1 ... 117 118 119 120 121 ... 129
Pete. (l33t FS)
Pete. (l33t FS) GRM+ Memberand MegaDork
9/17/24 2:20 a.m.

Wouldn't a cylinder that isn't firing cause a low MAP signal for the cylinder 90ish degrees further in the cycle because the engine slowed down due to the misfire, and thus would have reduced vacuum?

 

I was just thinking, how one can see a misfire with an oscilloscope and a MAP sensor on a single plenum engine.

Paul_VR6 (Forum Supporter)
Paul_VR6 (Forum Supporter) UltraDork
9/17/24 6:21 a.m.

I would think so, yes, unless they are tied together behind the tbs like for an jdle valve port thats shared

akylekoz
akylekoz UberDork
9/17/24 7:20 a.m.

I'm learning so much about how I will never try tuning until I have retirement time.  Also I will start on a 4 cylinder.

mke
mke Dork
9/17/24 9:03 a.m.
Pete. (l33t FS) said:

Wouldn't a cylinder that isn't firing cause a low MAP signal for the cylinder 90ish degrees further in the cycle because the engine slowed down due to the misfire, and thus would have reduced vacuum?

 

I was just thinking, how one can see a misfire with an oscilloscope and a MAP sensor on a single plenum engine.

How do you differentiate a low MAP due to misfire from a poor TB sync?

I don't have it working completely right yet, but I due have a misfire function that kind of mostly works by looking at the time between teeth on the flywheel.  Where it is right now, when it says all is well, all is well but when it sees issues something is off but its not  necessarily a misfire.  Its an "in my spare time" thing

mke
mke Dork
9/17/24 9:08 a.m.
akylekoz said:

I'm learning so much about how I will never try tuning until I have retirement time.  Also I will start on a 4 cylinder.

Its not supposed to be this hard.  I have known mechanical issues (the leaking head gasket) and untested software in the ECU and only scraps of a base line tune....its a challenge.  Normally the the software has been used by hundreds or thousands of people and at least one has shared the tuning info and the engine itself is fully functional and you're mostly just left with tweaking.....none of that here crying

mke
mke Dork
9/17/24 9:27 a.m.
Paul_VR6 (Forum Supporter) said:

I usually lock timing as I make these kind of adjustments. Get fuel consistent and go from there. Even idle a hair rich to give it wiggle room. Leaner than stoich itbs get grumpy faster. Then go play with timing then fuel again. 

I kind of lock the timing.....I set points to give stable timing in the idle range and that is what I thought I did here:

it was idling at 40-45kPa so I put the bottom of the table at 50, so below MAP does not impact timing.  But now its idling at 50+/-  and I have some higher rpm stuff below 50kPA.....this needs work.

I relooked at the program and everything on the MAP fuel correction looks right, there is no dead time in the math error I thought there might be.  So this one is confusing me......I think in the short term I will just throw in "% to use" kind of thing where 100% is where it's at, 0% would effectively be off  and play with that until my brain understands what the heck is going on.

Paul_VR6 (Forum Supporter)
Paul_VR6 (Forum Supporter) UltraDork
9/17/24 1:23 p.m.

Ye dead times could still be wrong. Most published data isn't with your ecu's driver. I have seen some .5ms off, I usually default .9-1ms at 13.2v for injectors I don't have data for. 
 

If missing the map should go up to near at least a few kpa for the missed cycle. It will be intermittent if its missing/not or just be worse map, depends on how fast the ecu is sampling and logging. 
 

look to see if there is an idle timing table. I like to use the bottom row for decel advance coming off fuel cut, and it would be hard with a single table. Other way to "fix" timing is just put 6 everywhere for mow like its 1998

mke
mke Dork
9/17/24 2:39 p.m.
Paul_VR6 (Forum Supporter) said:

Ye dead times could still be wrong. Most published data isn't with your ecu's driver. I have seen some .5ms off, I usually default .9-1ms at 13.2v for injectors I don't have data for. 

