Miata Turbo Forum - Boost cars, acquire cats.

Miata Turbo Forum - Boost cars, acquire cats. (https://www.miataturbo.net/)
-   MEGAsquirt (https://www.miataturbo.net/megasquirt-18/)
-   -   Reverant MS2, Lean under throttle tip-in + slight stumble, accel enrich wrong? (https://www.miataturbo.net/megasquirt-18/reverant-ms2-lean-under-throttle-tip-slight-stumble-accel-enrich-wrong-72463/)

sbcrx007 04-28-2013 08:39 PM

Reverant MS2, Lean under throttle tip-in + slight stumble, accel enrich wrong?
 
Right, so I'm running a Reverant MS2 basic standalone on a 93 1.6L car with RX8 injectors, a BMW vTPS, running NA only, have not installed turbo yet. I'm still learning tuner studio and a bit overwhelmed with all the options/menus that are present, and want to make sure I'm not missing something obvious here...

Issue:
The car runs well enough off idle, and will idle fine with some hunting around - which sounds common - except if you lightly dip into the throttle you get a stumble, or if you mash it from idle it falls on its face momentarily before revving up normally. During the stumble, it goes lean on the WB gauge and on the TS gauge, more on this in a moment From what I recall, it did not have this stumble when running the stock ecu.

Things done/tried:
1. Fuel system seems fine, new filter. Have not checked pressure yet, but assuming the pump can supply sufficient fuel at higher loads, light load shouldn't be an issue.
2. New plugs, factory gap (.035 i believe)
3. Newish NGK wires, stolen off my daily driver so they are known good w/ ~5k miles on them. Properly secured and w/ dielectric grease.
4. Intake is stock miata, thru the vane airflow meter. The meter is not hooked up, merely in the flow path so I can use the stock panel filter. Wondering if this was causing an issue, I removed the crossover tube from the vane airflow meter so that it was out of the path, yielded little to no change. Filter is new.
5. Checked resistance from block to chassis, battery to chassis. Block to chassis slightly high @ 3.5 ohms, added additional ground strap and brought it to 0.7, which seems good enough for me given that its a low end Fluke meter.
6. Double checked timing, running Reverant's timing map which seems pretty conservative and similar to DIY's. Timing matches when checked with timing light.
7. Prior to messing with accel enrich settings, I tried enriching the VE map in the 500/1000/1500 rpm range, from 35kpa to 100kpa. At times this seemed to help, maybe slightly but not much
8. Changed accel enrich settings from coming into play from 2500-5000 rpm to 800-5000 rpm (attempting to add more fuel at throttle tip-in) Also modified the curve from Reverant's original settings, testing it at more and more aggressive settings until achieving super rich readings at the same stumble. At this point the stumble felt/sounded a bit different, but was about the same length of hesitation. At this point I was using 100% TPSdot. I tried screwing with MAPdot as well with varying percentages, but it seems to just be a stumble with either a lean spike or rich spike at the stumble, no happy medium.
9. Also tried VE Analyzer Live, which was nifty but could not compensate for the stumble and ended up just giving me a hunting idle.
10. Lots of other random stuff that I tried in frustration, to no avail.

(Realized at some point that modifying the VE table as I was, enriching the hell out of the idle area, was really doing the same thing that the MAPdot enrichment under the Accel Enrichment settings was doing. So yes, my VE table is probably awful.)



So, am I missing something obvious? Is there a setting I'm not understanding that is set wrong?



Note: No, I don't have a datalog yet and haven't figured out how to export my .msq and such... working on it. Screenshots of VE and IGN tables instead:

sbcrx007 04-28-2013 08:45 PM

4 Attachment(s)
https://www.miataturbo.net/attachmen...ine=1367196510

https://www.miataturbo.net/attachmen...ine=1367196510

https://www.miataturbo.net/attachmen...ine=1367196510

https://www.miataturbo.net/attachmen...ine=1367196510

sbcrx007 04-28-2013 08:50 PM

3 Attachment(s)
https://www.miataturbo.net/attachmen...ine=1367196622

https://www.miataturbo.net/attachmen...ine=1367196622

https://www.miataturbo.net/attachmen...ine=1367196622

sbcrx007 04-28-2013 08:52 PM

Also have EGO correction set to "off" so that variable is out.

I've read and read, and searched and searched, tried the options I found... Appreciate any insight.

Jesse

sbcrx007 04-29-2013 12:36 PM

Also checked for vacuum leaks, checked for kinked vacuum line to MS2, and MS signal line already tees in @ fuel regulator.

I know my VE map is nuts, I feel like I'm compensating for something else when I add tons of fuel there or in the Accel Enrich wizard... So what am I missing?

southernmx5 04-29-2013 01:17 PM

Your settings look decent.

