Pelican Parts
Parts Catalog Accessories Catalog How To Articles Tech Forums
Call Pelican Parts at 888-280-7799
Shopping Cart Cart | Project List | Order Status | Help



Go Back   Pelican Parts Forums > Porsche Forums > 911 / 930 Turbo & Super Charging Forum


Reply
 
LinkBack Thread Tools Rating: Thread Rating: 3 votes, 5.00 average.
Author
Thread Post New Thread    Reply
Registered User
 
Join Date: May 2015
Posts: 762
Quote:
Originally Posted by Reanimotion View Post
if the max eWUR duty cycle is still set to 80 it might be worth increasing it a bit to 90 or 95 in order to mitigate that first few seconds of high CP on the second log
I'll set it higher, although it really didn't have any noticeable affect on the starting or the post-start idle.

Any clues on why cold starting has become an issue? I checked the cold start valve circuit, its all plugged in.
Old 01-25-2017, 04:58 AM
  Pelican Parts Catalog | Tech Articles | Promos & Specials    Reply With Quote #741 (permalink)
umop apisdn
 
Join Date: Jun 2014
Location: Brisbane, Australia
Posts: 485
logging during several cold starts should provide some clues as to the variables that might be in play

it's all going to be about balancing the transition from off to cranking to running

accumulator pressure - fuel pump turn on - cold start injector status - FV and CP ratios all need to be tweaked to ensure the system has and maintains fuel pressure and subsequent delivery during the startup

for example, until the engine turns faster than the "crank speed" setting, the system is essentially in fuel cut mode and the majority of fuel delivered will be from the cold start injector only. And the thermo time switch on the cold start circuit limits delivery to 8 seconds or so of total cranking.

warm start seems ok from the log but cold might be turning a little slower and not kicking over the set point
lowering the cranking speed setting from 300 to maybe 150 or less will bring the CP and FV control in earlier
__________________
Steve
1981 928S 4.7 ROW with KE3-Jetronic and Franken8 (AEM Inifinty) follow at [http://www.frankencis.com/Activity-Feed/userid/2]
Yes! mechanical/hydraulic constant flow injection can be managed by a modern EMS

www.FrankenCIS.com
Old 01-26-2017, 05:21 AM
  Pelican Parts Catalog | Tech Articles | Promos & Specials    Reply With Quote #742 (permalink)
Registered User
 
Join Date: May 2015
Posts: 762
Quote:
Originally Posted by Reanimotion View Post
logging during several cold starts should provide some clues as to the variables that might be in play

it's all going to be about balancing the transition from off to cranking to running

accumulator pressure - fuel pump turn on - cold start injector status - FV and CP ratios all need to be tweaked to ensure the system has and maintains fuel pressure and subsequent delivery during the startup

for example, until the engine turns faster than the "crank speed" setting, the system is essentially in fuel cut mode and the majority of fuel delivered will be from the cold start injector only. And the thermo time switch on the cold start circuit limits delivery to 8 seconds or so of total cranking.

warm start seems ok from the log but cold might be turning a little slower and not kicking over the set point
lowering the cranking speed setting from 300 to maybe 150 or less will bring the CP and FV control in earlier
I'll give that a shot. Trouble with cold start testing is that I only get one shot at it.
Old 01-26-2017, 09:13 AM
  Pelican Parts Catalog | Tech Articles | Promos & Specials    Reply With Quote #743 (permalink)
Registered User
 
Join Date: May 2015
Posts: 762
That helped with the cold start. Been doing some driving with autotune. Started out with gentle drives around the ramp, the up and down the access road, then a bit of highway cruising, then getting on it a little bit. Still got some way to go. Damn the combination of RarlyL8 headers, Hooligan, and the K&N filter is LOUD! The fweeeeee as the turbo spools up is quite audible, and the whooooosh sucking sound when you get on the turbo is so loud is scared the crap out of me the first time. Then I had to do it again just to hear it. No way I want to go on a trip with this exhaust on it.

Anyway, I set my idle mixture to 14.2 but the autotune seems to have made it much higher, possibly effecting hot starts. Here's my AFR and VE tables for #1.

Question Steve, should I somehow be doing and auto-tune with VE table #2 as well?


