There are failsafe, failsafe and failsafes...

[rj0150]

I have a flow switch for safety that ties into my stand alone ecu, if it fails the stand alone changes its tune on the car.

* Registered users of the site do not see these ads.

[Angle88GT]

Good

[Aquamist]

This thread is due for an update...

Aquamist is the first to develop a turbine flow sensor to read flow in real time in 2001. A good few years late, other companies followed with varied success. It was most probably using an off-the-shelf plastic body sensor in harsh engine bay environment. Over the next few years, every wmi makers settled on the turbine wheel type instead of the paddle wheel type due to cost. But still plastic body based whilst aquamist continue to employ the metal type. Totally reliability.

Fast forward this to 2013 and after some serious research, a couple of wmi manufactures have stopped offering the plastic flow sensor all together (supplied by GEMsensors.com) and used some alternative methods to detect flow problems. The messages was loud and clear to other wmi makers. These sensors just continued to fail in service, some just after a few months. Whilst this was going on, the aquamist's in house designed and manufactured sensors continue to perform flawlessly.

The most surprising outcome to this is the way the some wmi makers changed the way of marketing the same old sensor. Instead of sourcing a more reliable one, they just tell their customers to reduce the methanol concentration to 50% for safety reasons (wise words indeed) while still marketing their product as methanol injection systems. Unfortunately those sensors continued to be unreliable even after methanol ratio was reduced. There are hundreds of engines at risk out there without a failsafe due to flow sensor failure.

It is a real irony that a failsafe is the first item to fail...

[SUMATIC808]

Originally Posted by Richard L

This diagram is only applicable for a System2d system. The orange area is part of our wiring suggestion to the Sti boost control valve without triggering CEL activation.




Richard

with that "dummy load" resistor what size resistor would be used??
i am using an AEM meth injection system and looking for an alternative route for a failsafe. and this looks like it could work, could you explain more on how this route works? for example, like what happens when the relay is switched to the dummy load circuit, will there be a set boost psi that will used and how would it pull timing etc. is this as safe as using a limp mode as a failsafe?

again i am using an AEM water injection system and AEM failsafe system in my 06 sti. the style failsafe in the picture looks like it could be used in a universal system am i correct?

[boardinjorden]

thank you! this helped a ton

[Aquamist]

Originally Posted by SUMATIC808

with that "dummy load" resistor what size resistor would be used??
i am using an AEM meth injection system and looking for an alternative route for a failsafe. and this looks like it could work, could you explain more on how this route works? for example, like what happens when the relay is switched to the dummy load circuit, will there be a set boost psi that will used and how would it pull timing etc. is this as safe as using a limp mode as a failsafe?

again i am using an AEM water injection system and AEM failsafe system in my 06 sti. the style failsafe in the picture looks like it could be used in a universal system am i correct?

I believe this is best leave it to AEM's tech guys. I cannot verify or confirm this arrangement will work with the AEM system. All our newer system have this dummy resistor in built. the value will depend on different ECUs.

[Aquamist]

There has been a great deal discussions between the two most common failsafes algorithms. Progressive Failsafe vs Band-gap Failsafe methods. Here is our respond to the merit and pitfall of the two...

If you think the Aquamist failsafe can only detect "too little" or "too much" flow as expected from a fixed failsafe window width, think again. Our competition was preaching this to the masses, but they should have delved a bit deeper before launching their negative marketing campaign.

Simple answer, the “hi-lo” trip can detect a partial clog. If you set the "lo-trip” to 150cc/m and "hi-trip" to 500cc/m, anything outside this window will trip a failsafe. So though it seems Aquamist can only detect “too little” or “too much” flow, if one looks a little bit closer things become more clear…

Suppose the flow is dropped by 10%, (now 135c/m to 450cc/m) due to a partial clog. The failsafe will trip at the "lo side” of the window at the beginning of a spray event. Vice versa, if flow is increased by 10%, (now 165cc to 550cc/m) due to a leaky hose, the failsafe will trip at the "hi side” of the window. So who is not telling the truth?