They are injector dynamics injector so very well characterized and I use the data in the GM format.  The dead time table is 17x33 and about 1.16 with normal fuel pressure and voltage.  It could be wrong I guess but I did have the fuel pump fail and fuel pressure had dropped to 10psi before I noticed and mixture issues and same with system voltage all buggered the other day by the shorting starter and again no mixture issues I noticed so I'm thinking its at least close.

If missing the map should go up to near at least a few kpa for the missed cycle. It will be intermittent if its missing/not or just be worse map, depends on how fast the ecu is sampling and logging. 

The ECU reads the sensors at 2kHz so that is how it gets the global MAP reading from the multiMAP board, but the cylinder MAPs are being read from an expander over CAN so that reads when its thread is run, which is 10msec so 100Hz so I'm limited with what I can do with it.  

The crank tooth time thought is extremely fast and accurate.  That is a low level processor function that I can read out

look to see if there is an idle timing table. I like to use the bottom row for decel advance coming off fuel cut, and it would be hard with a single table. Other way to "fix" timing is just put 6 everywhere for mow like its 1998

There isn't one, but "right click,  add table" ....I can add anything I can dream up.    fully programmable means fully programmable :) 

I try to avoid table switching so I'll need to thing about how I want to set this up.  This is the first engine I''ve messed with where I used reduced timing to help control idle speed and that is changing how I normally do stuff.

Paul_VR6 (Forum Supporter)
Paul_VR6 (Forum Supporter) UltraDork
9/17/24 4:33 p.m.

Just turn it on rpm <1000, map<50, tps <2% etc 

mke
mke Dork
9/17/24 10:22 p.m.

Looking back, tonight was mostly wasted time.  I did add more sealer and doe a heat cycle.  I did type the wrong number into the throttle and get to hear it bouncing off the rev limiter which its nice to know works correctly, would have been even nicer to have the sound on video.  I did drop the mouse in the shut it down panic, and that was the clacking I heard pulling out.

I decided to add a 0-100% MAP correction number, so a little change in the model...just as I got started Lana called and I thought something was wrong so I unplugged and went to see....it was nothing so back to work.  The fuel we have is E10 so I had the wrong stoic and density numbers and I figured best to change it now before I'm very far into tuning.  I also notice something wrong with the MAP 7 reading at key on, it was 92kPa when it should be 100, so I turned off the MAP trim I'd just worked on.  I added row to the spark table to give myself more options and set idle timing to 8 this go.

Fired it up a few tweaks to get it idling okish, sealer in, cap ON.  Let it warm up....the MAP screen is saying one thing but my eyes are saying the TBs are not all alike so I used a feeler gauge and my hear and it was sounding better and MAP was around 45-50. A little drive and cylinders wanting to miss...It was very lean form the stoic number change.  Pulled it in and went to look at logs to tweak tables for tomorrow.

hmm....why is cylinder 1's injector pulse about 1/2 the others?  Because its trim factor is 0 not 1 like it should be.....1+ 0*1.1 is some how 0?  WTF is going on? I see the numbers, I see the equation and the answer is wrong? Best I can tell when I unplugged mid change I corrupted something and cylinder 1 injector was getting the deadtime pulse but no fuel pulse.   I went back out and loaded the model from Sunday and reloaded all my new tuning, same issue.  Reloaded the model from Sunday and the tuning from Sunday, problem solved so whatever is corrupt is in the tuning data... so I'll take some screen shots and rekey the new stuff it so I know I don't bring over any problems.

Also I didn't really give the #7 MAP issue the thought it deserved...the board outputs the lowest MAP reading as the Global value so WOT was 92kPa but the running reading seemed way more right.....hmmm.  I had a plastic bag over it to protect the board but its not sealed...yup, the over heat last week got water in the bag and there is some corrosion and 6 doesn't look any better. hopefully when I clean it all be well, if not I MIGHT have another sensor or 2?  All 12 need to be the same and I'm pretty sure the part I'm using has been replaced so lets not think about that.

