Right on.... I'll have to tweak a bit... the brackets that came w/ the blitz set are a little weak, and I think it's having a bit o' trouble staying where it needs to.... oh well... time to BUILD SHIZZLE!!!

 

Mark, you're funny, man...Those voltage readings I gave were the same as the readings I took a long time ago, when the stock intake system was still on the car (I've got a record of all normal input/output readings "on file" for any future reference relating to modifications, or troubleshooting, that I recorded just for purposes such as this) So, I'm not seeing "wrong readings", everything is as the FSM says it should be...Now, pray tell, if you actually know of someone that blew their motor because their AFM was at 90 (and no other reason, such as overboost, aged seals, etc..) then tell us...

A/F ratio is also running just as it was before the this AFM/ intake mod...Perfect .4 to .5v during cruise...

 

Then open it up and look. You'll see it's not counter-weighted. It has a damper flap mounted at 90deg so in fact there is twice as much mass to be affected by vertical movements. It's not counter-weighted because if you mount it correctly it does not need to be.



You can see pretty easily from this picture that any sharp movement in a vertical direction with the hinge horizontal will cause the flap to react. You may not necessarily feel it, and you probably won't see it on a DMM, but it's still movement that shouldn't be there.

When every single manufacturer without exception mounts their AFM's the same way (horizontally), you know it's because they done a hell of a lot more testing than anyone here and decided which way works best. The fact that they're never mounted vertically is a clear indication that there are good reasons not to. Manufacturers don't always do things the best way due to other considerations, but in this case the numbers are pretty overwhelming.

What I don't understand is why anyone would feel the need to mount it vertically anyway. Why not just do it right?

 

If these spikes are not picked up on a DMM, then in the real world how important is it to have the afm exactly level? Are you implying that the voltage change caused by a sharp vibration occurs in less time than is needed to actually pick up a change on the DMM?

It would be very easy to log the afm voltage input on the e-manage and find some nasty bumps to go over, but it would be nearly impossible to distinguish between a normal voltage flucation from slightly letting off the throttle, and a afm door that is being affected by the sharp bump.

Nearly every TID I've seen has at least 20-35 degree's of angle above horrizontal to it. To put the AFM back onto a completely horrizontal plane makes the creation of a TID much trickier.

 

No, I'm saying that the digital numerical display on a DMM can't show you small, momentary changes in readings very clearly (same reason you don't use numerical displays for tachos and boost gauges). You'd need a graphical display to see it, plus a fast sampling rate.
A slight upward inclination is not nearly as bad a 90deg rotation. It is easy to make a TID that keeps the AFM perfectly level, but most people (me included) are simply too lazy to weld two bends together instead of using just one.

 

See that is the key, he (WAYNE88N/A) is not even using a DMM, but an analog meter with a sweeping needle.

You need to scope it to see the real variations.

 

OK guys, trust me, I'm not a complete hard-headed idiot here, I'm the type that's always open to suggestions and new ways of "doing things", yet also the type that won't rush out and do something just because "everyone else is" until I've checked it out myself...No, I don't use a Fluke on the car because the damn thing is TOO sensitive, they jump all over the place, and you can't tell if the display is showing a smooth linear curve looking at a bunch of numbers (I've used them for years on the aircraft, don't forget, along with various types of meters and test equipment most of you guys have probably never heard of). The analog needle works better than most people give it credit for, for a good example of this, note the "signal" pickup (or lack thereof) that can be discerned on the injector circuits during operation using an analog. You'd have to have a duty cycle function on a digital meter to pick this up visually. I'm not dragging the o-scope out to the car, either, I see no need for it. As I posted earlier, the voltage inputs are instantaneous with peddle movement, no lag whatsoever, but the fuel changes are a different story, I would guess somewhere around .4 to.5 seconds elapsed , maybe more, before the engine reacted to the AFM inputs. In conclusion, high frequency "bumps" are not going to affect the fuel flow IMO.

As stated, everything posted here has been taken into consideration (including the fact RP in Dallas, noted for some of the fastest RX7's in the world, sells the Bonez kit which instructs the user to position the AFM in a 90* orientation), and if any form of performance degradation pops up, I'll move it back to the OEM position. Until then, I've satisfied myself that nothing weird is going on...

