BRN12345

I have tuned several Honda's using Hondata and have always wondered why in the upper rpm ranges timing is pulled 2-3 degrees...for example would go from say 30deg at 7000rpm to 27deg at 9000.

I dont have access to a dyno that i can use for testing such things, and want to know if anyone has tested retard on the top end VS a flat timing curve to redline.

Turbo-charged

thats 100% the opposite of what should be happening regardless of what engine management your using.

b19coupe

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by Turbo-charged &raquo;</TD></TR><TR><TD CLASS="quote">thats 100% the opposite of what should be happening regardless of what engine management your using.</TD></TR></TABLE>
I disagree. What kind of dyno are you using? We use a Dynapack and have found that the resolution can highlight the gains/losses of a 1/4 degree change in ignition timing. Our experience has been that a typical B-series motor wants peak timing at around 6900 rpm or so, and then rolls off above 7,700. We are only talking about gains of 1-3 hp here, so a Dynojet or a Mustang dyno may not offer the resolution to be able to discern these changes.

Next time I have the ITR on the dyno I will perform a little ignition timing test and post the results for all to see

kommon_sense

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by Turbo-charged &raquo;</TD></TR><TR><TD CLASS="quote">thats 100% the opposite of what should be happening regardless of what engine management your using.</TD></TR></TABLE>

Wouldn't it also depend on the fuel map and how much fuel was added at those points. For example, if the fuel curve starts to drop off then there would be less fuel to burn there.

Turbo-charged

VE will always follow the torque curve...torque is always going to drop at higher rpm, so your VE will drop as well. as torque drops, ignition timing should be brought in. ignition timing should not peak until after you have reached peak torque. therefor you are going to have less ignition timing at peak torque than you do at higher rpm.

lower power cars are pretty forgiving.....try to give a 800+ horsepower car too much timing in the mid range and it wont last very long.....

SOHC_MShue

Theres also that whole backround honda timing correction to consider that already adds in a boat load of timing based on rpm and increases with rpms. If you are seeing this then timing may be increasing in actuality, but not reflected in the maps.

Turbo-charged

very true. kind of a scary thought actually...

b19coupe

Had my car on the dyno and I put the timing theory to the test. The results are a little inconclusive, but I will post them anyway. Three runs, back to back, same intake and coolant temps at the start of each run-coolant at 190 and IAT's at 105.
The red run is with the timing at 30 degrees at 5500 ramping up to 31 degrees at 7000 and ramping back down to 30 degrees at 9,500.
The green run is with a flat 31 degrees from 5500-9500.
The blue run is with timing at 30 1/4 at 5500 ramping up to 32 1/4 degrees at 9,500.
The gains/losses are small-the biggest gain being 1 ft-lb. at 7,000 rpm on the blue run. Ironically, all three maps are running 31 degrees at 7,000 rpm.

The final timing map that made the most power had 30 1/4 degrees at 5500 rpm ramping up to 31 degrees at 7,000, 7,400, and 7,800 and dropped to 30 3/4 degrees from 8,200 and up. More or less timing lost power.
It would be tough to discern these changes on most dynos, but the Dynapack's resolution allows us to see very minute differences

Turbo-charged

what EMS?

b19coupe

Hondata S300.

Turbo-charged

i love s300 to death....but there is way too much "behind the scenes" stuff going on to accurately compare such minor changes IMO. i think your 7k rpm, same timing, and difference in the graph shows that.

none the less its cool to see your results. 1/4 of a degree, IMO, isnt really enough of a difference to say that the difference in timing was the result of the power difference in your final map from any of the others.

ive never used a dynapack. i hear they are very nice to work with though

baller status

So Hondata doesn't have a parameter for the actual output timing after all the corrections? WEAK.

I'm pretty sure the factory ignition maps will show the max timing output and that all the corrections PULL timing except perhaps for idle which will go either way to keep it at the target RPM. If the factory ecu advances timing beyond the ignition at all, I bet it would only be on knock controlled cars, mostly OBD2 and a few late model OBD1's. But if I had to put my money down I would say that even the knock function PULLS timing from the map.

I'd be interested to see some results on a load bearing dyno ie Mustang. If you were brave enough to hold it at those high RPMs long enough you could see the changes in power in real time instead of from one pull to the next.

I'd also like to see some bigger timing changes. Once you're at MBT there isn't any power difference for a few degrees in either direction.

I'm just talking out my *** though. b19coupe is clearly more qualified than myself.

SOHC_MShue

its been proven that there is a timing advance that raises with rpm in the stock code. There is a function to disable it in some programs.

Turbo-charged

shameless plug?

SOHC_MShue

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by Turbo-charged &raquo;</TD></TR><TR><TD CLASS="quote">shameless plug? </TD></TR></TABLE>
lol not really just trying to show that there really is something going on in the backround

i think ectune has it as well.

sewell94

eCtune has it as well

BRN12345

Since the dyno results are showing no real change in power.. another question comes to mind..

In an engine that sees constant high rpm use, would it be better to have less or more igntion timing up top if power does not change with timing change?

i.e. would retarding timing up top like in a stock itr calibration for example help in engine cooling? or engine longevity? or maybe has less chance of detonation?

