Failsafe activating after shift.

[Iworktofish]

Greetings,

My setup:
*97 BMW M3
*Technique Tuning Stage 2 turbo
*8psi wastegate, 13psi normal, 15psi with 50/50 meth
*HFS-6 with single 1mm jet and 0.7mm HSV restrictor
*Eboost2 wired to only allow HFS-6 to see IDC >13psi using auxiliary output to relay
*Coolingmist 1.5gal U shaped tank in trunk

During a single gear pull, system is spraying above 13psi as intended, indicated by 5-7 bars (best I can tell, hard to see & drive simultaneously). However, after I shift to the next gear, failsafe is activated dropping me to 8psi wastegate, and either stays active, or sometimes comes back after short delay.

I have tried adjusting the SC to ensure I'm getting ~5-7bars of lit LED's.
I adjusted WL 2 clicks CCW, WH 2 clicks CW to open up the window.
I have turned down the failsafe RST to minimum recover time.
Meth tank is not the greatest design and prone to sloshing (per Howerton) but it is full, and sensor is near bottom.

I am still getting intermittent failsafe activations after shifting to the next gear. Any ideas what would cause this and how to fix?

Thanks!

[Richard L]

This is interesting, the failsafe window is wide enough to cope.

Try two things:
1. increase the "FLSF SENS" trimmer by advancing it a few degress "clockwise". I gathered the naming of this trimmer is quite mis-leading. Clockwire to slow down the repsonse time. (range 0.2-0.7 seconds)

2. Unplug the tank level sensor so that it doesn't cause visual confusion to the fialsafe trip.

The amount of liquid you are injection is relately small compared to what the system is capable of. By inceasing the sensitiveity of the system with SC trimmer, you will be loosing some sampling resolutions. It causes more "over-shoot" and "under-shoot" of the flow senssing circuitry. By incrasing the "FLSF SEN" (clockwise), you are averaging the incoming signal so minimising the false trigger of the failsafe threshold.

[Iworktofish]

Thanks Richard,
So to make sure I understand, return the SC trimmer to normal 12 o'clock position, and instead use the FLSF SENS to slow down response time?

While driving the car again last night, I did notice that when the relay opens at 13psi and sends IDC to the HFS-6, I immediately get 5-7 red bars, but as RPM's climb it scales back down to 2-3 bars. I'm wondering if this is an unexpected result of using the relay to limit IDC signal. Perhaps this is what is triggering failsafe? Is this normal and if not, what setting do I have to adjust to increase flow in upper RPM range?

Thanks for tip on unplugging tank sensor...will try that first to see if that's the issue.

Lastly, by widening the failsafe windows to avoid false triggers, will they still come on to protect my engine in the event of a legitimate failure?

[Richard L]

Not quite, set the SC to display 5-6 bars under WOT. Set the WL and WH as normal. Just increase the SLSF SENS on the controller to 4 o'clock and let me know if this works.

[Iworktofish]

SC set to 12o'clock, WL at 11a, WH at 1p, and FLSF SENS set to 4pm as suggested and it appears the problem is gone. No more failsafe after shifts and no more tapering of spray, at least for the 2-3 passes I made. Should I start narrowing WL/WH or are the current settings ok and will provide failsafe protection?

2nd question. It takes ~1-2 seconds for the system to start spraying again after shifting and applying WOT, as indicated by the return of 5-6 LED's. My concern is that for this brief moment I'm boosting past the recommended 13psi without meth. Is this safe?

[Richard L]

We hav establish the caused of failsafe trip. I would set the FLSF SENS to 3pm and start narrowing of the WL and WH.

The HFS-6 is triggered by IDC% only. Boost and IDC is not always in unison. At low gear the boost tend to build at higher RPM. Vice vesa at high gear the boost builds up at lower RPM. the higher the load, the quicker the turbo builds up.

There are no delays on the system. What is the THRES ADJ trimmer set to? Reduce it a little bit and tell me if makes the system response to boost ramp sooner.

[Iworktofish]

ok will do.

THRES ADJ was left at 12o'clock. Since I've got the IDC masked until 13psi anyways, I will reduce the threshold lower and see if it quickens the delay after shifts.

