tianporn

Hi there. I hope you guys can help with my issue.

Motor is all motor JDM B18C ITR wits basic I/H/E mods
Running OBD1 ECU tuned on Crome - street and dyno

OK here's the issue...

Car was tuned with a reputable tuner and we tuned the car open loop, knock sensor disabled, O2 sensor disabled (which is best practise).

The tune was all good. I was there with the tuner. AFRs were safe, timing was good and car ran pretty good on the dyno. Once the tuning was finished the tuner re-enabled the O2 sensor but kept the knock sensor disabled and burnt the chip and that was that.

The issue now is that I'm hearing a lot of detonation (pinging/knocking) at partial throttle (10-20% throttle) from 1500rpm up to VTEC. This ONLY happens on light/partial throttle. At WOT there is no knocking what so ever. Also when the car is cold there is no knocking at the specified RPM range.

From these observations it appears to me that when the ECU is running closed loop, it's throwing the AFRs off? It almost seems counterintuaitive to me the O2 sensor would lean the car out so much?

Any thoughts?


Cliff Notes:
Car tuned with Crome in open loop
AFRs and timing was fine during tune
Map was burnt to the chip with O2 enabled (closed loop)
A lot of knocking can be heard at partial throttle which did not occur during the tune

OneBadTurboCRV

You can pull the plugs and read them for detonation to know for sure. If you don't have a wideband AFR gauge then you should buy one or take it back to your tuner. Until you know for sure what is going on with your AFRs then there's not much to say other than stating the obvious. Either the O2 closed loop settings aren't correct or something else is going on with the fuel system or the likes. A good tuner would have verified that closed loop was operating properly if that was going to be how the car was left.

tianporn

I've done so research. I can find similar occurances with Hondata users but nothing on pgmfi?

http://www.hondata.com/forum/viewtopic.php?f=19&t=16281&sid=1906ca558d2510652b5 2910c234fa41e&start=0

http://www.hondata.com/forum/viewtopic.php?f=19&t=12827&start=0

tianporn

Bump

JKTUNING

Did you guys actually log the short term fuel trim from the o2 sensor?

If the sensor is bad or un-responsive and it "could" be pulling a bunch of fuel out it would certainly cause the car to go super lean under low load conditions.

tianporn

I don't think CROME logs short term fuel trim.

Unless you are talking about checking A/Fs while running closed loop?

motoxxxman

bump
I'm having this exact same issue with my car. the tune is nearly flawless in open loop through most weather conditions, but the instant closed loop is enabled it goes stupid lean. pr4 ecu tuned on crome. I've tried using my lc-1 to simulate narrowband, and have also tried a used oem honda sensor and bought a brand new oem o2 sensor, and the results were the same with all 3.
I cant seem to find anything that provided a fix for this for crome users. I mainly just want an active o2 for DD'ing when humidity levels change, as that seems to be the only thing throwing off my af's.
anyone know the cause of this or how to solve it?

1990 prelude with built b21a1, super 60 t3 @9psi, stock map sensor, dsm bluetop 450cc injectors, ELD disabled as my car does not have ELD
I did a lot of testing with all different values in the ect fuel compensation, making changes to max bounds, closed loop, and 0016fh, and nothing made it better, some of those changes actually made it worse

ddd4114

Are you sure the closed-loop feedback isn't simply just removing enrichment you had in certain areas? The ECU will always try to maintain a stoichiometric mixture in closed loop, so if you were trying to run richer at higher loads in open loop, it would cancel that out.

Can you provide some information about the AFR in closed loop vs. open loop for the areas in which you're experiencing knocking? Can you be specific about the engine speeds and loads that have a knocking problem?

joe b

Just run the car in open loop .

motoxxxman

I set closed loop threshold by throttle position obviously varying by rpm but to be right at 9"/hg vacuum at all rpm's. when I run in open loop it's rock solid 14.6-14.8 from 25"/hg to 11"/hg also at all rpms. idle varies from 14.2-14.8 depending on weather and oil temp and iat sensor heatsoak, but is usually right around the 14.6 area. beyond 9"/hg it jumps to high 13's, then a gradual drop to 12's around 2-3psi of boost, and 11.8-11.9 at all boost levels above that up to max of 9.5psi.
when I run in closed loop it immediately reads 16's to 18's at all throttle positions and manifold pressures below 9"/hg, but 8"/hg and higher is identical to open loop operation. occasionally it does try to compensate and gradually dips to 15's then freaks out and goes back to 16's-18's again.
Ignition timing is spot on everywhere, or close enough to to show zero signs of heat or det or low power; according to 2 different dyno's, reading plugs weekly and after any spirited driving as well as after long highway cruises, and according to the borescope looking at pistons and valves.