Old 01-28-2017, 02:47 PM
  Pelican Parts Catalog | Tech Articles | Promos & Specials    Reply With Quote #744 (permalink)
umop apisdn
 
Join Date: Jun 2014
Location: Brisbane, Australia
Posts: 485
as VE Table 2 has control of the frequency valve it should be the primary tune target for mixture
VE 1 can then be relegated to coarse tuning if required or left at 100 where not needed
__________________
Steve
1981 928S 4.7 ROW with KE3-Jetronic and Franken8 (AEM Inifinty) follow at [http://www.frankencis.com/Activity-Feed/userid/2]
Yes! mechanical/hydraulic constant flow injection can be managed by a modern EMS

www.FrankenCIS.com
Old 01-28-2017, 03:12 PM
  Pelican Parts Catalog | Tech Articles | Promos & Specials    Reply With Quote #745 (permalink)
Registered User
 
Join Date: May 2015
Posts: 762
Quote:
Originally Posted by Reanimotion View Post
as VE Table 2 has control of the frequency valve it should be the primary tune target for mixture
VE 1 can then be relegated to coarse tuning if required or left at 100 where not needed
So do you recommend I put VE table 1 back to 100's and restart auto-tune with table 2?
Old 01-28-2017, 07:10 PM
  Pelican Parts Catalog | Tech Articles | Promos & Specials    Reply With Quote #746 (permalink)
umop apisdn
 
Join Date: Jun 2014
Location: Brisbane, Australia
Posts: 485
Yes, I would initially leave Table 1 at flat 100 and do my primary tuning with Table 2 and the frequency valve

Think of the eWUR and Table 1 as manipulating air plate position and therefore fuelling range for a given airflow
and
the Frequency Valve and Table 2 as directly controlling the fuel delivery at the injectors with a +/- on the base line delivery provided by the current position of the air plate

one will have an effect on the other of course, and the both may need to be adjusted for a particular load and/or RPM point to get the desired result

theoretically, either of the Control Pressure or the Frequency Valve will adjust fuelling and therefore provide a reasonable end result. But, the Frequency Valve is directly manipulating the differential pressure at the metering valves within the head, and the eWUR Control Pressure has a mechanical lag as it requires the air plate to move in response to the change in control pressure at the top of the piston.
__________________
Steve
1981 928S 4.7 ROW with KE3-Jetronic and Franken8 (AEM Inifinty) follow at [http://www.frankencis.com/Activity-Feed/userid/2]
Yes! mechanical/hydraulic constant flow injection can be managed by a modern EMS

www.FrankenCIS.com
Old 01-28-2017, 09:36 PM
  Pelican Parts Catalog | Tech Articles | Promos & Specials    Reply With Quote #747 (permalink)
Registered User
 
Join Date: May 2015
Posts: 762
Quick update on where I am with this.

Reset the #1 VE table to 100s, started auto tune using VE table #2. Started with some easy parking lot driving, then round the block a bit, a little highway, some cruising. Have gotten on the turbo hard a couple of times, but haven't really auto tuned long hard pulls through the gears yet. When I was doing the beginning auto tune I did notice a little change in the engine feel. Didn't expect much as we're starting from a pretty good base tune, this is about ensuring fueling with extra boost.

Lessons learned so far:
-Power should come off the delayed action relays as there is an interruption of power when the ignition moves from Crank to Run which was causing problems
-Due to the much higher fuel pressures in the 930 a larger fuel injector is needed in the eWUR. I used a 48lb injector from Five O motorsports. I have still seen some bouncing fuel pressures still, and would probably opt for the next larger injector size
-Grounds are critical, make sure sensor ground used for every sensor, and ground the TPS input also
-Ignition pickup off the TD output on a 6 pin CDI works fine, use VR1+ and leave VR1- floating

In addition I used a sandbag and a small body panel hammer to add an additional 1/4" relief in the heat shield behind the bumper to give a little more gap between it and the top of the AFR sensor.

I'll follow up with some logs shortly.
Old 02-06-2017, 02:03 PM
  Pelican Parts Catalog | Tech Articles | Promos & Specials    Reply With Quote #748 (permalink)
umop apisdn
 
Join Date: Jun 2014
Location: Brisbane, Australia
Posts: 485
Excellent, and thanks for the ongoing feedback
__________________
Steve
1981 928S 4.7 ROW with KE3-Jetronic and Franken8 (AEM Inifinty) follow at [http://www.frankencis.com/Activity-Feed/userid/2]
Yes! mechanical/hydraulic constant flow injection can be managed by a modern EMS

www.FrankenCIS.com
Old 02-06-2017, 02:18 PM
  Pelican Parts Catalog | Tech Articles | Promos & Specials    Reply With Quote #749 (permalink)
Registered User
 
Join Date: May 2015
Posts: 762
Quote:
Originally Posted by Reanimotion View Post
Excellent, and thanks for the ongoing feedback
No problem, here's some more:

-Lambda valve regularly blows a 2A fuse. Replaced with a 5A, seems fine so far


Also, still definitely seeing the wild swings in eWUR duty cycle chasing wild swings in Control Pressure. Unsure why this is happening. Does not seem to happen all the time, gut says it happens more often on hot starts, but once it starts I can't find a way to stop it. Here's some more logs.

