NefMoto

So, new pump, fpr and injectors are fitted and everything is dialed in 100%, LTFTs are down to 0.4% on partial and 0.2% on idle. (this was made during the summer)
timing etc are dialed in for maximum torque/power, and the car runs agressive and pulls strong, yet really smooth, no issues whatsoever.


NOW the winter is here its getting cold outside, 1.5dm of snow just fell, and the tuning for smooth starting and warm-up begin.

i understand that under 9 degC the alcohol part wont really evaporate/ignite well, and so im using injectors with insane atomization and high fuel pressure to counter this, and it works really well, but as it gets down to -40 here in the winters i really need to tune/perfect the cold starts.


people tell me to add several hundreds of percent to cranking and cold-start maps and so ive been experimenting alot..
on really cold days the engine still fires up allmost instantly, but then dies as soon as i even look at the thottle.. and even if i rev it to keep it running, there is just no power there at all, and i end up driving out of the parking-lot at 3000+ rpms with a white cloud behind the car.. :P

and so now i need to know which maps to fiddle with..!


the ones i have tried are:
FKSTT_0_A, FKSTT_1_A - which i assume is the cold start enrichment?
KFWKSTT_0_A, KFWKSTT_1_A - which i figured would be some form of cranking enrichment?

lately someone mentioned these maps KFFWL_0_A, KFFWL_1_A, which seem to be warm-up maps, which might actually be what i am looking for.. so now im confused as to which maps does what.. ???


i know i will need to add more fuel on cranking, but how/where?
and i know i will need to add more fuel when the engine has fired up, based on engine-temp from -40 up to ~20degC, i.e warm-up, but in which map?


Any help/thoughts is appreciated! :)

(btw. the engine is an 1.8t with ME7.5 wideband.)

To start with - why is it holding back. Are you lean or rich when this happens?

*KSTT maps instructs ecu to fuel during crank and start of your engine (few secs), then it switches over to KFFWL and they fuel your car up to 90deg

im having difficulties reading the wideband-sensor since it is not in working temp-range.. but once it reaches temp it tells me Lambda ~0.8ish.
i dunno what it should read, as were basically pushing ~75% of the fuel straight thru the engine here.. only igniting ~25% of it in really cold starts.. you think it can be calibrated using lambda even tho the ethanol is present and left unburnt/vaporized into a white smog?




masterj: are you 100% sure of this? if so, then i havent even touched the warm-up tables yet, and that makes a huge room for improvement, haha  ;D

after a quick look in FR it seems the FKSTT maps are the basic factor, with this and the weighting factors KFWKSTT, KFWKSTN i should be able to get the cold-starts sorted.. and get onto the warm-up which seems to be the issue when driving from the parking lot.
modifying KFFWL and KFFWLW should sort this if i understand things correctly..

i will need to dig into the ESSTT, ESWL modules a bit further.. thanks for pointing me in the right direction!  ;D



more E85 cold-start tuning experience is very much welcome, suggestions of increments, temps, times, timing etc etc  :)

Any progress with multi-spark as discussed here:

http://nefariousmotorsports.com/forum/index.php?topic=2213.15

yeah, im using multi-spark right now, been adjusting it a bit, but trying to add more sparks at higher rpm,>2000rpm resulted in missfire.. guessing i need to lower the dwelltime or lenghten the time between sparks.

adding more sparks down low, <1200rpm, did definitely result in cleaner startups with the E85.

Could you possibly detail your EV14 cold-start anecdotes?  What about your experiences with dwell time?
Your comment two weeks ago about the true linearity of EV14's post TVUB/KRKTE centering is so accurate.
The car pulls so fucking hard right now and I'm still on 15 psi.  Obviously I'm using our mutual friends timing maps too...

Well, try to use a separate gauge that will warm up the O2 sensor with just ignition on, so that when you start you see the actual mixture.
As for your second question, I don't see how the amount of fuel pushed straight through the engine would affect the lambda reading. So yeah, it definitely can and should be calibrated using lambda.

i would use innovate mtx-l gauge to fix warm up fueling ;) just wait for o2 to heat up on ignition without starting the car

roger that, ill hook up another wideband and see what it reads during cold starts, if this can indeed be calibrated with the lambda then it shouldnt be too hard to get it spot on..
id imagine the map will look pretty strange at the point where the alcohol start to ignite.. since it will go from ~2-300% enrichment down to 5-15%.. :P

ty for the advice, ill look into this  :)

I was going to suggest looking at the dwell time and spark, but it looks like you've already covered that.  Following this thread for my future E85 reference.

using NGK BCPR9ES-11 sparkplugs, these help tremendously with e85 starts and smoothness, them are colder and with a wider gap, and this im combination with MSD produces multiple really strong sparks, dont think i can improve it much further without adding plasma discharge.. :P

why i chose these plugs is because:
1. copper electrode, very efficient heat transfer, wont glow or cause pre-ingitions like the iridium and platinum plugs tend to do when using this fuel, E85 is supposedly 8 times more sensitive to this then normal pump-gas.