Thanks much for your assistance.

[Richard L]

Since you have masked the iDC signal, you can set the "THRES ADJ" to 9 'clock. Let me know.

I think the reason for the delay was due to the IDC% (set to 12 o'clock or ~42%) was not low enough under certain conditions when the boost reaches 13psi. Loweing the iDC threshold should overcome this.

[Iworktofish]

Richard,
I set the THRES ADJ to minimum CCW (7 o'clock?) before I read your above post and the delay is still there. I guess that is how long it takes for the turbo to spool back past 13psi after shifts?

[Iworktofish]

Hi Richard,
up until now I've been testing on local roads in 2nd and part of 3rd gear. Just the past 2 days I've driven the car on the highway testing in 3rd and 4th gear and discovered the original problem is still there. Once I shift, failsafe is activated and the duration of the next gear is on wastegate at 8psi down from 15psi.

Here's what I've done recently:
1. set FLSF SENS back to 4o'clock (max) from 3o'clock
2. disconnected tank level sensor
3. set Eboost to trigger relay to 12psi, 11psi, 10psi and then anything greater than 100rpm so HFS-6 always sees IDC

After all of the above, failsafe is still activating after I shift. I am out of ideas...please advise.

[Richard L]

Two things come to mind...

1. I have heard some discussion on other forum about the water pick up problems on the coolingmist tank. Here is the link (http://www.e90post.com/forums/showthread.php?t=420249&page=2). But one of the things to check out.

2. I believe you are the one who fitted a pressure switch inline with the IDC sensing wire. This may cause the some problem for the system to ramp up fast enough from 0% DC to 50% DC. This delay may be your problem as the system was seeing nothing and sudden expect to calculate the correct flow for a 50% DC.

Please disconnect the inline presure switch and see if the problem persists.

[Iworktofish]

I believe you are the one who fitted a pressure switch inline with the IDC sensing wire. This may cause the some problem for the system to ramp up fast enough from 0% DC to 50% DC. This delay may be your problem as the system was seeing nothing and sudden expect to calculate the correct flow for a 50% DC.

Please disconnect the inline presure switch and see if the problem persists. No pressure switch. I was the one that used the Eboost2's programmable auxiliary output wire to trigger a relay to mask the IDC signal. On bimmerforums you mentioned there should not be any ill effects to this. However, in my last post, I mentioned just for eliminating it as a possible culprit, I reduced the trigger from 12, then 11, then 10psi, then all the way to 100rpm so that it is effectively disabled and always sees IDC. No effect, so I'm certain this isn't causing the failsafe. It may be causing the delay though. For now I am leaving it set to "off" and turned up the THRES ADJ from 9 to 2 o'clock so that it is not activating prematurely. Again, the tuner suggests no meth before 13psi.

I have heard some discussion on other forum about the water pick up problems on the coolingmist tank. Here is the link (http://www.e90post.com/forums/showthread.php?t=420249&page=2). But one of the things to check out. Thank you for that link...I think you're on to something. Here's an excerpt from the thread:

After you install the barb fitting into the tank with the drill, remove the fitting by reversing the drill and cut a 7 inch section of hose and insert into the bottom of the fitting just as shown. No glue or tape is required. Now install the fitting by inserting the hose in the tank and tightening the fitting by hand. Once its hand tighten you can use a wrench to completely tighten. This hose is a pickup tap that keeps the system from drawing in air and allows you to use almost the entire tank before filling up. You must install the fitting as below so it goes all the way across the tank. Failure to do this will cause the system to draw air in. The plastic hose will fit in the fitting just like below without any glue. It will stay in place.

I bought the tank from Howerton, and it didn't have these instructions nor any diagrams. Jeff said he had them for a long time. The barb was pre-drilled into the tank with no 7" section of hose. I will see if I can contact coolingmist for a copy of the instructions to see if I can salvage this tank. Shuasha is using this tank on his car without any problems, but I'm thinking this has to be the culprit. While shifting, the fluid is possibly sloshing (even though it's full?) and is drawing air, causing the delay and/or failsafe. If this ends up being the cause, I will bite the bullet and purchase a better tank from Howerton.

