johnyblaze

How do you tune using a piggy-back controller (like GReddy E-Manage) while the stock ECU is in closed-loop and still adjusting the long term fuel maps?

We're struggling with this, trying to tune with a wideband O2 on street and dyno, but we can't reproduce the AFRs deterministically between runs because the ECU is still "learning" the fuel curves and adjusting them. Of course this is a non-issue for WOT tuning, but I also need to tune while not under load so that day-to-day driving works too.

What are people doing to circumvent this? And what happens if I complete a tuning session only to have the ECU continue its learning cycle and re-adjust the long term maps?

Another thing to note too is that I've been having to reset my ECU for various things (mostly to clear OBD-II rear O2 sensor codes) which means my ECU never gets to complete its learning cycle.

This is for a 2002 Honda Civic Ex (D17a2) with upgraded 370cc injectors. I'm running 8psi boost. We're using the e-Manage airflow map and the additional injection map for fuel tuning. The axis for the maps are as follows...

Airflow: Adjust MAP voltage according to TPS & RPM
Additional Injection: Add injector duty cycle according to MAP voltage & RPM


What else is odd is that we're NOT clamping the stock MAP sensor voltage and the stock ECU is NOT complaining -- Honda ECUs are notorious for throwing codes when MAP sensors see boost. Anyone know why I'm not seeing MAP codes? I am however seeing a P1106 code indicating out-of-range voltage from the barometric pressure sensor (which is located in the stock ECU casing). This may or may not be related to not clamping the MAP voltage. Has anyone else seen this or know how to resolve it?

If you're interested, here's some info on Honda ECUs' learning process: http://www.hondata.com/techclosed.html

Thanks in advance for any help!

--J.

Marauder

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by johnyblaze &raquo;</TD></TR><TR><TD CLASS="quote">
What else is odd is that we're NOT clamping the stock MAP sensor voltage and the stock ECU is NOT complaining -- Honda ECUs are notorious for throwing codes when MAP sensors see boost. Anyone know why I'm not seeing MAP codes? I am however seeing a P1106 code indicating out-of-range voltage from the barometric pressure sensor (which is located in the stock ECU casing). This may or may not be related to not clamping the MAP voltage. Has anyone else seen this or know how to resolve it?
</TD></TR></TABLE>

I know that in '99, Honda changed the point at which the ECU will throw a CEL for boost. Before that, even with 1 psi of boost the ECU would throw a CEL. '99 and later civic's seem to throw a CEL for boost around 10-11 psi of boost.

johnyblaze

That would explain it.

What about the P1106 BARO code? Anyone familiar with this, or with the BARO sensor in general?

It's rarely talked-about.

Marauder

I think I've looked up the barametric pressure code for someone before, the only fix is a new ECU.

johnyblaze

That's interesting. But what would cause failure of this sensor? It's built onto the ECU board!

Could incorrect E-Manage wiring or configuration affect this?

I wonder what the local Honda stealership will say if I need a new ECU -- don't suppose it'll be covered under warranty still, even though the wiring harness is all spliced into

Marauder

The test procedure for the barometric pressure sensor is:

1) reset ECU
2) Start the car and hold at 3000 rpm until radiator fan turns on
3) Connect SCS service conector
4) Test drive with A/T in 2nd gear, M/T in 4th gear
5) Accelerate for 5 seconds using wide open throttle

Does DTC P1106 come up?

Yes.....swap the ECU, if problem is fixed, leave good ECU in the car.

No.....intermittent failure

Think the BARO sensor is just internal to the ECU. I have no idea if or how the e-manage would screw with it.

Boo_Daddy

Hey man I am having the same issue. I have the emanage with a B18B engine and a chipped p28 ECU (chipped for a ls)

did you ever get your ecu to stay in open loop?

Thanks!

johnyblaze

I have not, and I haven't found a good answer for this question. Argh -- very frustrating.

I'm considering moving to standalone, if I can't get the deterministic control I need with a piggy-back controller.

scttydb411

baro sensor typically comes on when the map voltage is corrected like it is typically with the hack. it trips when the map and baro voltages are not equal and trips thinking there is something wrong. doesn't seem to affect timing or fueling on my setup and it's on all the time for baro. i just check the codes periodically to make sure.

i'd also like to know how to trip the ecu into full time open loop if it's possible.

johnyblaze

I suspected that MAP voltage clamping and the BARO code were related.

