lbatalha

Can you not use ITB mode? I got it working nicely when I had ITBs.

Shouldn't your MS3Basic have dual sensors already? I don't think any version came without them.

freedomgli

Will eventually go to ITB Load Mode. But you cannot start there. First I must collect data driving around on Alpha-N tune to determine switch points. And right now I don’t even have my single MAP sensor hooked up to my vacuum manifold to collect the data as last time I tried it gave me issues. Only newer MS3Basic (V3?) come with two MAP sensors. Mine is circa 2014. So I’ve got more work to do yet. It’s easy to get 90% there. It’s that last 10% that’s a challenge.

DNMakinson

iTrader: (3)

Update:
I had two idle issues on 1.5.1 and on Pre-1.5.2 Beta 4:

Sometimes, right after start, PWM Duty % would suddenly drop, instead of following AfterStart Taper.
It turned out that I was dropping out of CLI. What was happening was that, after the start flair, RPMDot would go to about -600, which was below my low deceleration RPMDot setting of -300: set according to the instructions shown in the first post. I changed to -675 and have had no issues since.

When A/C came on, I would also drop out of CLI for a moment and car would struggle until it was re-established. This was solved by adjusting the Startup/Idle, A/C parameter “Idle-Up CL loaf Allowance (%)”. I had it at 10 and had to drop it to 8. I admit that I had this issue in 1.4.X as well, but did not know the cause. I still don’t really understand that parameter, but issue is now resolved.

EDIT: Re: EBC, for 1.5.X, I have ended up with nearly all same settings from 1.4.X, save I am using 60kPa Delta to prevent Integral windup, and Integral I had to change from “60” to “7”.

Botttom line, I’m on 1.5 now.

lbatalha

I started in MAP mode just fine, tuned it as best as possible, determined switch points and enabled ITB mode. I never touched Alpha-N.

The second MAP sensor won't be too much issue if you enable full time EGO and tune it right, but it does help with fast elevation changes.


Im loving 1.5.2 beta4. I just discovered you can have separate authority tables for EGO, one for adding fuel and one for removing, was this added now? I dont recall seeing it before.
This combined with no longer reseting EGO to 100% on overrun and such, makes it so much better

Ted75zcar

iTrader: (2)

Separate EGO is new in 1.5.2 beta

DNMakinson

iTrader: (3)

Do you know if they have changed the PID type yet? There was talk of moving from a Type C to a Type A or B, in order to obtain faster action.

Ted75zcar

iTrader: (2)

Latest beta is running a type B

DNMakinson

iTrader: (3)

Thanks. I will look into my settings, then. I have not noticed that the action has changed drastically, save for the not going back to 100% on over-run cut.

Also, I do not see the use case for (2) EGO Authority tables. Care to elaborate on that?

Ted75zcar

iTrader: (2)

MSExtra Forum Topic

Edit: not sure if your question is about the table part, or the separate authority part. I don't really see the table as being a big value add, the separate authority is something I requested.

lbatalha

Say you want full time EGO, but are afraid that if you give it too much authority under boost, a sensor malfunction could lean out too much and kill your engine, with this you can limit the amount of negative EGO, while maintaing however much positive EGO you want.
For instance, I have 5% positive correction above 100kpa now, but only 2% negative (previously I ran 2% both ways since I calculated that to be a good safety margin in case of wideband malfunction)

Also yeah, type B PID, seems to respond much faster with no other changes (especially the short term part)

Thanks @Ted75zcar for requesting it, It's something quite simple but improves flexibility a lot.

poormxdad

Does the MS interpolate between rows and columns in the authority tables, or do just get whatever's in the box?

Thanks,

lbatalha

Yes full interpolation as far as I can tell.
My current tables look like this:



Previously the negative table was the main table, but with this feature I gave it some extra pep

Skamba

With the information in this thread I just updated my MS3 Basic to 1.5.2 RC11. Seems to work quite well.

As I was kind of worried about doing a firmware update, I created a quick sheet with the steps to do the update based on this thread. If anyone else wants to use it, you can find it here. So far the car seems to drive well. Engine states are quite intuitive and the updated EGO seems to be quite good.

Since the update, I'm running on wastegate pressure - I still have to retune the Boost PID. Could someone share their Boost Control Settings as a starting point for me?