I'll advise my findings after I makeshift some sort of 7" section of hose. Your support is truly appreciated.

[Richard L]

Please keep in touch, I would like to get to the bottom of this. It i spossible for me to modify the controller to allow you to dial in a longer failsafe-trigger delay but it will not be good for the long term.

[Iworktofish]

Hi Richard,
I was able to get hold of Coolingmist on their tank instructions and installed a 7" hose in the tank to draw from the open area where it will not see air. I made sure the hose stayed in this position and I also filled the tank to where it is completely full.

I just took a 15min drive. Unfortunately, still experiencing the failsafe. In fact, perhaps because my windows are set too wide (WL at 11o'clock, WH at 1o'clock, FLSF SENS to max) I'm actually seeing conditions where I'm boosting at 15psi with no bars and the HFS-6 is not interrupting the Eboost's solenoid to wastegate. This happened on 4 consecutive attempts, so I reduced the EBC to 13psi to prevent from catastrophic failure.

Since the tank was disconnected, I'm going to try and purge the lines to see if there's trapped air. It doesn't specify in the manual, but is it necessary to purge using the TST jumper each time the line is disconnected from the pump?

Any other suggestions...pinched meth line, trapped air, clogged jet or FCM, bad module, etc, etc, etc? I'm fairly certain my install was correct, but I'm willing to tear the car back apart and inspect each component until I can get this to work 100%.

[Howerton Engineering]

I just skimmed this and am wondering about the failsafe settings. When we set them up it is in this way:

Adjust the WL full CCW and WH full CW. Set failsafe delay to 12 o'clock.

Put the system in test for 1 minute and measure your jets into a container to ensure the system is flowing according to spec. While spraying 100%, turn the WH down(CCW) until the "B" light starts to flash or goes off, turn back to right 2 clicks. The SC and WH have now been set.

Take the car out on the road, in a high gear roll on the throttle very gently until the system starts to spray. Do this a number of times to get a feel for where it starts, turn the WL up a couple click and repeat process. Do this until the "B" light comes on goes off right at cut in then turn back to right two clicks.

Now try driving the car and see if the failsafe trips in normal circumstances. You can now fine tune the WH WL and delay to tighten the window even more as desired. Sometimes with smaller jet systems the SC may need to be at 7 bars at 100% for a good spread under normal driving, and other circumstances can call for some other adjustments as well.

I think the WH and WL just turned from noon two clicks may not have been an adequate adjustment if that is what was done.

I apologize if this is redundant, but want to see if the failsafe can be tweaked a bit.

[Iworktofish]

Jeff, thanks much for your input. As of now my WL is at 11 and WH is at 1o'clock. I will try this later tonight and will advise.

FYI, I think for WL you meant to say "Do this until the "B" light comes on goes off right at cut in then turn back to LEFT two clicks", correct?

[Howerton Engineering]

Yes, I meant left. Sorry for the confusion.

[Iworktofish]

Well, I did further testing and still no luck.

I forgot to mention, yesterday I purged the lines by running it on TST for about 3-4min. I also set the EBC down to 13.5psi, a safe level the car can run with or without meth. Tonight, I set WL to full CCW and WH to full CW, FLSF SENS to 12o'clock. Same thing...would come on fine at WOT once threshold is met, but after shifting it would go into failsafe. I then even set FLSF SENS back to max full CW, basically now having all failsafe windows open fully, no change.

At my next opportunity, I will test the system and measure the jets spraying into the container for 1min just to be sure it's flowing to spec. I will also completely bypass the relay for the IDC, even though the eboost is set low enough to where the relay is effectively bypassed, just to be sure. And I will check the pickup tube in the CM tank to make sure it's in the correct forward position, even though I can't see how it would suck in air with the tank at full.

For whatever reason, something is happening when I lift throttle and shift gears that is continuing to trigger failsafe. This is getting frustrating and I'm all out of ideas.

[Richard L]

We may have missed the obvious, a particially clogged jet?