So is there no workaround? What if I used a Missing Link check valve to hide boost from the MAP and just used the GReddy pressure sensor with the e-Manage for tuning boost vs rpm?

Regarding forcing open loop, I don't know still. I'm hearing/reading all sorts of crazy ideas like providing a steady 5v signal to the ECU as the TPS signal so that the ECU thinks it's at WOT all the time (and stay in open loop). Not sure what other implications that would have.

Another idea was to use an O2Sim that we typically use to simulate the rear O2 sensor signal but instead use it also for the front O2 sensor signal so that the ECU thinks the mixture is stoich and will stop adjusting. It will still be in closed loop, but at least we'd be able to tune deterministically (since the ECU won't be making adjustments).

I don't know -- both of these seem pretty hokey to me! Your thoughts?

ImAcracker

I know OBDI all you have to do is disconect your o2 sensor and it will be in open loop all the time.

johnyblaze

So is this what you're doing with your OBD-I car to keep it in open loop?

If I were a betting man, I'd bet that pulling the O2 sensor(s) on my OBD-II car would cause the ECU to throw a MIL and possibly kick it into limp-home mode :-(

Anyone with OBD-II have experience with this?

ImAcracker

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by johnyblaze &raquo;</TD></TR><TR><TD CLASS="quote">So is this what you're doing with your OBD-I car to keep it in open loop?

</TD></TR></TABLE>
Yes that is what Im doing on my OBDI to keep it in open loop.

Del_Slowest

i guess i should start thinking about how to do this also.....


Modified by Del_Slowest at 1:07 PM 3/28/2004

opto_isolator

Yo JB - yes its me, the one who made all of those "hokey" suggestions

Anyways - I suggested using a O2 simulator for the FRONT O2 sensor because it is the one which controls the Air/Fuel mixture (as well as the MAP and IAT). If you run an O2 sim for the front, you can tune it to stay at stoch all the time.

The only downfall of this is that the car may run crappy some of the time due to the fact that changes in temperature and pressure can affect the A/F ratio...As far as I've been reading.

I've attempted to email Kenji Sumini from Greddy, posing as an interested RSX turbo kit buyer. Greddy has gotten the RSX kit to work with an ODB-II ECU without any issues. As to how, I don't know. Kenji is the Greddy Engineer who actually flew around the country fixing their kits....However he has been not a big help in the past when I've mentioned that I am trying to use the E-manage on a 2001 civic. I asked him exactly how they have gotten around the issues with closed-loop tuning. Hopefully he will answer his emails!

BTW, are you having a shop do the tuning, or are you doing it yourself? If they are an authorized reseller of greddy products than technically they can get support from greddy.....

Good luck, keep us updated!

Joseph Davis

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by johnyblaze &raquo;</TD></TR><TR><TD CLASS="quote">How do you tune using a piggy-back controller (like GReddy E-Manage) while the stock ECU is in closed-loop and still adjusting the long term fuel maps?</TD></TR></TABLE>

You can't.

In my experience, you can GUESS by monitoring your O2 sensor output... if the base fuel pulse is set very high, the closed loop rich/lean switching action biases itself toward the rich side. AFR still goes lean, but due to excess fuel it takes a hair longer to switch to lean than it does to switch from lean to rich. I've picked up a few mpg by fiddling low throttle fuel trim in the bad old days of the AFC hack... just keep in mind that part throttle boosting is now a no-no. Part throttle boosting should be a no-no with anything OBD2, anyway. OBD1 can be problematic, but is fine with OBD0; OBD0 sticks itself in open loop if the MAP sensor input changes quickly enough (spooling up) or if the engine is accelerating fast enough (17% throttle, 6 psi should do it).

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by ImAcracker &raquo;</TD></TR><TR><TD CLASS="quote">I know OBDI all you have to do is disconect your o2 sensor and it will be in open loop all the time.</TD></TR></TABLE>

Nope. OBD1 vehicles stumble at cruise with the O2 sensor disconnected. If you bothered to monitor your disconnected O2 sensor, that the ECU isn't seeing, you would know that since the ECU ASSumes a lean condition and dumps in a shitpile of fuel. If the ECU responding to the O2 sensor (or what it ASSumes is the O2 sensor) isn't the definition of closed loop, I don't know what is.

The only way to disable closed loop in an ECU is by fiddling the ECUs code. If you can do that, then a piggyback hack is an expensive waste of your time.

opto_isolator

First thing I would do if I were you is to get a MIL for the second O2 sensor so you can get rid of the damn error code.

