[OpenFlash]

Lots of talk about methanol injection failsafes lately. Quite a bit of misunderstanding. I'll try to explain the differences and why certain units are offered over others.

First, all the current US built methanol injection systems (CM, DO, AEM, Labonte, etc,.) use off-the-shelf flow meters supplied by companies like Gems. There are two popular types. Turbine-wheel type and Paddle-wheel type.

Turbine-wheel Type:



The turbine wheel is set in the path of a fluid stream. The flowing fluid impinges on the turbine blades, imparting a force to the blade surface and setting the rotor in motion. When a steady rotation speed has been reached, the speed is proportional to fluid velocity.
- For: ............. free flowing
-Against:......... Not as linear to flow as the paddle until an optimum speed is reached

Paddle-wheel Type:


It consists of a simple impeller with radial vanes, impinged upon by a single jet. like the paddle wheel on a Mississippi riverboat.
- For:............ linear to flow
- Against:....... quite restrictive

As you can see each design has their own strengths and weaknesses. And it's quite easy to identify which one you have in your particular failsafe. The theoretical benefits to each sensor is clear. However, in the real world, neither are suitable for methanol injection applications due to the materials used. As we have all seen, many times over, all these sensors eventually start to work erratically (or not at all) when running in a methanol injection system for any length of time. Depending on methanol concentration and hours of usage, these sensors can last a matter of days, weeks or months. The bearings, housing and rotation assembly start to swell and deform in the presence of methanol. This is because they were not designed for this particular application. But they are cheap, readily available and semi-suitable when volume flow and methanol concentrations were low (under 25%).

Faced with this situation, there are 4 alternatives for the US meth kit manufacturers:

1) To ignore the problem, recommend that people run a low meth %, and replace the units as they fail.

2) To design and manufacture their own custom-made flow meter built with components suitable for methanol injection.

3) To purchase and resell a flow sensor from another company (ie, Aquamist UK) that is suitable for methanol use.

4) To convince the uninformed that you don't want a flow sensor AT ALL and there are better alternatives to measuring flow.

Strategy #1 is a popular. It is followed by most US meth injection kit sellers. And this is why failsafes have gotten a bad rep among us enthusiasts. Over the life of a meth kit, it isn't unusual to have replaced the failsafe a few times. The blame is usually "hey, you used too high of a meth concentration" or "there was some foreign material that lodged in the flow wheel."

Strategy #2 is too expensive and really quite difficult. Designing and manufacturing a flow sensor from scratch requires a pretty big monetary investment and some good mechanical understanding. Time and money is often spent elsewhere (other product development, marketing, etc,.)

Strategy #3 is too expensive and not very feasible. Because instead of getting their own failsafe boxes made for say $40, they would have to purchase an Aquamist sensor for $120 and then explain to their customers why they have an Aquamist sensor under their hood.

Strategy #4 is a new one. And one that is trying to find a foothold among the uninformed. The argument is that all flow sensors are crap/unreliable and that the best way is to do away with them completely and monitor something like pump current (ie, the amount of electrical "work" accomplished). The problem with this is that pump current is NOT a direct representation of injection flow. Nor is current even linear with injection volume. And current changes as a function of pump temp, injection fluid mixture, pump wear, system pressure, etc,. A good analogy would be to measure vehicle speed (mph) by monitoring fuel consumption. Is there a relationship between the two? Sure. Is it controlled and does it exclude other variables? Not at all. Trying to sell such a product is like an engine builder who is incapable of building a good engine saying "engines are bad, let's do away with it and put a wind sail on the roof. It's far more reliable."

So when you are in the market for a failsafe, do your research. Understand the issues. It will have you a lot of hassles (and maybe even an engine) in the long run.

My 2c,
Shiv

[Sniz]

my 2c, go with Aquamist.

I spent $995 on my meth system with no tank, dont regret it at all. Totally generic, had to be adjusted for use in my vehicle. No issues. Great control.

