Notices
Tuning Reflashes, Piggybacks, Standalone ECUs

Haltech Burbles and Pops on Decel Transition

Thread Tools
 
Search this Thread
 
Old Sep 20, 2010 | 08:21 AM
  #41  
aln's Avatar
aln
Registered User
 
Joined: Jun 2008
Posts: 177
Likes: 0
From: Dublins
Default

Originally Posted by mx594
To each his own I guess. To me it just sounds like a crappy running, poorly tuned car.
Reply
Old Sep 20, 2010 | 08:48 AM
  #42  
midz350's Avatar
midz350
New Member
iTrader: (4)
 
Joined: Aug 2007
Posts: 4,054
Likes: 24
From: around.
Default

Originally Posted by mx594
No.
This what was binder talking about:



As you can see, its in seconds.
Reply
Old Sep 20, 2010 | 08:52 AM
  #43  
aln's Avatar
aln
Registered User
 
Joined: Jun 2008
Posts: 177
Likes: 0
From: Dublins
Default

Originally Posted by mx594
I am still getting some loud pops on decel. I don't dare rev match on downshifts anymore, because it pops loudly every time. It's somewhat embarrassing. I miss the stock ECU
loud? like real loud or..>? Can you record the pop?
Reply
Old Sep 21, 2010 | 07:19 PM
  #44  
mx594's Avatar
mx594
Thread Starter
New Member
iTrader: (4)
 
Joined: Dec 2005
Posts: 608
Likes: 1
From: Grand Rapids, MI
Default

Alright I did some testing today. Here is what I have concluded:

Decel cut is based on APP, not TPS. I know this because I have logged many cases where the Decel cut activates, despite the TPS still reading above 2% (2% is my what my threshold is set to). I know it says "Throttle Value" in the ecu manager but I promise you, it is using APP. If you don't believe me, do some datalogs.

The second thing I have concluded is that my problem is definitely being caused by the "Injection Time Threshold" setting. If I set the threshold to 1.5 ms, the decel cut will activate right about 0.5 seconds after the APP drops below 2%. This is exactly what should happen, because the time "Delay to Fuel Cut" is set to 0.5 seconds.

However, if I set the "Injection Time Threshold" to 0.85 ms (the highest it can go while still allowing the cruise to work) then the delay between 2% APP and decel cut is more like 1.0 seconds. So it seems as though the "Injection Time Threshold" is what is causing the excessive delay between throttle lift and decel cut, because it should take no longer than the "Delay to Fuel Cut" setting, which is 0.5 seconds.

See the attached datalog for a comparison between 1.5 and 0.85 ms "Injection Time Threshold". Notice the time difference. This is repeatable every time.

What it comes down to is that the Haltech needs to be able to recognize when the cruise is engaged and prevent fuel cut from happening. Because an "Injection Time Threshold" of 0.85 ms is not high enough to have a proper decel cut. It sees the 0% APP when you take your foot off the throttle after engaging the cruise, and thinks it is time to go into decel cut - unless you set the "Injection Time Threshold" excessively low (lower than the stock ECU would ever call for when cruising).

At least that's what I believe to be true. What do you guys think?
Attached Thumbnails Haltech Burbles and Pops on Decel Transition-decel-cut-1.5-vs-.85-threshold.jpg  
Reply
Old Sep 21, 2010 | 07:28 PM
  #45  
mx594's Avatar
mx594
Thread Starter
New Member
iTrader: (4)
 
Joined: Dec 2005
Posts: 608
Likes: 1
From: Grand Rapids, MI
Default

Originally Posted by midz350
This what was binder talking about:



As you can see, its in seconds.
No that's not what binder was talking about. Changing the "Delay to Fuel Cut" time wouldn't affect the cruise, only the "Injection Time Threshold" will cause that problem. Since he mentioned that it affected his cruise, and since he typed "ms" after every number in his post, I think it's safe to say he was talking about "Injection Time Threshold".