O R X O N

!!!! holy ****

A few weeks ago I moved from 1.4.x on the TruboKitty site to, per my last log header, "MS3 Format 0566.05 : MS3 1.5.1 release 20171006 16:30BST (c) JSM/KC ********"

An issue I did find was fixed; was when VVT hit the minimum CLT in 1.4.x it would sputter pretty violently. Just one ignition cycle was a complete wash, and it always annoyed me. I haven't seen this in 1.5.1 so far yet. It's possible this is from changes in ramping, but I don't see this reflected in "VVT Target" nor did I change any PID settings.

I do have some idle problems but they're my fault; I haven't spent the time to reconfigure everything with states just yet. When I have high load (such as AC) and go into CL idle, it stays up there pretty high, for a while. Probably have to adjust my PID and bring my gain up to ~800s.

As for EGO correction, I actually noticed it was a bit more sluggish, but again, PID retune is needed I guess. EGO correction is disabled while I iron out VVT and hand-smooth my VE table. My #1 issue right now has been solving this nightmare of a 3500 torque dip (N/A for now, ignore boosted rows).


I planted my foot at WOT and LFB'd to hold at 3500, and scrolled from 0 to 40 and to 0 again on my VVT table while watching my AFRs. Unlike literally every single person here, I make peak torque (by a huge margin) at only ~12* advance. Which weirds me out, because that's "the absolute value of VVTuner's base table minus the camshaft range of travel."

Mystralhawk

Not sure if related, but I actually had an issue on 1.5.1 where when Minimum CLT was met on my VVT settings, my car would have a sudden jolt. Turns out injection timing would freak out and was running at ~376 prior to CLT being 140.

I upgraded to MegaSquirt 1.5.2 beta 7 because apparently 1.5.1 had the wrong angle when adjusting injector timing. The release notes for 1.5.2 beta 7 specifically mentions "Injector timing adjusted based on actual VVT angle was using wrong angle." Release notes: https://www.msextra.com/downloads/de...-1-5-2/beta-7/

Once I started the beta version, I didn't have a jolt when VVT CLT minimum temp was met.

O R X O N

I noticed the jolt on 1.5.1 below, just took some time. I jumped up to 1.5.2 RC11 and it's been great so far. My massive torque dip above was my factory cats and the terrible shape they'd been in all these years.

Using DNMakison's EAE settings from a post in 2019 and EGO with something like a P:25, I:1, D:10 value works great. Ran into an issue where very low TPS (MAP in the 20s @ 2.5K-ish, but definitely not overrun) would start jerking with EGO adding/removing a few percent at a time. I think it was crossing in and out of 0% authority, but, it's low enough load that I'll just take more authority away until 35kpa or so.

O R X O N

All my gripes above have been essentially tuned out. But I do notice that launch control now only engages ~25% of the time.

Most of the time with clutch in, as soon as I push the gas, the car goes full rich and dies on the spot. I haven't found anything conclusive in my logs. Sensor inputs look fine; but I want to rule out my wiring work first. Just a bit of a petty coincidence it started after the firmware. Anyone see this before?

DNMakinson

iTrader: (3)

Reviving this thread because I'm considering going back to 1.4.1. Now that I have utilized 1.5.2 Beta 4 to look at the boost PID parameters, I don't think I have a need for any the advanced features of 1.5.X. They did finally fix the Limp Mode RPM Limit switch in, I think, Beta 5; but I had already developed a work-around for my use case.

Thoughts @18psi @Reverant ?

DNM

18psi

iTrader: (76)

I still hate engine states.

DNMakinson

iTrader: (3)

But are there useful items in 1.5.X that cause you to use the newer FW in spite of Engine States? Or do you tune Miatas with the newer only at the owner's request?

I can't recall anything that was fixed in 1.5 over 1.4, and I don't recall that I'm actually making use of any of the new features.

EDIT from the MSExtra Release notes says, no features that have meaning to me:

MS3 1.5 firmware
Key changes and Gotchas - from 1.4.x and earlier
=======================
1. Engine states
These must be configured for idle, overrun, idle VE, idle advance.
2. Dome control.
Initial test implementation now replaced by implementation integrated with
regular boost control. New table settings will be required.
3. Fan control.
New settings for TPS and VSS shutoff.
4. A/C idleup.
New settings for TPS, VSS and RPM shutoff.
5. Internally the handling of I/O pins has changed.
6. Nitrous progressive VSS and time based - **fuel adder handling changed**
7. Noise filter internal logic changed.

DNM