Turbo-charged

in that case it would be "safer" to keep the timing advanced up top....but drop it down in the midrange.

most tuning related engine failures happen at peak torque. just because you may not notice the damage until the upper rpm's.....peak torque area is where the damage usually starts happening and then it goes down hill from there.

also if you add timing and you dont see a gain....you have already achieved mbt, or possibly even passed it. using factory ecu based tuning programs i keep timing very conservative because you dont have complete and total control of it.

98vtec

if you dont see a change in power after adding ignition, i would reduce until it loses power then add a degree to put it back into the peak. Less is more, IMO.

blundar

I love Dynapacks, but you have to be really careful with them. They're my dyno of choice, but certain aspects of how they operate are troublesome at best and scary at worst. It all boils down to ramp rate.

With a Dynapack, as part of the F11 run setup you must enter a "settle time" that controls how long the dyno will apply sufficient force to hold the engine at the start RPM. I don't know how much attention you have paid to the real time torque output while the dyno is settling, but it is all over the place. (like a couple percentage points - far more than the spread on your tests, b19coupe). Point #1: systems derived from OEM Honda code (Hondata, Neptune, eCtune, Crome, Uberdata) behave totally differently in dynamic conditions (i.e. acceleration) compared to steady-state.

One of the greatest strengths (and weaknesses) of a Dynapack is the way it regulates ramp rate during a run in a linear manner. This is a great strength because you can essentially "zoom in" and see the behavior of a motor at a great resolution. This is a huge weakness because it is trivial to produce conditions that bear no resemblance to how the car behaves in reality. Easiest way to do it: pull from very low RPMs to redline with a turbocharged car and a turbo that does not spool till close to redline. The dyno will limit the car's acceleration to maintain a constant rate of acceleration, regardless of how much torque it is making. In the real world, the car's rate of acceleration would be a function of how much torque it was making combined with speed and various other factors. For a system that does dynamic corrections based on acceleration, you're [freak]ed if you tune too close to the edge. Moral of the story: failure to choose appropriate ramprates and/or windows is critical.

Dynojet's are not without issue either. The rate of roller acceleration is a function of torque, which at face value is better than a Dynapack. Unfortunately, you have no way of controlling ramp rate so you're stuck with letting small horsepower engines do very long pulls and high horsepower engines do very short pulls. Also, how fast the roller is spinning (changes in frictional losses that do not factor into Dynojet math is my explanation for the SHAPE of Dynojet torque being higher at higher rollerspeeds compared to a Dynapack - and yes I ran the same car on the same day on both Dynos) seems to have quite an effect, too.

Supposedly, Mustang has a "virtual inertia" feature that tries to approximate the rate of acceleration of a vehicle but I have yet to play on one of their dynos that has this feature.

Bottom line: if you use a dyno that does not accurately mimic how fast the car will accelerate on the street/track, you will have to worry about dynamic changes to ignition based on conditions you cannot accurately simulate. Timing settings that may work appropriately in one gear / acceleration rate may cause issues in another gear/acceleration rate due to differences in dynamic tuning.

Also, it is worth noting that K-Pro is significantly worse in this repect than any OBD1 systems. The largest completely unaccessible portion of the tuning information present in the stock K series car is a whole series of routines loosely termed "catalytic converter protection." Think about it: the cat is practically the third most expensive engine component, after the engine/transmission. It makes perfect sense to try to control engine conditions to avoid melting down or fouling a cat that would have to be replaced under warranty. As of the last time I looked at the K-Pro software (today) there was no option to disable or modify any of these corrections. Don't believe me? Try doing pulls over and over and over again and watch your AFRs plummet, even with closed loop turned off. That's not a wack IAT trim...


Modified by blundar at 2:29 AM 6/4/2008


Modified by blundar at 2:30 AM 6/4/2008

SOHFAST94

I have to agree with you. I do factory calibrations (OEM) and Catalytic algorithms are just as important to the calibration as ignition timing and fueling to the OEMs and without being able to modify or turn the functions or scalars off that can be a very dangerous and troublsome thing. Strange I never thought of this before while tuning Kpro. I guess I just assumed that by turning off "OBD2" covered functions such as this, I guess that's not the case. Kpro just went down a couple of pegs in my book

dmotoguy

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by blundar &raquo;</TD></TR><TR><TD CLASS="quote"> Point #1: systems derived from OEM Honda code (Hondata, Neptune, eCtune, Crome, Uberdata) behave totally differently in dynamic conditions (i.e. acceleration) compared to steady-state.</TD></TR></TABLE>

with eCtune and Neptune, if you have the background calculations turned off, would this still be the case?

blundar

doesn't matter. you can't tune every cell in steady state and expect the car to run remotely close in acceleration. It will be off, every time, every system.

Garrett

That's a good point Blundar

with Hptuners (for GM OBd-2) I can turn off the Cat over heat protection and it help a lot

SOHFAST94

Yeah the only problem with that is when you overheat the cats and melt one then what?