If you want help I suggest you make a datalog and take a screenshot of it or use MLV's convenient save to jpeg/png feature and post it.

edit: or post a full datalog :)

sbcrx007 04-29-2013 01:38 PM

Fair enough, I'll get on it tonight.

Braineack 04-29-2013 01:44 PM


Originally Posted by southernmx5 (Post 1006434)
take a screenshot of it or use MLV's convenient save to jpeg/png feature and post it.


not good enough.

sbcrx007 04-29-2013 02:12 PM

Well I've been watching the "diagnostics" tab with the live data streams in TS - I imagine a screenshot from Mega Viewer would be similar anyhow.

Brain, what would be sufficient? the whole .msq file? Or just a real datalog? I'll do a datalog tonight - will basically be with the car at a warm idle and me slapping the gas pedal quick like, or stabbing it to the floor pretty quick.

Also, I have a bmw vTPS that seems to be operating correctly and has been calibrated.
Also, the ground for the wideband is straight to chassis, does not share with radio or ign components.
Also, the MS2 unit is velcro'd to the carpet, so it is NOT grounded (per some reading here), and only connected via Rev's adapter harness to the factory wiring.

Braineack 04-29-2013 02:20 PM

how can a screenshot show me over 40 different data points at various points of a log

southernmx5 04-29-2013 02:48 PM

The lc1 should be grounded to the engine block at the same point as the factory grounds to avoid offset and noise issues.

sbcrx007 04-29-2013 09:19 PM

4 Attachment(s)
11.22 - Cold start. Wideband is warming up until ~24 seconds in. Spike at the end is a quick slap on the throttle about 3/4 to the floor.

24.34 - After warm up. Two quick throttle blips, followed by a quick slap to the floor, followed by a slow press of the pedal. Stumble in the rapid blips, no stumble on the slow throttle input.

The car seems to drive fine, power and drivability are near stock. The quick throttle input only really comes into play leaving from a stop... Which coupled with an aggressive clutch isn't any fun. Also, once its above 2000-2500 rpm, this stumble with rapid throttle input seems to go away.. not sure if that's just a function of things happening faster or what. Naturally, I didn't think to log that - can do if it's important.

Thank you

Grantofhell 04-29-2013 11:02 PM

Mine is the same, but it has now gotten to the point where the car will stumble at any RPM if I increase %TPS or stall if I disengage the clutch anywhere above 4k rpm and let the idles drop.

Gt2560rMiata 04-29-2013 11:32 PM

Not too much MS experience here. But you need a good VE table before being able to dial in Accel enrichments. Go run VEAL and do some driving then play with the Accel enrichments again

sbcrx007 04-29-2013 11:45 PM

Agreed on the good VE table, but I started with a decent table from Reverant, and was able to autotune it a bit better... This same stumble was present the whole time. My current lumpy VE table is only a result of trying to dump in more fuel at the stumble.

I was thinking that maybe somewhere there was a lag setting awry, or I'm not setting the accel enrich curves steep enough, or some other mystery setting that I don't know about...

Braineack 04-30-2013 09:57 AM


Originally Posted by Grantofhell (Post 1006638)
Mine is the same, but it has now gotten to the point where the car will stumble at any RPM if I increase %TPS or stall if I disengage the clutch anywhere above 4k rpm and let the idles drop.

these are two different issues. lets first address the idle droop.

is this during overrun, or anytime over 4K?

there could be a pleathora of reasons why is does this: needs more dashpot, poorly tuned VE table, overrun returning fuel too late and the mechanical delay between injection and combustion isn't fast enough, etc etc.

sbcrx007 04-30-2013 01:58 PM

Overrun/Decel fuel cut is disabled, as is EGO correction.
Again, driving around it feels fine because your rpms are almost always up off idle... My understanding is that it should idle ok @ ~800-900 rpm with the RX8 425/460cc yellows, without any mystery hesitation.

Any insight from the logs?

Braineack 04-30-2013 02:29 PM

I ahvent looked at logs yet.

sbcrx007 05-05-2013 09:04 PM

Update:

Playing around with AE wizard a bit more this weekend, I was able to get rid of most of the lean stumble that was present. previously the graph was starting at 50%/s, I brought it all the way down to 1%/s so I get extra enrichment right away, with less throttle input/speed.

Now I only get a lean spike if I just slightly touch the pedal... I don't get it, it goes lean for a split second and hesitates just slightly before revving normally. My understanding was that adding a vTPS would make this go away entirely.

No one sees anything wrong with any of the screenshots or log I posted?

thank you

Braineack 05-06-2013 08:29 AM

what is your TPS threshold %/sec in AE?


All times are GMT -4. The time now is 08:04 AM.


© 2024 MH Sub I, LLC dba Internet Brands