That said, he wasn't making any sense.
Reply
Old Sep 22, 2010 | 03:04 PM
  #46  
binder's Avatar
binder
New Member
iTrader: (8)
 
Joined: Jan 2007
Posts: 6,457
Likes: 7
From: terre haute, IN; STL, MO
Default

injection time threshold is what i was talking about in ms.
Reply
Old Sep 22, 2010 | 03:13 PM
  #47  
binder's Avatar
binder
New Member
iTrader: (8)
 
Joined: Jan 2007
Posts: 6,457
Likes: 7
From: terre haute, IN; STL, MO
Default

Originally Posted by mx594
2.5 ms IS really high. That's my point. If it is set too high (above ~.85 ms) the cruise won't work. I have never heard of the cruise having problems if this is too low...I'm not sure what you are talking about. The Haltech basemap has this parameter set to 2.5 ms. So with the base map provided by Haltech, the cruise won't work.

At 2.5 ms there are no pops, no burbles, just clean quiet decel. Just like stock.

I'm pretty sure the Haltech uses the APP for fuel cut. Otherwise we wouldn't have the problem of going into fuel cut when using the cruise control. When you engage the cruise and take your foot of the gas, APP goes to 0% but I really doubt the TPS goes below 2%.

And I STILL don't see the injection time going to 0 during decel cut. It only goes down to about ~1.2 ms. Are you guys sure it is supposed to go completely to zero?
i've never had a problem with cruise not working. My problem was when going downhill with cruise on so the TPS WAS actually going below 2% so it has to be tied to the TPS not the APS because it wouldn't work at all when cruise was on since APS is 0% then. In cruise my TPS will go below 2% on downhill which is when my decel has issues.

When i talked to hal on the phone about this as i was driving he told me most engines have an injection time threshold of .550ms and it works fine but mine needed to be bumped up. I bumped it up to .650ms on my vortech and my cruise worked fine even on decel. On my turbo system i had to increase it to .750ms to keep the car from bucking when on cruise and going downhill.
Reply
Old Sep 22, 2010 | 08:33 PM
  #48  
mx594's Avatar
mx594
Thread Starter
New Member
iTrader: (4)
 
Joined: Dec 2005
Posts: 608
Likes: 1
From: Grand Rapids, MI
Default

binder,

do you even know what "Injection Time Threshold" means? Seriously. Because what you are saying does not make sense.
Reply
Old Sep 22, 2010 | 11:04 PM
  #49  
djamps's Avatar
djamps
Registered User
iTrader: (1)
 
Joined: Dec 2008
Posts: 4,492
Likes: 10
From: MD
Default

Based on the values I'd venture to guess 'minimum effective injector latency'? On my Uprev I've got it set to 0.4ms which is what my DW injectors specify.
Reply
Old Sep 23, 2010 | 05:39 AM
  #50  
mx594's Avatar
mx594
Thread Starter
New Member
iTrader: (4)
 
Joined: Dec 2005
Posts: 608
Likes: 1
From: Grand Rapids, MI
Default

Originally Posted by djamps
Based on the values I'd venture to guess 'minimum effective injector latency'? On my Uprev I've got it set to 0.4ms which is what my DW injectors specify.
Huh? We are not talking about injector latency. Are you talking about the minimum effective pulsewidth? Because you can change that with Osiris.

"Injection Time Threshold" in the Haltech refers to the injection time in ms that the STOCK ECU must fall below in order to signal the Haltech that it is ok to begin decel cut. Don't forget that even with the Haltech, the stock computer still thinks it is in control and still tries to fuel the engine accordingly. The Haltech monitors the injection pulses that the stock computer is calling for. Heck, you can even copy them through if you want, and effectively run the stock computer through the Haltech.
Reply
Old Sep 23, 2010 | 09:08 AM
  #51  
str8dum1's Avatar
str8dum1
New Member
iTrader: (11)
 
Joined: Oct 2007
Posts: 8,807
Likes: 7
From: raleigh-wood NC
Default

how/why does that affect the cruise control if the value is too high?

I assume you dont want the Haltech to trigger the decel cut while using the CC?
Reply
Old Sep 23, 2010 | 09:56 AM
  #52  
mx594's Avatar
mx594
Thread Starter
New Member
iTrader: (4)
 
Joined: Dec 2005
Posts: 608
Likes: 1
From: Grand Rapids, MI
Default

See the screenshot posted by midz350 above.

There are three conditions which must be met for the Haltech to activate decel cut.

1.Throttle Value - I believe this to be based on accelerator pedal position (APP), not throttle position (TPS). I have seen in my datalogs many instances where decel cut will activate when TPS is still above the Throttle Value setting.