[tscdennab]

Vishnu meth thread of the current hour

[themyst]

So where does SnowPerformance fall under? I've been using this thing since way back when you came to NYC in January, no issues except for one day in 100+ degree weather.

BTW, DO doesn't even offer a flow sensor.

[GeorgiaTech335coupe]

Measuring pump current to acquire flow would also give inacurrate readings for PPS systems during start up of the pump which requires more current, skewing the current v flow model used for conversion. This could cause on/off throttle transitions to give false high readings of meth, as well as post shift for MTs. It is arguable that the amount of time for the discrepancy could be negligible as well as when this occurs, but it is definitely not optimal.

[OpenFlash]

Originally Posted by GeorgiaTech335coupe

Measuring pump current to acquire flow would also give inacurrate readings for PPS systems during start up of the pump which requires more current, skewing the current v flow model used for conversion. This could cause on/off throttle transitions to give false high readings of meth, as well as post shift for MTs. It is arguable that the amount of time for the discrepancy could be negligible as well as when this occurs, but it is definitely not optimal.

This can be "solved" by allowing for a delay before a failsafe event is triggered. And by allowing for a wide enough window of measured current. But yes, both "fixes" takes a few steps away from the idea of a failsafe (ie, to catch the problem at the beginning, not when it becomes a gross flow failure). I'm afraid that many people don't understand this and look at these devices as magic boxes. People should understand what is going on because they are, quite literally, betting their engines on it.

Shiv

[messi]

man.. am i happy that i read these last couple threads on this. I had a CM kit on a car and that flipping flow meter gave me so much beans, inaccurate readings, low flow etc all kinda crap.. would definitely not go with anything cheap again. It also gave erratic readings as it the flow meter would light up all the way and then taper down and i used to wonder how that made sense. I was told this happens because the flow meter was close to the pump. Not sure if that made sense? Maybe the techs can fill me in.

[Rotorocious]

The 911 method

[rwalker]

The Aquamist sensor is a turbine-wheel type made of methanol-resistant materials?

[335i101]

BMS has a great WW kit, that's what I bought and would recommend to all!!

[///M3THOD]

Is it possible to integrate a traditional turbine or paddle failsafe with a reading of the pump current?

Its seems to me that in any given system/s where the failure of a single component could result in a catastrophic failure, there are redundancies in place to prevent this.

Since the traditional failsafes are physical and inhibit flow, perhaps supplementing this with a current reading, while not the most accurate method, could prevent complete failures simply by offering a redundancy. It seems that the only other way to do this with the current selection of failsafes would be with 2 traditional failsafes which would force the pump to work harder and decrease longevity.

Just a thought.

Originally Posted by 335i101

BMS has a great WW kit, that's what I bought and would recommend to all!!

Can we please not do this ^^^^^^^^^

[OpenFlash]

Originally Posted by rwalker

The Aquamist sensor is a turbine-wheel type made of methanol-resistant materials?

Yes absolutely. It was designed and manufactured in in-house since 2002, and being constantly updated and improved. It combines the principle of above types, resulted in good linearity and reasonable ease of flow. Fluid is redirected to flow on the bore surface by an internal scroll, driving the tip of the turbine so rotation speed is very linear to flow. With this design, there is no need for jetting (paddle type) and guarantee turbine-tip contact than the loosely defined fluid/turbine blade contact point (turbine type). It's also 100% methanol compatible. Unlike every other methanol flow sensor currently available.

Here's a see thru version of it:

[OpenFlash]

Originally Posted by ///M3THOD

Is it possible to integrate a traditional turbine or paddle failsafe with a reading of the pump current?

Its seems to me that in any given system/s where the failure of a single component could result in a catastrophic failure, there are redundancies in place to prevent this.

Since the traditional failsafes are physical and inhibit flow, perhaps supplementing this with a current reading, while not the most accurate method, could prevent complete failures simply by offering a redundancy. It seems that the only other way to do this with the current selection of failsafes would be with 2 traditional failsafes which would force the pump to work harder and decrease longevity.

