculator' and massaged the axes to match what we had been using for RPM and MAP. The car seemed more responsive, but was less tunable.

I was going to redo it, but realized that the 'stock' settings for the calculator are not too different from we were using. So I used them instead. After tuning, there was no bucking and very little hesitation.

I did some logging and tried to tweak the VE table to get rid of the problem. Worked a bit and then got worse again. Since the spark advance table got me this close, maybe the final answer is hidden there. When I learn a bit more about the spark table, I'll tackle it again.

Smells lean at idle and surges, but changes to the VE table at idle values hasn't helped.

For now, though, I want to restart AE and warmup enhancements. But not for a few days.

Here's a tune and log from today. I welcome your comments.  June_15.zip ( 462.59k ) Number of downloads: 73

Have not had time to check out your latest logs yet but just wanted to add a quick comment on your spark table. As you have already found out with your fuel table removing resolution where it is not needed in order to increase resolution where it is can be quite helpfull. In most cases your advance is going to be all in by 3500 RPM(probably even less then that) So any values above 3500RPM are all going to be the same, and thus wasted resolution. My cars dont need this much resolution with stock cams and the fact that megasquirt averages out the in between values, but with your cam it might help.

[JimN73]

James, it seems to run better - smoother, quieter and more responsive, so I'm inclined to leave well enough alone.

What I'm seeing now, maybe it was there before and I missed it, is an erratic pulse width. The MAP trace is now fairly smooth but the PW is not, and it seems to be causing erratic RPMs. That may not be true, I'm not sure of the interactions of the pieces and don't necessarily know the villains from the victims. I don't have any idea about what could be interfering.

Mark has suggested trying a lower RPM lag factor. I'll try that.

[jpnovak]

I just took a look at your log file and tables. A few observations.

Why do you have 11 bins for rpm that stop at 3000 and one bin at 5000? This is not the kind of scaling you need. You will completely miss your peak torque fuel delivery and all the normal driving loads on the table.

Overall, your AFRs are lean in many spots.

The MAP values should again be rescaled. You pull about 40kPa under closed throttle. This should be your lowest row.

You pull 60-70 kPa at steady throttle cruise. 70 should be the middle row and 60 should be in the bottom third of your Y axis. Then fit as many bins as possible. I don't think you need an 87, 94 and 101 kPa bin. Just a 90 and 100 will do. You are basically at WOT at this point.

Take the time to setup the table now and it will make tuning easier in the long run.

Are you using MegalogViewer (MLV) or TunerStudio to analyze your datalogs? This will help immensely.

[JimN73]

I just took a look at your log file and tables. A few observations.

Why do you have 11 bins for rpm that stop at 3000 and one bin at 5000? This is not the kind of scaling you need. You will completely miss your peak torque fuel delivery and all the normal driving loads on the table.

Overall, your AFRs are lean in many spots.

The MAP values should again be rescaled. You pull about 40kPa under closed throttle. This should be your lowest row.

You pull 60-70 kPa at steady throttle cruise. 70 should be the middle row and 60 should be in the bottom third of your Y axis. Then fit as many bins as possible. I don't think you need an 87, 94 and 101 kPa bin. Just a 90 and 100 will do. You are basically at WOT at this point.

Take the time to setup the table now and it will make tuning easier in the long run.

Are you using MegalogViewer (MLV) or TunerStudio to analyze your datalogs? This will help immensely.

thanks, Jamie. I appreciate the specifics. The tables are pretty much what I started with. Most of the changes that I've made have been to try and eliminate the severe bucking that was happening, including concentrating on the top half of the VE table. If I can eliminate a row at the top of the table, I'll be able to tune the bottom half better

Most recently, I used the 'spark table calculator', that's what you see now. The car seems to run a lot smoother is more responsive. I just recalculated spark using my redline of 6200 rpm and got a table with columns for 3300, 3700, 4000 and 6200 rpm. The y axis seems to be fixed.

Onward.

[jpnovak]

Jim, PM me your email address. I will send you a corrected file that will rescale your MAP values and rpm bins based on the 6200 redline. You will have to start tuning. If you can send me datalogs I can help you work through the issues - but will need specific descriptions about the problems you are having.

When I am tuning solo I usually have the datalog going. I create a custom gauge in TS that shows AFR, RPM and MAP as large as possible. This way, when the car is acting up I can quickly glance at the screen and quickly pick up three numbers that tell me rich/lean, what rpm, and what load. Then I can look back at the datalog after pulling off somewhere safe and preferably in the shade. (IMG:style_emoticons/default/smile.gif)

I do not like the automatic table generators for these cars. There is just not a good correlation between the values and reality. The easiest to generate spark tables is to follow a distributor curve. Once you have the fuel mapped in, you can start making adjustments to timing. As you increase timing, you can run more lean at cruise to gain mpg. Again, a quick glance at the custom gauge will tell you MAP readings as a function of rpm and speed on level cruise. Use these values to increase timing and decrease fuel.

[JimN73]

Quarterly update.

Over the last months I have been going thru the iterative process of tuning, and reached a point where the RPM/MAP/PW all looked like a hacksaw blade. I took this to mean that I had tuned everything that was tunable and the display showed the effect of a very light flywheel and/or a high overlap cam. Guys at WCR told Mark that a 12 lb flywheel is really light and that might be part of the problem.

We changed the flywheel without much effect. I went back to tuning and was able to eliminate bucking on steady throttle and get rid of almost all of it on deceleration. Another member told me that he has bucking on decel with a mild cam, so maybe I was going to have to live with it. but I was going to fight it.

Then, when talking to Mark, we decided to try a firmware upgrade, maybe there was something in the new firmware that would help.

Wrong!! The upgrade took out the AFR, and restoring to the original firmware didn't restore driveability.

So, has anyone seen this before and do you know what the cure is?

[JamesM]

In my experience, with a firmware upgrade you will want to manually re-enter ALL your settings including your VE and spark tables. The locations that these values are stored in memory can change so you will no longer have the values you did before the upgrade.

As for the decal issues, are you talking light deceleration or overrun with the throttle completely closed? The majority of the issues I had with the overrun area when I did my first tune were timing related (to much advance) but really you need to log your AFRs and take a close look.