I have been for years because I can't sort this issue out, even with multiple ecu's. and I already stated why I want to run closed loop. I'm sick of tweaking the tune on such a regular basis due to new england's constantly changing humidity levels, and typical variances in fuel quality and such. lately iat sensor heatsoak has become an apparent issue as well. still working on figuring out how to prevent that without relocating the sensor to a spot that will give false readings

ddd4114

Have you tried datalogging O2 voltage when using a narrowband sensor? What does it do when you're seeing 16-18 AFR in closed loop? I wonder if you have some kind of calibration or electrical problem.

7rrivera7

iTrader: (1)

Why not try out another tuning platform like Neptune?

VegasInvasion

if it's properly tuned there's virtually no reason to run closed loop, and the stock O2 is useless on a performance engine.

if you're using a gold rom, which is the only valid option for crome, then you can run closed loop with wideband settings and it performs perfectly.

read my manual for better details.

Black R

What is the closed loop target a/f ratio for you guys having trouble?

Make sure it's not set at those lean values!

motoxxxman

the o2 voltages correlate perfectly to a honda narrowband signal when datalogging through the ecu, both with the oem sensor and with the wideband simulating narrowband (never simultaneously of course). when it does occasionaly richen up to stoich, voltages hover around 0.44-0.50, then when it kicks back to 16's-18's the o2 voltages are also instant to show super lean.

I'm a broke bastard lol. when I have the extra cash, I'll definitely be switching to s300 v3. but crome in open loop is mostly fine for now, besides the constant tweaking to the tune when humidity changes.

how is a stock o2 useless on a performance engine? seriously, I'm curious. the o2 is only active when the engine is running within the programmed closed loop parameters, and I've properly adjusted those settings to have closed loop only be active below 11" of vacuum.
and how is a gold rom the only valid option? I've been tuning with crome for over a decade, loooong before gold roms even existed. gold roms obviously open up a whole new world of parameters for tuning, but I've never had this issue before with any other vehicle using standard roms; only my car.

14.7 from full vacuum to 11", then tapering down to 12.8 at 0"/0psi, then to 12.0 from 4psi to 8psi, and 11.8 past that

VegasInvasion

The ST/LT correction on an 8 bit ECU is so unbelievably slow and basic that it's more likely your open loop values are more accurate anyway, and more dynamic.
Stoich is not the most effective part-throttle AFR, it's a smog standard. Sure, you can set the CL target to 16:1 if you don't mind idling at 1200 to keep it from dying.
The P30 scalars are crap, and the QD2 protocol is too slow. You of all people should know that if you're been using Crome for 10 years. It might fly on a stock engine, or mild boost. But no reputable tuner relies on a rom with those kind of limitations on a built engine.

motoxxxman

ok, good to know, for the parts I can understand at least. my engine is non-vtec and limited to 7k-7200 and pulls up to 26" of vacuum, so I always re-scale the rpm's and manifold pressures in the actual tables to correlate better for the peaks and valleys of my torque curve (leaving the lowest and highest untouched). I have tried the stock table scales to see if there was a connection, but no change, so I was able to rule that out.
I also run only 9psi at the moment, which is why I'm still sticking with crome for the time being. still working out some bugs in the design of my motor and correcting a few mistakes I made and the machineshop made during the build of my current engine before raising the boost.
the part I dont understand though is QD2 protocol. what is that?

also a random possibly bogus question, but is there a such thing as or an ability to wire up a multi-position switch (potentiometer type) to make a universal fuel correction? like a way to alter LT trims on the fly for an ecu running fulltime in openloop? (not an afc that hacks the map signal and alters ignition timing as well)

OneBadTurboCRV

LOL... you CAN'T be serious...

ddd4114

I didn't think it was such an unreasonable question. People have been doing stuff like that for years before real-time calibration became common.