2. colder, 9 instead of oem 6, is chosen because we really dont need the plug to be so hot that it burns of soot and other layers, the E85 will clean it more then enough, and then good heat transfer to the cyl-head = only positive.

3. wider gap, from oem 0.8mm to 1.1mm, for a stronger and wider spark gap, these ignition-coils can handle it with very positive results.

4. single electrode, for a good and strong spark in the center, plugs are good enough for ~60000km imo, and cheap compared to the oem platinum ones, infact very much so, might even swap em after 30000km..  :)


just thought id throw this info in here if someone were interested, even tho it is a bit off-topic, it helps alot to have these things sorted before tuning the fueling and cold starts.. as with the oem plugs i noted pre-ignitions, surge/hesitation and popping sounds in the exhaust all the time, and also had to crank even harder to get the engine running.

Have you looked at KFZWWLRL for timing?

completely missed that one, ive been looking for something similar, very usefull, thanks!  :)

any suggestion as to increase or decrease the timing advance during cold starts with this fuel?


..also KFZWWLNM looks very interesting, is this map allways active?

Decrease

Anyone tune FZWSTTM for this purpose?

i have it defined, and i actually dont know WHY i havent been using it for e85.. as i was using it earlier on my 91oct tune..
i will give it a go.. but KFZWWLRL did smooth things out real nice in cold weather.

currently running e85 in -10degC, fires right up on first crank  :P
if i wait ~20sec, i can drive from the parking-lot real smooth and without any hestitations at all..
BUT if i fire it up and try to go right away there is just no power there.. im thinking i would have to increase the warm-up fueling by ~300% to get only the gasoline part to ignite to cure this problem.. as the ethanol part is having a hard time igniting untill cumbustion chamber is warm.. other suggestions?

i have a set of special e85 injectors going into the engine real soon.. with extreme atomization for this fuel.. ill see if i can get the ethanol part to ignite at lower temperatures by reducing dropplet size and allowing it more surface area.. aswell as a better spray-pattern.

You are all aware of oil dilution are you? The "enrichment" doesn't go out the exhaust.

What is the optimal lambda for these really cold starts? Mine is at 10-11AFR (gas gauge) and it takes about three starts with pumping the gas above 2000rpms until it settles. I'm leaning it out 10% increments but no luck so far.

Whats that? you mean fuel ends up mixing with lunricating oil?..or it gets trapped in catalyst?

It ends up in the oil, which is detrimental to it's lubricating/cooling properties.

But, why should it end up in oil, if the sealing at piston ring is proper.

More fuel should not mean that it will go in the crank case...

The rings are not that tight when cold.

how about the target AFR?

Do some googling of "oil dilution" "incomplete combustion" and "cold engine"

Thanks for info

Unburnt fuel also washes the cylinder walls of lubrication, causing top end wear.

Would such massive enrichment be needed at like +5-10*c?

Planning on running e85 at summer and 98 at winter, but early/late summer it will go down towards like +5*c.

http://papers.sae.org/2011-32-0582/

at around 10degC, no problems at all using the normal gasoline enrichment levels for coldstarts.. at 5degC, about 15-25% extra needed the first 20-30 seconds for a smooth start in first crank.. (this using normal injectors.)
so i wouldnt worry in those temps.

yesterday it was -24degC outside, and with some tweaking of the tables i got the car to fire up on the 4th crank..
no power was present when i tried to pull out of the parking-lot tho..
waited ~30 seconds for the cumbustion chambers to heat up and it was a smooth ride from there.. im thinking a small cache-tank with fuel pre-heating and circulation thru the rail before start might be good to implement for these temperatures.. i have an idea on how to solve this..! =P
(warm fuel = wont have to enrich more then with gasoline = VERY good for the engine, starts and economy)

Is there any reason we can't get %STADAP to handle startup enrichment?

If you are maxing out the enrichment adaptation, then the limit can just be increased.