Just a thought.

I'm not sure how a proper failsafe can fail in a way that could cause engine damage. In the case of the Aquamist flow sensor that we use, we read the raw digital signal provided by a hall effects sensor which reads turbine rotation. And the turbine won't rotate unless there is meth flowing past it. I can see a failsafe that outputs an analog voltage potentially failing and spitting out a false voltage when there is no flow. But even that is rare. The biggest issue with the current US-built flow sensors is that the internals actually deteriorate in the presence of meth. Which prevents the turbine/paddle wheels from rotating freely. So they end up giving lower flow readings that what is actually being sprayed. So at least they will fail in the desirable (for engine safety) direction. The biggest hassle, in this case, is just not knowing if meth is being sprayed or not.

Shiv

Can we please not do this ^^^^^^^^^

+1

[rwalker]

Originally Posted by shiv@vishnu

Yes absolutely ... It combines the principle of above types, resulted in good linearity and reasonable ease of flow.

Are pump internals subject to deterioration under high meth concentrations?

[E90SoFlo]

Originally Posted by ///M3THOD

Is it possible to integrate a traditional turbine or paddle failsafe with a reading of the pump current?

Its seems to me that in any given system/s where the failure of a single component could result in a catastrophic failure, there are redundancies in place to prevent this.

Since the traditional failsafes are physical and inhibit flow, perhaps supplementing this with a current reading, while not the most accurate method, could prevent complete failures simply by offering a redundancy. It seems that the only other way to do this with the current selection of failsafes would be with 2 traditional failsafes which would force the pump to work harder and decrease longevity.

Just a thought.

Can we please not do this ^^^^^^^^^

Isn't that what shiv does in every thread? GTFO

[secretsquirrel]

Originally Posted by E90SoFlo

Isn't that what shiv does in every thread? GTFO

No, the difference is he actually provides technical details as to why one method is superior to the other. Even if you don't like his methods, its hard to argue with 90% of what he says.

[Syndicategt]

Originally Posted by secretsquirrel

No, the difference is he actually provides technical details as to why one method is superior to the other. Even if you don't like his methods, its hard to argue with 90% of what he says.

Great information Shiv! Agreed with above, too bad 90% of members on this forums provide 0% technical information to these threads.

[Brey335i]

I was under the impression that the Aquamist flow sensor was also smaller than most of the other failsafes and therefore easier to position close to the nozzle. Or is that incorrect?

[OpenFlash]

Originally Posted by Brey335i

I was under the impression that the Aquamist flow sensor was also smaller than most of the other failsafes and therefore easier to position close to the nozzle. Or is that incorrect?

Yes, they are pretty compact. The customized flow sensor we use in our kit is even smaller since the Procede reads the raw hall effect sensor signal (no need for any additional circuitry to convert digital to analog. In fact, it's just a small tube with a sensor clamped on to it. And it screws direct to the inlet side of the injection valve.

The flow sensor is screwed on to the right of the valve. That is the hall effect sensor connector that clamps around the tube (where the turbine is located). Flow goes from right to left.

[hemi to e90]

shiv, is the analog sensor only getting voltage/signal from the moving components? am i understanding this correctly?

[Q4P]

Man, I have no idea why everyone is shitting on Shiv. He is not my best friend but he does provide facts with every single post he makes and backs them up. I doubt anyone's knowledge is as advanced on the subject as his. It seems that nearly every useful JB feature / innovation has been copied from the Procede. The fact is that there were two engine failures recently that people want explanations to and as it just so happens both were running JB tunes. While this may have no relation to the actual engine failure, the fact remains that they were both running JB tunes and until someone can accurately pinpoint what happened it is quite low to resort to name calling.

[GeorgeSmooth]

So when can we see a flow sensor upgrade kit on the old "not so good anymore" kit you use to sell?