I wouldn't recommend doing something like that for long-term use simply because the risk of it not working properly could cost you your engine. I actually don't think it's even possible in Crome Pro, but I'm pretty sure you can make something like that work with a Gold ROM. You can use a pot to adjust a 0-5V signal going to an available ECU pin (or 0-3.8V for the O2 pin) and configure a table to adjust the overall trim based on the voltage. It's definitely a band-aid fix, but it's possible.

OneBadTurboCRV

You must be focusing on the last part of his post. I was talking about the whole thing. For someone who has been using CROME for a decade you would think they would know about the protocols for data logging, messing with the scalars can cause corruption (I have had it happen on multiple occasions without touching the beginning or ends), and the 9 psi part. I have been hack tuning my Honda stuff since about 2002 and understand the stages that tuning has went through. Not that I am a big time O.G. or anything because of that, but it's on par with his decade claim.

ddd4114

Yeah, I thought you were just talking about the last part. Whatever the case, he doesn't seem like a typical internet jackass to me, so I think he deserves a little respect. He's open to feedback, writes coherently, and wants to learn, and that's better than 95% of the bullshit I read on this forum. That being said... I did roll my eyes at the 9 psi comment, but I didn't want to derail the thread and bring attention to it (like we are now).

OneBadTurboCRV

Sorry...

motoxxxman

thank you guys, I do appreciate that. I'm here to learn for sure any way you look at it. that is why I asked about this whole situation afterall lol, to see if anyone has made any headway on the topic at hand. and I'll be the first to admit experience isnt everything, especially when it comes to tuning. I have been tuning with crome for over a decade, but there will always be things I don't know that others hopefully will know that I can learn from. while I do agree with ddd's 95% comment haha, I make no claim to fame

to expand a bit, my car is my DD when the engine is healthy. I say that because I use my personal car as a test subject for all sorts of various things to see what parts and procedures work or dont work, and what each of those things' actual limit is roughly. which of course equates to engine and drivetrain damage on occasion, which I'm ok with because thats kind of the point of it all lol.
I'm also the only person in the US that I'm aware of that runs higher than 9.5:1 compression with any amount of boost through our bastard B engine (b20a/b21), on pump gas nonetheless. the engine is not an excellent platform for high power, but the chassis sure is especially for an autox and circuit track car. which is why I have only been running 9psi (sometimes higher on the street) in the 2 chassis' I've raced; 252whp/249tq with nearly zero spool time in 2nd gear and up is pretty nasty on a course and city streets with only 2200lbs on the pavement.

as for datalogging protocols, I've never had an issue with datalogging except the one time I had to reinstall and forgot to change the com port's latency. so I've never really had to do any research into the protocols and am unfamiliar with what QD2 is. if you can link me to a good resource on that, I'd love to find out.

and the scaling part, every time I've changed any of the lowest or highest values, corruption has occurred instantly. I knew that before I ever even tried but wanted to see for myself. all values in between though, I've only had a corruption occur from changing them 2 times over the years. luckily I keep backups of everything, so I simply started from scratch and copy/pasted from the previous save, then changed all the scale values to the same values that corrupted it the previous time, and it never happened again. I blame crome glitches lol.

I however am still rather determined to find out what's causing this CL issue though that apparently many other people have also experienced. I still cant figure out if its my chassis or my ecu (pr4 board). I've only had the opportunity to try one other ecu (p06 board), which is also a 1720 like mine.

can anyone who has the same issue as myself confirm what board their ecu has? whether its a 1720 or not?
also, do any of you suspect it may have anything to do with ELD not being present, requiring it to be disabled in crome?

my car is a 1990, but is all obd1 from the factory, with the exception of not having ELD. also to clarify, my distributor contains only the tdc and ckp sensors, while the cyl sensor has it's own separate housing on the exhaust cam. I have verified though that tooth count and phasing for each is 100% identical to a p72 distributor.

motoxxxman

aaaaand, I totally forgot I had actually started my own thread on this topic in 2013, with some info I forgot about, but very little help. there's a vid too, and several posts containing my awkward finds and questions. if you guys could read through it quick and check out the video and post back here in this thread, I'd much appreciate it especially relating to ELD, and if it's possible to wire ELD into my car, and any relation it could have on this whole issue.

linky:
https://honda-tech.com/engine-manage...-tune-3169975/

Natural Aspirations

The ELD "should not" have an effect on AFR's.

ELD info Click Me

I suspect that you have the fuel multiplier messed up somewhere.