Only range zero would have to be tweaked I think, maybe adjust the temp ranges themselves slightly too.

hmm, care to elaborate on STADAP works?
i think it would be great if this can be used.. from what i can tell by a quick look in FR, the ecu will "learn" the best startup enrichment itself?
i wonder how that works for different temperatures, as the ethanol in this case is far from linear.. if it could handle this, then yeey!  :)

were talking pretty massive enrichments here for -5 degC and downwards..

im designing a fuel pre-heater to eliminate the need for that massive enrichment in this -25-30degC weather we have had the last couple of days.. as in -5> there is really not much enrichment needed.. should radically improve fuel economy and driveability.. as well as no unburnt E85 ending up in the oil-pan.

This function creates an adaptive factor, kstaa, for the injection factor at start fst_w.

Coolant temp at start, tmst, is evaluted into three temp ranges, zero, one, and two. Each temp range has it's own adapted value. There are also interpolating factors, WAD*T*, to allow a percentage of influence between ranges.

If the number of ignitions during start reaches a threshold, ZSYNAUSW, then the number of combustions is checked against ZBURNSOL and ZBURNDEC and the decision is made to increment kstaa for too lean or decrement kstaa for too rich. The amount to adjust kstaa is determined in DECKSTAI (amount to decrement) and INCKSTAI (amount to increment) based on temp range.

I think that the only a few small things might need to be adjusted. I'm not sure of the temperature ranges that cause problems and if it gets increasingly worse as temps drop. I was thinking have range 0 exclusively handle the problem range, but if there is a large difference within the problem range, then you could have range 0 for extreme cold and range 1 for slight problem temps. Range two would then be for normal operation during the warmer months. The interpolation factors may need a good amount of tweaking to balance everything.

KSTAI : to fudge the temp ranges slightly
KSTAMX0 : the max cap for range 0
WAD0T1 : to adjust range 0's influence on range 1
WAD0T2 : to adjust range 0's influence on range 2

The other interpolation factors may need some adjustment as well and I think that you would want very minimal if any influence between range 2 and the colder ranges 0 and 1.

IMO it will be much better to allow the ECU to adapt the amount of fuel required for startup and dynamically adjust rather than trying to fight for the optimal enrichment.

I've had a look at this function before but had no need to use it as it never gets cold enough here.

It is disabled out of the factory.  Max adaptation is set to 0, and temp ranges are 0C 1C 2C.  Time since start values are a little off too. 


Rick

The values you see in KSTAI are not in *C. The axis is tmst (coolant temp at start) in *C and the map values define which range is active (0-2).

Yes I realise that.  But if you look, some values for active range are set at 255

Yep...these are inactive temp regions. Anything other than 0, 1, or 2 will not be active.

What are these axes about in KFFFANZ? They should be nmot and tmot?

(https://lh6.googleusercontent.com/-16JdySTqmJU/UTYa7DGZfLI/AAAAAAAACBM/PNpKOhuG7uw/s941/KFFFANZ.GIF)

I figure this STADAP only affects non start issues and  has no affect on warm up adaptation. I have no problems on initial start but the engine stalls right after the start if I dont give it plenty of throttle and keeping it over 2000rpms. This is at colder than minus10C. My KFFWL tables are maxed already so now Im looking for msd, KFFWLW and Stadap.

I should add: ev14 96# @4bar and E85

The adaptation factor is also used in afterstart enrichment.

This function is not the right way to handle E85. Calibrate your start for the fuel you use. You need a different map for gasoline and high ethanol fuels anyway. Using all the adaptation stuff will not get optimal results. OEM calibrations have to use it because of the big fuel quality span you get in different regions of the world. If you know "your fuel" there is no need for that.
STADAP is typically used for fuels down to E25. Higher ethanol contends use other adaptation strategies because the difference in behaviour compared to regular gasoline fuels is just too big.

As I have mentioned before I don't think it's the best idea to use E85 on such cold ambient conditions because of the oil dillution problem. If you don't warm up your engine really good (oil warmer than 90°C for at least half an hour) you can literally see your oil level go up. You can change your oil after 2000km then.

Thanks
As I have already maxed my KFFWL_ tables next calibration is the weighting MAP KFFWLW. I added 20% over zero temps and now I need to wait for the next really cold morning.

I have eberspaecher engine/interior heater but I'd like to get this sorted when or if my heater decides not to function. This morning was only  minus 1C and no problems but I only have that 20% on KFFWLW and MSD active under 1400 rpms, but the KFFFANZ is a bit mystery. I have the 40 and 80 columns at 2 sofar.

E85 is the most economical blend over here so thats why I use it over winter too.

Started and run just fine at minuc 16 C,  which is good as the heater started to act up yesterday and only fired on 3rd try. It would be nice to know which change affected the most, msd or stadap.