1
2
3
4
5
6
7
Old 02-07-2017, 01:59 PM
  Pelican Parts Catalog | Tech Articles | Promos & Specials    Reply With Quote #750 (permalink)
Registered User
 
Join Date: May 2015
Posts: 762
Also, I'm a a bit confused on what auto tune is doing.

EGO control is set to above 1,500 rpm and above 60 MAP
My idle is 950 and around 52 MAP
Idle mixture was manually set to 14.2 which matches the target AFR in the AFR tables

Q1: So why any changes in idle bins at all?
Q2: And why such large changes in the idle bins if already set to that mixture?
Q3: The numbers are above 100 which should be richer, so why is my idle AFR now leaner?


Old 02-07-2017, 02:07 PM
  Pelican Parts Catalog | Tech Articles | Promos & Specials    Reply With Quote #751 (permalink)
umop apisdn
 
Join Date: Jun 2014
Location: Brisbane, Australia
Posts: 485
Auto tune has it's own settings on the advanced tab to the right, separate to those of the ECU EGO control parameters
__________________
Steve
1981 928S 4.7 ROW with KE3-Jetronic and Franken8 (AEM Inifinty) follow at [http://www.frankencis.com/Activity-Feed/userid/2]
Yes! mechanical/hydraulic constant flow injection can be managed by a modern EMS

www.FrankenCIS.com
Old 02-07-2017, 02:10 PM
  Pelican Parts Catalog | Tech Articles | Promos & Specials    Reply With Quote #752 (permalink)
 
Registered User
 
Join Date: May 2015
Posts: 762
Quote:
Originally Posted by Reanimotion View Post
Auto tune has it's own settings on the advanced tab to the right, separate to those of the ECU EGO control parameters
ok thanks. Please advise if you have any insight into the cp and dc swinging/needle chasing.
Old 02-07-2017, 02:14 PM
  Pelican Parts Catalog | Tech Articles | Promos & Specials    Reply With Quote #753 (permalink)
umop apisdn
 
Join Date: Jun 2014
Location: Brisbane, Australia
Posts: 485
I've got part way through reviewing the logs and #5 @16:12:46 seems to be the best hint at what might be going on

below are my settings for the log viewer which helps keep things visually clear between different logs


the first oscillation during deceleration looks very much like the mythical air plate flutter
on a one second cycle with around a half second lag on the wideband sensor
it goes rich
the control pressure drops
the duty then drops to compensate
it goes lean
control pressure increases
duty increases to compensate

it definitely still tracks the target pressure and seems to settle when the engine loads up, but the data is not conclusive without maybe putting the mechanical WUR back in place and recoding control pressure without the eWUR as a variable

adding a configuration for the control pressure PID reaction to allow it to be tuned might be valuable, but from the looks of it, it seems to be an external influence rather than some form of oscillation as the CP is changing direction before the duty turns around
__________________
Steve
1981 928S 4.7 ROW with KE3-Jetronic and Franken8 (AEM Inifinty) follow at [http://www.frankencis.com/Activity-Feed/userid/2]
Yes! mechanical/hydraulic constant flow injection can be managed by a modern EMS

www.FrankenCIS.com
Old 02-07-2017, 07:32 PM
  Pelican Parts Catalog | Tech Articles | Promos & Specials    Reply With Quote #754 (permalink)
Registered User
 
Join Date: May 2015
Posts: 762
Quote:
Originally Posted by Reanimotion View Post
the first oscillation during deceleration looks very much like the mythical air plate flutter
on a one second cycle with around a half second lag on the wideband sensor
Possibly, although these are just samples from the drive home. My observations are that it either is there from start up, or it isn't. If it is, there's no way to get rid of it.

Quote:
Originally Posted by Reanimotion View Post
it goes rich
the control pressure drops
the duty then drops to compensate
it goes lean
control pressure increases
duty increases to compensate
Right, classic chasing the needle. Given the built-in and unavoidable delay in responding to mixture changes I think some sort of hysteresis or some attempt at prediction is needed to get ahead of the pulsing.