Tomorrow is pull plugs and clean plugs, key data, clean the MAP and maybe get another heat cycle if it hasn't gotten dark and MAP is working

mke
mke Dork
9/17/24 10:35 p.m.
Paul_VR6 (Forum Supporter) said:

Just turn it on rpm <1000, map<50, tps <2% etc 

Yes, that's easy and when it click-on it will settle into idle.  When it clicks off timing will go from whatever I like for idle to whenever runs best so , 6 or 8 at idle to 15 or 20 which usually causes an off idle surge, at least it did last time I played with it and decided I didn't like table switching.  I could add some kind of interpolation  to make it like a 4D table to bring it off smoothly,  but I should be able to do something similar by adding rows/columns maybe? My thought is anything under 1200 or maybe 1500 rpm is not for driving real and would need a decent amount of throttle to be usable, so I'll try to define the idle space so it doesn't bugger anything important. It is where I plan to start and complicated it as needed I guess. .....if I ever get past the issues I keep finding/creating.

mke
mke Dork
9/18/24 10:49 p.m.

Today was clean the plug...very black today so clearly I was going the wrong way after changing the stoic value...oops. And lots of variation with the MAP trim off and me messing with the sync by ear...another oops I guess. On a good note, 1, 3, 12 all dry, 6 & 7 just damp so I'm guessing 1 more run section and the heads will be sealed...hopefully

I rekeyed all the new timing mapping in. Did not rekey the bad tuning. Did rekey the MAP trim % function and its working correctly.

That bought me to the # 7 on the multiMAP issue. I did have 1 new MAP sensor!

Attachment:


...and installing it made the problem worse. It ended up being corrosion in the board itself killed power to MAP 7 so for now I added a jumper and its fixed.


....but long term I guess I need to be thinking about a new board. If anyone is interested in a multiMAP group buy/build let me know.

An aside a buddy txt'd asking if I want a set of 4:1 scale levers...yes?

I'd never seen them but google brought me a pick from...GRM


An amazon had a 400lb scale on lighting deal at $11 but the other 3 I had to pay full price so it ended up $58 for 4.

I'll drive up and grab the levers Saturday so another $20-$30 in gas and find out what this car weighs this weekend....kind of curious....and that is no doubt encourage weight reduction efforts

Paul_VR6 (Forum Supporter)
Paul_VR6 (Forum Supporter) UltraDork
9/19/24 7:39 a.m.

Yea you could probably set that to 1200-1500 easily but you also can keep it a little tighter later once you get rhe rest tuned in to catch issues. Like a dropped cyl will have high map and kick you out if idle and alert there is an issue. Different strategy for different goals

What kpa maps? I have a bunch of spare 250s here

edit: I have those same scale bars, they work great

mke
mke Dork
9/19/24 11:00 a.m.

In reply to Paul_VR6 (Forum Supporter) :

That is exactly what I did, at least for now.  This think will idle...well I had it as low as 600 but factory spec on this just like stock 308 is 1000.  Best I can tell its about oil pressure...it starts to get pretty low at low rpm once the engine is good and warm so the idle will probably need to be right around 1000....although I see liquid moly has 10w60 oil I might try.  Anyway, I though in some more rows and will probably add a few columns so I can get more aggressive adding timing off idle while still getting it to actually drop into idle.  But its a start.

The MAPs are 105kPa, the new versions are 110 I believe.  But I had 1 new, I have another installed on a test board and I think the one I removed still works so I think I should be ok for that.  The dead power pin on the board concerns me though so I'll likely make a new one.  If I do, I'll offer up the old one, its designed so it can literally be bandsawed to get the number of cylinder you want, this is still a perfectly good 10 cyl or less board.

Detecting errors is something I need to improve for sure.  Like here I should be looking at all the MAPs plus baro at key on prior to cranking and flagging an error if they aren't matching within a % or so.  Its a long to-do list.......