Do you have an OBDII reader? If so is that your method of monitoring the ECU? I'd be interested to know how much the A/F is changing...

What specifically did Greddy say to you about using the injector correction change parameter? The main reason to use it is to help with tuning in general. According to their manual, you make your adjustment with the injector correction, tune your car, and if you ever need to swap injectors just change the injector correction and you shouldn't have to retune. All it does is chop the voltage from the MAP sensor from the get-go vs having to go through the airflow map and change all of the values (which there are a lot of).

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by J. Davis &raquo;</TD></TR><TR><TD CLASS="quote">

You can't.

In my experience, you can GUESS by monitoring your O2 sensor output... if the base fuel pulse is set very high, the closed loop rich/lean switching action biases itself toward the rich side. AFR still goes lean, but due to excess fuel it takes a hair longer to switch to lean than it does to switch from lean to rich. I've picked up a few mpg by fiddling low throttle fuel trim in the bad old days of the AFC hack... just keep in mind that part throttle boosting is now a no-no. Part throttle boosting should be a no-no with anything OBD2, anyway. OBD1 can be problematic, but is fine with OBD0; OBD0 sticks itself in open loop if the MAP sensor input changes quickly enough (spooling up) or if the engine is accelerating fast enough (17% throttle, 6 psi should do it).</TD></TR></TABLE>

Some how greddy managed to at least get a base tune for the RSX though....which is interesting....

Joseph Davis

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by opto_isolator &raquo;</TD></TR><TR><TD CLASS="quote">Some how greddy managed to at least get a base tune for the RSX though....which is interesting....</TD></TR></TABLE>

My point was you can't adequately tune your close loop part throttle fuel with a piggyback. Greddy wasn't able to do that either.

ImAcracker

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by J. Davis &raquo;</TD></TR><TR><TD CLASS="quote">

Nope. OBD1 vehicles stumble at cruise with the O2 sensor disconnected. If you bothered to monitor your disconnected O2 sensor, that the ECU isn't seeing, you would know that since the ECU ASSumes a lean condition and dumps in a shitpile of fuel. If the ECU responding to the O2 sensor (or what it ASSumes is the O2 sensor) isn't the definition of closed loop, I don't know what is.

The only way to disable closed loop in an ECU is by fiddling the ECUs code. If you can do that, then a piggyback hack is an expensive waste of your time.
</TD></TR></TABLE>

I always heard that if you disconect your Primary o2 sensor that the ECU will run strictly off of the fuel maps. And I also noticed the when I had my o2 sensor disconected my ECU would not try and correct a bad lean tune in the idle portion of my fuel map. My tune was originally running lean when I started the car and I could see the ECU slowy start to richen it up (this is with the o2 sensor connected). But when I disconected my Primary o2 sensor when i started the car it would just sit there running real lean. The ECU would not richen up the mixture. I then burned a chip with more fuel in the idle portion of the fuel map and could see my changes stay exactly where I put them (when the o2 sensor is disconnected). So im conffused as to why the ECU is not richening up the mixture for me when the ECU ASSumes a lean condition. Is there another reason why the ECU would not be correcting the mixture with the Primary o2 sensor disconected?


Oh. And all these A/F's were monitord using a wideband.

johnyblaze

It starting to sound like closed-loop tuning with the e-Manage is futile and keeping off boost while in closed-loop is the only safe alternative. Pretty disappointing.

I'm considering full standalone AEM EMS or Autronic at this point, though the cost will be more than I can (or am willing to) bear :-(

BTW, I am getting a shop to tune my car and my mechanic has had some luck getting tech support from Kenji @ GReddy as well, but nothing earth shattering. But it was "Matt" @ GReddy who suggested doing away with the Injector Correction feature. He didn't say why, he just said not to use it when we asked why the Injector Correction wasn't working well and how to debug. It may have been because of a corrupt .gsc file though -- we will try it again.

In any case, from everything I've read on this board and from other sources, it really seems that the e-Manage (or any piggy-back for that matter) is not going to be the way to get that perfect tune. If the injector correction works, I think the drivability tune will be fine in closed loop WITHOUT BOOST. With closed loop boosting, I'm sure it'll run lean, or I could richen it up and sacrafice drivability while not in boost (and in closed loop).