Quote:
Originally Posted by Reanimotion View Post
it definitely still tracks the target pressure and seems to settle when the engine loads up,
The oscillations dampen out, but don't go away, then when we go back to idle or overrun, they come back.

Quote:
Originally Posted by Reanimotion View Post
but the data is not conclusive without maybe putting the mechanical WUR back in place and recoding control pressure without the eWUR as a variable
I'd be happy to try that, however I'm not sure how to do that. It will require some additional fittings I think, to tap into the fuel line between the WUR and the fuel head.

Quote:
Originally Posted by Reanimotion View Post
adding a configuration for the control pressure PID reaction to allow it to be tuned might be valuable, but from the looks of it, it seems to be an external influence rather than some form of oscillation as the CP is changing direction before the duty turns around
Possibly. However I've never seen this behavior with the mechanical WUR in place. Never observed any pulsing or oscillation in RPM. Since the mechanical WUR isn't reactive to mixture changes, I don't see how it would get into a needle chase.

Have you seen this behavior on non-turbo applications?
Old 02-08-2017, 06:58 AM
  Pelican Parts Catalog | Tech Articles | Promos & Specials    Reply With Quote #755 (permalink)
Registered User
 
Join Date: Sep 2004
Location: Central Va
Posts: 3,624
flightlead404 and Reanimotion,

I simply want to say I REALLY appreciate you two troubleshooting this system on the Pelican forum. This stuff is fascinating.

THANKS!
__________________
-Scott
'87 911 turbo, '10 Cayman S
In the past - '71 T, '77 S 3.6, '80 SC, '88 3.6 cab
Old 02-08-2017, 10:32 AM
  Pelican Parts Catalog | Tech Articles | Promos & Specials    Reply With Quote #756 (permalink)
Registered User
 
fasteddie313's Avatar
 
Join Date: Mar 2014
Posts: 37
Garage
Can you just take the WUR off of autotune, figure out what duty cycles give you what CP with no plate fluctuations, and then set the duty cycle to that, with a warmup curve and a boost curve?

If you plate is bouncing and it is trying to chase it, wouldn't setting the duty cycle solid let it bounce and fluctuate but still keep an aveage CP of your set duty cycle? More like a regular WUR..

I think I would try tuning the CP manually to get you close to what ARF you want but a bit rich/safe all the way through boost, set that and forget it, and then use the FV tuning to lean it out as much as you dare to make the best power and achieve the best fuel efficency..

I'm not real familiar with what you can and cannot tell the MS to do for you yet so I don't really know..


I'm getting pretty close with my car..
I have the Fuel Distributor mounting, intake side plumbing, and fuel feed/return lines done.. I'm confident in the siamese injector lines but still waiting on pieces for that.. I also have my head back on, working in the intake vacuum routing with a bosch idle valve, have the MS harness through the firewall nice and WB half installed..
I'm a bit hung up at the moment, need to buy bolts, wiring supplies, couple taps, nickle dime stuff I can grab soon.

https://www.youtube.com/watch?v=il4Dwv4dMQM
https://www.youtube.com/watch?v=IpUGzFvWECY
Some others on that channel..

About all this wiring..
I'm leaning towards doing my pigtails with solder and shrink wrap, and some I have those sensor ends for with the crimp pins.. But like my wideband needs to go through my dash so I think I should put shielded spade crimp connectors on that inside the car so I can still take my gauge/switch plate out for other things..
I can make circuits alright but I'm not experienced in doing more professional style automotive connections, what do you suggest? Any resistance problems with solder? How should I do these?

Hope I'm not disrupting your tuning conversation too much..
__________________
80 931S LSD-G31 Big Breaks FMIC
1 Bar so far
Old 02-08-2017, 11:01 AM
  Pelican Parts Catalog | Tech Articles | Promos & Specials    Reply With Quote #757 (permalink)
umop apisdn
 
Join Date: Jun 2014
Location: Brisbane, Australia
Posts: 485
Quote:
Originally Posted by flightlead404 View Post
However I've never seen this behavior with the mechanical WUR in place. Never observed any pulsing or oscillation in RPM. Since the mechanical WUR isn't reactive to mixture changes, I don't see how it would get into a needle chase.

Have you seen this behavior on non-turbo applications?
looking at the log we've been discussing, the pressure fluctuations may be a red herring. The chasing should have a damping effect if the pressure fluctuations are indeed just caused by the air plate fluctuating.

base question - is there a detrimental effect on mixture or engine behavior, and if so where?

the first on overrun seems benign, the second during idle has VE2 and the frequency valve impacting the AFR and idle

IF the idle area of VE2 can be flattened out, does idle and AFR improve?