Aquamist has tried the dynamic window method but it performed badly under a fast dynamic load change on the bench as well as in real world tests. Simple tasks such as a normal gear change will trip the failsafe at 10% error band. The failsafe continues to trip until it is set beyond 40% error band. This is caused by the turbine flow sensor over and under shooting and not keeping pace with the flow change produced by the fast acting valve. The only way to solve this problem is by applying a powerful electronic low-pass filter, but we have discovered that even doing so, the overall failsafe response time is too slow, much slower than the current setup. Consequently, we have decided not to offer this option until we can address the problem satisfactorily. At the present we prefer 10% error to 40%.

Maybe the much hyped up "dynamic-window" failsafe system only works with pump speed systems, which is probably due to the system's inability to respond quickly load changes. A few Aquamist users tried to mix and match and ended up with a very loosely-defined failsafe window and eventually returned to the original band gap method (failsafe window).

The latest HFS4 is equipped with an additional 0-5V input. It can be used in parallel with the standard failsafe. If AFR, EGT or IAT exceeds a certain set level, it will drop boost or switch maps. To us, looking ahead or capturing a problem as it happens, is a far better way, even if there are delays and over generous error bands. Take a closer look at the distribution of the captured flow curves in the advertised images, especially at the lower section of the flow range.

[JMlegacy]

Great thread Richard.

I'm interested in a DIY abnormal pressure detection failsafe

pressure too low > trip relay for failsafe
pressure too high > trip relay for failsafe

thing is, i don't know what device i can use for this. I'd much rather NOT use two hobbs switches for this.

Isn't there some device like a digital pressure sensor that can set two pressure points and trips if pressure is outside of this range?
I'd install the pickup at the nozzle

[JMlegacy]

also, for boost threshold/trigger point, what switch can I use for that since the hobbs may not be suited to the high temps of engine bay?

[buggeye1428]

I know u said diy. But I'm not sure if u have heard about the aem fail safe. It is a gauge to measure the pressure of the meth or water, and if it goes to low it lowers ur boost so u don't blow ur engine.( I'm not exactly sure how it works just wanted to throw that out there)

U could rig a gage into the line to measure the pressure of the meth or water. You could rig something though that if it's an electronic gage to a boost controller to drop the boost and get the same effect as the aem fail safe.

[buggeye1428]

Not sure if that is what u were looking for, just wanted to throw it out there!

[JMlegacy]

Yes i'm aware of most of the failsafes out there.
I'm currently studying to build an arduino circuit that would control via PWM a fast acting soleniod that is mapped to MAF voltage.

If richard / aquamist can chime in, I'd love that.

BTW, this is why I LOVE aquamist guys. They share, and share and share. IF i had the money I'd have gone with them flat out! But i also like the hobby of electronics and computer science

[buggeye1428]

That sounds lik fun to build it you self! Lotta planing and work but fun!!!!!

I just posted a thread on alky intros vs AEM, could u give me some input?

You seem knowledgeable on the subject! I'm just getting into meth injection!

[Aquamist]

Originally Posted by JMlegacy

Great thread Richard.

I'm interested in a DIY abnormal pressure detection failsafe

pressure too low > trip relay for failsafe
pressure too high > trip relay for failsafe

thing is, i don't know what device i can use for this. I'd much rather NOT use two hobbs switches for this.

Isn't there some device like a digital pressure sensor that can set two pressure points and trips if pressure is outside of this range?
I'd install the pick up at the nozzle

Originally Posted by JMlegacy

also, for boost threshold/trigger point, what switch can I use for that since the hobbs may not be suited to the high temps of engine bay?

This set up should work well for a "single point" spray system where operational pressure is clearly defined. You can use an electronic pressure sensor but you need to build a high/low comparator circuit to make sensor of it.

Mechanical Pressure switch is prone to drift over time in the engine bay environment and seasons.

[JMlegacy]

OK. Nice.