Seems pretty futile.

johnyblaze

Oh, and yes, I have an ECU reader to read the OBD-II codes being thrown. But it's a very simple tool and does not let me read more of the ECU than the codes.

opto_isolator

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by johnyblaze &raquo;</TD></TR><TR><TD CLASS="quote">It starting to sound like closed-loop tuning with the e-Manage is futile and keeping off boost while in closed-loop is the only safe alternative. Pretty disappointing.</TD></TR></TABLE>

Keep bugging greddy about it - if they can get the RSX emanage to work, and if the S2000 guys can get it work with superchargers, than surely it will work for a 2001 civic....I'm guessing the key is to tune for part throttle, which is what is so difficult. I've read a few posts on the hondata forums, because they have the same issues. According to them however, to tune properly you need to tune in open loop mode, than switch back to closed loop after you've tuned. Only problem is that our ECU doesn't allow for that (hondata reflashed K20 ECUs will). Like I said, one suggestion is to try an O2 simulator temporarily in for the first O2 sensor while you are tuning, to fool the car that its fuel is at stoch....I'm not sure how well it would work but its worth a shot.


<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote &raquo;</TD></TR><TR><TD CLASS="quote">
I'm considering full standalone AEM EMS or Autronic at this point, though the cost will be more than I can (or am willing to) bear :-(</TD></TR></TABLE>

Not out yet for our car. Only other solution may be motec, but that retails for $bling$

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote &raquo;</TD></TR><TR><TD CLASS="quote">
BTW, I am getting a shop to tune my car and my mechanic has had some luck getting tech support from Kenji @ GReddy as well, but nothing earth shattering. But it was "Matt" @ GReddy who suggested doing away with the Injector Correction feature. He didn't say why, he just said not to use it when we asked why the Injector Correction wasn't working well and how to debug. It may have been because of a corrupt .gsc file though -- we will try it again.
</TD></TR></TABLE>

Really surprised they haven't been more help....considering the situation (an authorized tuner actually calling)

I don't think its futile, otherwise SF's fuel kit wouldn't work either, or the MF2 - the car would try to lean out the fuel mixture if it saw more fuel in the O2 sensor with those as well.....

Joseph Davis

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote &raquo;</TD></TR><TR><TD CLASS="quote">Quote, originally posted by johnyblaze ยป
It starting to sound like closed-loop tuning with the e-Manage is futile and keeping off boost while in closed-loop is the only safe alternative. Pretty disappointing. </TD></TR></TABLE>
<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by opto_isolator &raquo;</TD></TR><TR><TD CLASS="quote">
Keep bugging greddy about it - if they can get the RSX emanage to work, and if the S2000 guys can get it work with superchargers, than surely it will work for a 2001 civic....</TD></TR></TABLE>

But, you see, Greddy didn't get anything working in re closed loop boosting with the e-manage on any platform. You are aware that the e-manage is widely considered a shoddy hack?

As far as full standalones and $$$... Megasquirt AVR. You can probably find someone local who is comfortable with kit electronics if you are uincomfortable assembling your own, and wiring will be required.

johnyblaze

I'm fairly convinced that the e-Manage will always be sub-par at best. The perfect/optimal/safest tune will only be possible with standalone. I suppose that's the nature of any piggy-back system.

I'm still unsure if I'll settle with the "okay" tune that the e-Manage can give me (and try to stay off boost while in closed loop), or if I want to pay the price for standalone.

As far as the AEM EMS goes, correct, it is not yet available for the 7thgen civic as plug-and-play, but I'm having conversations with http://www.torquefreaks.com (who has one of the nations best AEM EMS tuners) about doing a custom job, which is definitely feasible.

I'm also considering Autronic as there is a local dealer/installer/tuner here in WA.

I completely forgot about Motec -- I'm sure it's comparable to Autronic both in terms of price and features. I'm not sure if these systems can control accessories like A/C, power steering pump, cruise servo, etc. I've been told the AEM EMS can control these.

Anyways, I'll see how things continue with the e-Manage. Going to try employing the injector size correction feature again.

Cheers.

hoggy

Have any of you taken a look at uberdata. It works so much like a hondata system there is very little reason to not run it since its free if you are budget minded. Its the best management for a honda if you are a broke like me or just not wealthy. It may seem like more work to some, but its not nearly as hard to get a good tune compared to what it seems like you guys are going through.

Check out
http://www.ecimulti.org/uberdata/forum