__________________
Steve
1981 928S 4.7 ROW with KE3-Jetronic and Franken8 (AEM Inifinty) follow at [http://www.frankencis.com/Activity-Feed/userid/2]
Yes! mechanical/hydraulic constant flow injection can be managed by a modern EMS

www.FrankenCIS.com
Old 02-08-2017, 02:58 PM
  Pelican Parts Catalog | Tech Articles | Promos & Specials    Reply With Quote #758 (permalink)
Registered User
 
Alan L's Avatar
 
Join Date: Apr 2006
Location: New Zealand
Posts: 3,233
I have had bad WUR CP pulsing just like you experience here. I had it before my UTCIS install, then it went away. Then it came back with the injector version of the UTCIS. I could feel the pulsing in the WUR lines. The airplate had a fast vibration to it. (In response). If I put my finger on the airplate it stopped and everything was stable. Until I blipped the throttle and the system tried to adjust again. This was an issue with the UTCIS and injector parameters as far as we could determine. UT concluded my fuel head had a lot less leakage than their test unit - which did not exhibit this behaviour. We tried a few tweeks to fix it, but it never went away. The unit went back to UT, never to be seen again. But it would seem like a mis calibration of injector function in the CP line? I don't know what set the earlier pulsing off, when I was originally working on the engine. But the pulsing exhibited an off note in the engine (mixture) and the tach behaved oddly - I can't remember the exact symptom, but I could tell inside the car from the tach when this pulsing state had begun. It will be very bad for the WUR and fuel head diaphragms.
Regards
Alan
__________________
83 SC, 82 930 (track) - Stock except for RarlyL8 race headers, RarlyL8 Zork, K27-7006, 22/28 T bars, 007 Fuel head, short 3&4 gears, NGK AFR, Greddy EBC (on the slippery slope), Wevo engine mounts, ERP rear camber adjust and mono balls, Tarret front monoball camber adjust, Elgin cams, 38mm ported heads, 964 IC. 380rwhp @ 0.8bar Apart from above, bone stock:-)
Old 02-08-2017, 03:34 PM
  Pelican Parts Catalog | Tech Articles | Promos & Specials    Reply With Quote #759 (permalink)
umop apisdn
 
Join Date: Jun 2014
Location: Brisbane, Australia
Posts: 485
Thanks Alan,
This isn't the injector hammering we saw in the early testing before we added the fuel damper, this is a slow and steady once per second surge or cycle.

Our injector is running at 100 cycles a second and the pressure /duty feedback loop is on 1000 cycles per second so I'm comfortable it's not injector pulse related

from appearances it's a nice cyclic curve in control pressure with a 3 or 4 pound swing +/- our target. we might be able to tune the control/feedback loop to hit harder and arrest the wobble or attack it in another way by not trying to correct a deviation that lasts less than a second.
Or adding the factory damper in the control line could be an option if it's not there now.

Hang on, Flightlead404 - is the factory damper in the circuit on your car? I had one on mine that had to be removed as it was shuttling a pulse between the original system damper and the eWUR integral unit

Primary question out there now is - is AFR being affected?
IF AFR and idle/running is good then a wobble visible at the WUR may not be affecting the fuel delivery at all
If we lock the duty cycle to an assumed pressure as Eddie suggested and UTCIS did on their unit we lose the damping effect of the PID control and probably still have the wobble in the pressure.

this mess below is from post 100 and no damper at 1/10th of a second sampling
__________________
Steve
1981 928S 4.7 ROW with KE3-Jetronic and Franken8 (AEM Inifinty) follow at [http://www.frankencis.com/Activity-Feed/userid/2]
Yes! mechanical/hydraulic constant flow injection can be managed by a modern EMS

www.FrankenCIS.com
Old 02-08-2017, 09:22 PM
  Pelican Parts Catalog | Tech Articles | Promos & Specials    Reply With Quote #760 (permalink)
Reply

Thread Tools
Rate This Thread
Rate This Thread:

 


All times are GMT -8. The time now is 09:09 AM.


 
Powered by vBulletin® Version 3.8.7
Copyright ©2000 - 2018, vBulletin Solutions, Inc.
Search Engine Optimization by vBSEO 3.6.0
Copyright 2018 Pelican Parts, LLC - Posts may be archived for display on the Pelican Parts Website -    DMCA Registered Agent Contact Page
 

DTO Garage Plus vBulletin Plugins by Drive Thru Online, Inc.