On the stoic thing this feels like the right moment to get flex-fuel setup.  I don't actually program the ECU, I build a model that calls library function.   Most functions are very generic, like "math" or "if-else"  so its a pretty generic controller that could just as well run commercial equipment but there are a couple functions specifically for running an engine.   One of them is calculation of injector time

I use this function and the very last input is "flex_stoic_ratio".  The function instruction say"Input11 - FLEX_FUEL_STOICH_AF - This input can be used with a flex fuel sensor for calculating the stoichiometry of a mixed fuel system." Whatever that means and the example screen shot above its set to 0 so that is where mine is set and it works correctly.  It might by % alcohol?  It might be 0 disables it?  Its possible to run the model on the PC so I can figure out what its doing but I haven't done it in years now and when I tried yesterday the only function that wouldn't run right was...LSFS.    I think I probably just need to put a 0.1 in there for 10% alcohol but I'll figure it out and I already added that variably to the user tuning screen so I can just set it until I add a sensor for it.

 

 

Paul_VR6 (Forum Supporter)
Paul_VR6 (Forum Supporter) UltraDork
9/19/24 11:52 a.m.

I'll see if I have any spare 110s I only used them a few times and likely won't need them. 

mke
mke Dork
9/19/24 3:40 p.m.
Paul_VR6 (Forum Supporter) said:

edit: I have those same scale bars, they work great

What are you using for scales?  My  great bathroom scales shut off after 30 secons or so and won't come back on unless I unlad them.  for $16-20 I can switch to mechanical scales that I know will work or at like $32-40 shipping scales with controllers that might work...they auto shut off but pressing the on buttom might get numbers back or might send it into tare fuction, not sure how on know without buying them.

mke
mke Dork
9/19/24 4:09 p.m.
mke said:

I use this function and the very last input is "flex_stoic_ratio".  The function instruction say"Input11 - FLEX_FUEL_STOICH_AF - This input can be used with a flex fuel sensor for calculating the stoichiometry of a mixed fuel system." Whatever that means and the example screen shot above its set to 0 so that is where mine is set and it works correctly.  It might by % alcohol?  It might be 0 disables it?  Its possible to run the model on the PC so I can figure out what its doing but I haven't done it in years now and when I tried yesterday the only function that wouldn't run right was...LSFS.    I think I probably just need to put a 0.1 in there for 10% alcohol but I'll figure it out and I already added that variably to the user tuning screen so I can just set it until I add a sensor for it.

It looks like 0 at power on is a disable for the feature, proper use apprears to be just set the defualt at 14.7 then read a flex sensor and use a lookup table to convert the sensor reading to an A/F number that the function will read.  For now I just set it to default to 14.13  for E10 and it seems to be working exactly right and I have to start pulling fuel out.  It was running really nice but I think still rich and loaded up the plugs and I started loosing cylinders so after it cools a little I'll give it another go.  Idle is 45kPa right now but TBs are about as closed as they want to go so I'll probably drop the timing a little more to leave me so adjustment....but full hot it's 1000rpm +/- so fine I guess.

 

mke
mke Dork
9/20/24 10:49 a.m.
mke said:
Paul_VR6 (Forum Supporter) said:

edit: I have those same scale bars, they work great

What are you using for scales?  My  great bathroom scales shut off after 30 secons or so and won't come back on unless I unlad them.  for $16-20 I can switch to mechanical scales that I know will work or at like $32-40 shipping scales with controllers that might work...they auto shut off but pressing the on buttom might get numbers back or might send it into tare fuction, not sure how on know without buying them.

I found some $32 shipping scales that are dual power (battery or plug) and specifically say they do NOT auto shutoff by default.  Arriving Saturday.

I'm other news a fellow 308 own has taken pity on me after the multiMAP failure and offered to and offered to design me a better one with the CAN function built right in.  That will clean up the wiring a bit.

Yesterday's second run I leaned it much too much.....so later today I'll put fuel back and it will run better.  The cylinders shutting down seemed to be water, 6, 7 and now 12 was wet again.  More sealer.