2. Injection Time Threshold - This is the time in ms which the STOCK ECU injector pulsewidth must fall below to trigger decel fuel cut in the Haltech. Unfortunately, the Haltech does not allow logging of the stock pulsewidth. (In fact that just gave me and idea - I could log the stock pulsewidth with Cipher and see if it truly does fall below the threshold when I hit the cruise...)

3. Delay Till Fuel Cut - After both the above conditions are met the Haltech will wait this amount of time before activating decel cut. I am not sure if this time begins when condition 1 is met or when both 1 and 2 are met though.

When you activate the cruise control and take your foot off the pedal, the APP or "Throttle Value" goes to 0. So condition 1 is met. Now the only thing keeping the Haltech from thinking it's time for decel cut is the Injection Time Threshold. Thus, if this value is set too high the stock ECU may command an injector pulsewidth less than this value (when it is regulating the speed of the vehicle) and send the Haltech into decel cut. Once the cut happens, the stock computer sees the deceleration and applies more throttle (more throttle = more fuel) and the stock injection time rises above the Injection Time Threshold again and the decel cut stops momentarily before it happens all over again. That's why when you set the Injection Time Threshold too high, the cruise will cycle on-off-on-off about every .5 seconds.

At least that is what I believe to be true.

Last edited by mx594; Sep 23, 2010 at 10:00 AM.
Reply
Old Sep 24, 2010 | 06:55 AM
  #53  
binder's Avatar
binder
New Member
iTrader: (8)
 
Joined: Jan 2007
Posts: 6,457
Likes: 7
From: terre haute, IN; STL, MO
Default

I know what it is.

I'm telling you EXACTLY what hal told me to do to fix my cruise problems when going down a hill and car goes into decel with cruise enabled.

I bumped it from .55 to .65 the first time and the bucking going downhills on cruise stopped. Then on these injectors i continued to raise it until it stopped. .750ms was the time in which my cruise was smooth going down hills.

If your cruise isn't working in straight and level then something else is going funny. I know hal sets most cars to .550ms injection threshold because he said that is a good starting point and usually it has to be increased slightly for larger injectors. He told me to do steps of .010ms until the problem stopped. That's why i'm saying 2.5ms is way too high. Your injector pulsewidth will probably be lower than that all the time therefore never going into decel cut i would assume. I'm not an expert but i've asked these questions to Hal who is the expert.

Also, on another note i increased my fuel enrich from 5% to 10% and that greatly reduced my lean buck when i touch the throttle after decel. I'm going to go alittle higher and it should fix it completely.
Reply
Old Sep 24, 2010 | 06:48 PM
  #54  
mx594's Avatar
mx594
Thread Starter
New Member
iTrader: (4)
 
Joined: Dec 2005
Posts: 608
Likes: 1
From: Grand Rapids, MI
Default

Originally Posted by binder
That's why i'm saying 2.5ms is way too high. Your injector pulsewidth will probably be lower than that all the time therefore never going into decel cut i would assume.

Dude, you are thinking of it backwards. Decel cut happens when the stock ecu pulsewidth falls BELOW the threshold. So if your Threshold is too low decel cut will be take longer to activate, as I proved in the datalog attached previously.

I have never heard of anyone raising the threshold to fix a cruise problem. If you had it too low it might cause problems with the cruise as well, but I don't know that for sure becuase I have never tried it.

I am using 2.0 ms right now, and on the highway I can get the cruise to work if I am not going down hill. If I activate the cruise and very carefully lift my foot off the throttle it will stay engaged, but if I try to lower my speed with the coast button or start going downhill, the stock ecu will output a pulsewidth of less than 2.0 ms and the car will go into it's on-off-on bucking cycle.

The best part about using 2.0 ms is that the car NEVER pops anymore, not even on rev matching downshifts. It sounds like stock and the gear changes are still smooth as can be. To me that is more important than a working cruise control, so I am going to leave it as is for now. I am gonig to contact Haltech and see what they have to say. If there was a way to make the threshold adjustable depending on whether or not the cruise "set" light was on, it would work much better I think.
Reply
Old Sep 25, 2010 | 03:07 PM
  #55  
binder's Avatar
binder
New Member
iTrader: (8)
 
Joined: Jan 2007
Posts: 6,457
Likes: 7
From: terre haute, IN; STL, MO
Default

Originally Posted by mx594

I have never heard of anyone raising the threshold to fix a cruise problem. If you had it too low it might cause problems with the cruise as well, but I don't know that for sure becuase I have never tried it.
email hal, he'll tell you exactly what he told me which is the same thing i posted. he starts tuning at .550 ms and slow raises from there. If a pro tuner starts at .550ms do you think that your 2.5ms is a little excessive?