The reset timer after failsafe trigger can be trimmed by using the FLSF RST

[Iworktofish]

thanks Richard. I'm not an engineer, but would a partially clogged jet cause it to work perfectly when applying throttle and reaching threshold, but then go into failsafe after I lift throttle and shift? I will swap out the 1.0mm jet for the next smaller one for testing.

I've played with FLSF RST before, setting it to minimum, but ended up putting it back to 12o'clock. Reason is because when the system was working fine (or I should say "better") and had delay inbetween shifts, I wanted to see if the delay was the failsafe being triggered at shifts and recovering. I suppose I can set it back to minimum, but I really want to get at the cause of what is triggering the failsafe.

Gameplan:
1. measure output into container for 1min to check flow
2. try smaller jet, perhaps current 1.0mm jet is partially clogged
3. jump leads 87-30 on eboost relay to completely bypass to ensure IDC signal is always present.

I could buy another tank for testing as the CM tank has been questioned before, but the pickup is at the bottom and I'm unsure how it can draw air with the tank completely full.

Let me know if you have other suggestions.

[Richard L]

The lift off coudl have something to do with the mis-matching between the IDC trigger point and boost pressure switch setting. I have not seen this happen before.

Do you use an inline restrictor?

I think the game plan is good idea. Don't get a tank until we have sorted this out.

What is you WL and WH setting now?

[Iworktofish]

WL is full CCW, WH is full CW, and FLSF SENS is full CW so the parameters for the failsafe window are fully open.

Per the instructions on pg9 of the manual:
0.5mm restrictor .................... 0 - 380cc/min
0.7mm restrictor .................... 0 - 680cc/min
0.9mm restrictor .................. 0 - 1080cc/min

From memory I believe I needed 550-600cc/min so I used the 0.7mm inline restrictor in the FCM. Let me know if I should have used the 0.9mm or none at all.

[Richard L]

I would just remove the restrictor for the time being in case there is a partial block there. This is a strange one for me especially on lift-off.

How difficult is it for you to run the system without the pressure switch, just to check if there is a mis-match of the two trigger points.

[Iworktofish]

No problem, I will remove the restrictor and give it a shot.

No pressure switch in the system. See post #12 on our discussion about the Eboost2 aux output and relay. And in my post earlier today, I mentioned I've already set the output on the Eboost2 so that the relay is triggered allowing full IDC signal to pass on any RPM >100, and as a precaution will be completely bypassing the relay by jumping leads 87-30 in case signal is still being cut for some unknown reason.

Thanks again for your help.

[Iworktofish]

I was able to measure output into a container as recommended by Jeff. It only flowed 250cc/min on the 1.0mm jet. I switched jets to 0.9mm and same, ~250cc.

I then removed the 0.7mm restrictor. It did not appear to have any contaminants as I could clearly see through the jet. Testing again, I now got ~500cc which was closer to spec.

Buttoned it up without any restrictor, went for a ride and now got full 7-8 bars even after shifting. Only did 2nd and part of 3rd gear, but after 3 tries still no failsafe.

I then turned the boost back up to 15psi, and set the Eboost2 aux to trigger the relay to send IDC signal at 13psi. No failure after another couple passes in 2nd shifting into 3rd. Looks like this was the problem.

So at this point, what is the diagnosis? Partial clog in the restrictor? Restrictor not supposed to be there at all?

Later this week I will work on closing the failsafe windows, as currently they are all set to max.

Thanks Richard and Jeff for your assistance.

[Howerton Engineering]

The pump could possibly be at a lower pressure than spec. There is an adjustment screw on the nose of the pump that adjusts this. I'm not sure if you need to at this point as you now seem to have it running as it should.

It's hard to say what the problem was without some troubleshooting and looking into the parts.

[Iworktofish]

Any drawbacks to not using a restrictor as specified in the instructions?

[Howerton Engineering]

The restrictor is there to place a similar orifice on each side of the valve to give it some linearity across the entire injection range. If running from a low IDC setting, you may see a hump in the injection curve in the middle IDC region without a restrictor. Depending on how you run the system, triggering it at the point you are, you may not encounter this issue. In reality, it just runs a bit fatter in the midrange, unless you are already seriously rich, it won't hurt at all.