Paul_VR6 (Forum Supporter)
Paul_VR6 (Forum Supporter) UltraDork
9/20/24 11:52 a.m.

I use cheap mechanical scales. I looked into other options but they got close to what a real scale system costs. 
 

What plugs do you run? A little rich if stoich shouldn't be fouling plugs that quick. Edit: or is that more the water issue?

mke
mke Dork
9/20/24 2:17 p.m.
Paul_VR6 (Forum Supporter) said:

What plugs do you run? A little rich if stoich shouldn't be fouling plugs that quick. Edit: or is that more the water issue?

Stock TR - DR8EB

SO part of it is me doing foolish things, part mechanical with the water making me think I'm doing the right thing when I'm not and part mechanical with what sure seems like injectors acting differently from each other.  

I'm going to head into the trim tables and see if I can't even out the colors a bit.  I have 2 (6&12) that look fuel glazed which is likely why I kept thinking I need to lean.  Then 7, 9, 5 are pretty white.  The plan is see if I can change the color of these plugs to match the others before attempting much more actual tuning.

 

mke
mke Dork
9/20/24 4:06 p.m.

I did some rework on the fuel map and got into the cylinder trim tables a bit trying to even out plug colors and they are not so bad after like 30-40 minutes sitting at idle.  I play a bit more with that but in general I'm pretty happy.

7 has been white all along and today I added 20% more fuel....still white so I'm thinking it just leaking so bad its steam cleaning itself but the sealer is really trying to work and was able to filled the plug gap

 

dave215
dave215 New Reader
9/20/24 5:27 p.m.

Idea .that stuff can certainly bridge gaps .I am running a davies craig electric water pump on my lsi -low pressure but keeps it cool .Since you are running high rpms would seem to be pretty significant pressures which may not enble the sealant to set up .

dave215
dave215 New Reader
9/21/24 10:57 a.m.

If you want to experiment I will send (on me)  you an old one (8 years old -noisy bearing) but works .davies craig 150ewp.they have a lot of s;pecifications on their website

mke
mke Dork
9/21/24 9:30 p.m.

I drove the 3 hours each way to pick up the scale levers and they came with a pair of used but still usable yokahama A052 295 tire...man they are wide.  The 285 sumitomo's I have are about 10" of tread, these are closer to 12 so I'm REALLY glad I got a free pair to test fit so a big thank you to Aaron.  He tried to talk me into taking the 315s he had but these already scare me.

Weights (with at least 5 gal fuel)
Total 2970
Front 1228
Rear 1744
Right 1498
Left 1504

Now we know.

 

Getting it sealed I idled it for 45 minutes....that's right it frikin sat there idling for 45 minutes and nothing broke!
 

but 7 is leaking badly.  I pulled a sample of the coolant while it was still pretty hot and shot some onto the cold welding bench....it did nothing....so I hit it with a torch to see it react and the water boiled on and what was left was a little gick, not the white ceramic you need a sander to remove.  I now believe this stuff is dead at 50 minutes.  I was remembering needing to do multiple heat cycles so I tried to use the wayback machine site to find older instructions to see what changed but no luck.  So I dug back though the forum to see what I said...yeah...this is not the sealer I used last time or ever, no idea why I thought it was.  

Tomorrow I will go get the barsleak 1100 headgasket sealer that worked but needs at least 3 heat cycles and a well flushed system and hopefully get this think sealed.

mke
mke Dork
9/21/24 9:36 p.m.

In reply to dave215 :

Thanks for the offer Dave!  Tonight I realized that I'm not using the sealer that actually worked so tomorrow I'll start over with the right stuff and hopefully get it sorted.  When the sealer it working I'm supposed to just let it idle.....tomorrow I'll do better with that.

1 ... 117 118 119 120 121 ... 129

You'll need to log in to post.

Our Preferred Partners
0tOVeBSgmtyCTUsK2pPCea10r6HctkvTvfZPl1qh8mTPiaElCdY1I00Vln6bkMfV