Originally Posted by mx594

I am using 2.0 ms right now, and on the highway I can get the cruise to work if I am not going down hill. If I activate the cruise and very carefully lift my foot off the throttle it will stay engaged, but if I try to lower my speed with the coast button or start going downhill, the stock ecu will output a pulsewidth of less than 2.0 ms and the car will go into it's on-off-on bucking cycle.

The best part about using 2.0 ms is that the car NEVER pops anymore, not even on rev matching downshifts. It sounds like stock and the gear changes are still smooth as can be. To me that is more important than a working cruise control, so I am going to leave it as is for now. I am gonig to contact Haltech and see what they have to say. If there was a way to make the threshold adjustable depending on whether or not the cruise "set" light was on, it would work much better I think.
email hal, he's going to tell you to lower it.

You have something else going on in your fuel table
Attached Thumbnails Haltech Burbles and Pops on Decel Transition-decel_values.jpg  
Reply
Old Sep 26, 2010 | 06:33 PM
  #56  
mx594's Avatar
mx594
Thread Starter
New Member
iTrader: (4)
 
Joined: Dec 2005
Posts: 608
Likes: 1
From: Grand Rapids, MI
Default

I don't know why I waste my time doing all of this testing, spending hours analyzing datalogs and posting the results on here. I have proven that the decel cut does not work properly with threshold settings that low, and your only response is "well Hal said..."

Maybe this is just a shortcoming of the Haltech?

I paid Hal $100 for my basemap by the way. The threshold was set to .65 ms.

I did email Hal about this months ago telling him about my success with raising the threshold, and the only thing he had to say was "Have you tried using cruise control since those changes? You may encounter trouble there on downhills". He then went on to say "If it is popping it sounds like you have other settings that need to be setup, such as APP cal, start point, decel rpm, etc".

I assure you that my APP and TPS calibrations are 100% correct. The decel rpm table is setup correctly - in fact I haven't changed it from what Hal had set up. I don't know what start point is.
Reply
Old Sep 27, 2010 | 04:57 AM
  #57  
djamps's Avatar
djamps
Registered User
iTrader: (1)
 
Joined: Dec 2008
Posts: 4,492
Likes: 10
From: MD
Default

Do you have a flash on your stock ECU? I wonder if the injection timing is modified somehow.
Reply
Old Sep 27, 2010 | 05:01 AM
  #58  
Alberto's Avatar
Alberto
Cranky FI Owner
Premier Member
iTrader: (14)
 
Joined: Mar 2003
Posts: 34,715
Likes: 8
From: DMV
Default

My Haltech tune from Forged did this like crazy. I actually liked it.

I then had the car retuned and it all went away. I would compare maps and try to figure out the differences but my laptop was stolen from the car with the maps on it
Reply
Old Sep 27, 2010 | 08:51 AM
  #59  
mx594's Avatar
mx594
Thread Starter
New Member
iTrader: (4)
 
Joined: Dec 2005
Posts: 608
Likes: 1
From: Grand Rapids, MI
Default

Originally Posted by djamps
Do you have a flash on your stock ECU? I wonder if the injection timing is modified somehow.
I was using Osiris tuner for a short period of time, but I flashed it back to stock before before removing the UTEC and installing the Haltech.
Reply
Old Sep 27, 2010 | 08:53 AM
  #60  
mx594's Avatar
mx594
Thread Starter
New Member
iTrader: (4)
 
Joined: Dec 2005
Posts: 608
Likes: 1
From: Grand Rapids, MI
Default

Originally Posted by Alberto
My Haltech tune from Forged did this like crazy. I actually liked it.

I then had the car retuned and it all went away. I would compare maps and try to figure out the differences but my laptop was stolen from the car with the maps on it
Aww man that sucks! I would be pissed.

Have you gotten a new laptop and downloaded your current map since then? If so, would you mind me taking a look at it to compare it to mine?
Reply



All times are GMT -8. The time now is 06:23 PM.