You guys keep talking about the "weight" of this door in relation to gravity- in effect, because of the spring, it has no weight, or at the least, weight is negligible when taking into account air flow pressures and spring forces working against each other...

So, in short, I'm hearing you guys, I'm not trying to convince everyone to put their AFM's in the hinge down positions, just that so far, after more than 3000 miles with it in this position, I have had no performance or fuel economy problems, nor see any electrically...

 

mine looks like this and i dont have a problem

 

Again you are missing the issue... its not the weight (although that may have some minor effect on it) it is the sensor's reactions to impacts not absorbed by the suspension (in other words= BUMPS IN THE ROAD).

So, a large bump with the AFM in the 90 degree off position and suddenly the AFM Flapper moves more than just air flow would result in.

This is clearly an error. The sensor is no longer reporting accurately. This means the car is not at optimal levels.

You can argue your way all you want, but there is no escaping that fact.

Of course if the flapper was dirty and or the spring was bad I would imagine that you too would add the problem of the flapper moving too much anyway.

Now if the AFM was positioned properly, then the only time any bump in the road can effect operation is if you were to hit something head on.

And as far as electrically if I had an analog meter that has a huge huge buffer (as most do) you probably would not see any difference. You can keep using that if it makes you feel better (although I would not ever use one on a modern car or truck). Of course when I tested I used a hand held scope on it with mapping so it is pretty easy to play back after the fact and see the problems.

But for all you joes who claim it works fine 90 or even 45 out, go do a road test and see. Drive a road course and check your time, then put the AFM right and drive it again...I know which one will be faster.

 

You are implying by that statement that RP believes the AFM should be mounted vertically, which I highly doubt they do. They simply sell a kit as-is to make money.
As Mark has said, it's the flap's reaction to the car's rapid vertical movements that are the issue, not the actual weight of the flap. Because the flap is hinged in the horizontal plane, when the AFM body moves up or down suddenly, the flap lags behind it slightly despite the spring tension. So the reading is briefly wrong.
True, but you seem very reluctant to admit that there is an undesirable effect from doing this. And I still don't understant why anyone would want to do this in the first place. Why not just do it right?

 

Ok, now my AFM doesm't look a thing like that.....

 

That's because you have a Series 5, which has a sliding cone AFM, as opposed to the Series 4's flap AFM.

 

i'm just gonna make mine fully open all the time, and see what it does

 

your car won't run because the ECU will be injecting more fuel than the engine can even swallow at idle.

 

Well you may be right, that slug of metal on the opposite side of the pivot does look a little small but I would have to take the flapper out and hang it from a string before I would say that it isn't counterweighted.

Each leg looks about 120 degrees apart. that slug would only need to weigh as much as one flapper leg.

The direction you mount the flapper is more likely based on the quality of the bearing the door would move on in each orientation.

ed

 

My AFM is mounted about 15% from horizontal. I have a voltmeter hooked to it while driving (its called an SAFC) and even HARSH bumps do not affect airflow significantly on cruise control. (0.1% MAX)... almost nothing. If that. Most of the time there is no difference at all.

While not on cruise control I think that my foot will move more due to the bump than the AFM flap.

 

You mean 15deg? That's probably fine, as it'll have considerably less effect than 90deg. In fact according to the trig it's about 7% of the effect.

 

I like the post about how the foot will be more/ just as likely to cause a change in the afm's output. A dose of common sense. My afm is installed in the factory position and when I do intall my K und N, it's installed in the factory position. The foot remark makes sense though.

 

sorry messed up

 

Wow, somebody brought this back from the dead, lol...

In the interest of covering the entire spectrum of opinions given here, I mounted mine back to 0 degrees just to see if anything changed. It's mounted less rigid in this position (had 2 metal bracket mounts in the 90* position), and rests on the wheel well structure at the corner (a little RTV in between to prevent actual metal-to-metal contact) due to the cold air bulkhead installed, but it's good enough for now. Performance increase? Eh, can't tell, maybe so, maybe psychological...

Also correlated something interesting from the O2 sensor inputs- while cruising, the needle bounces between .4 and .5 about twice a second...Now, without further "frequency type" inputs from any of the other sensors to compare with, I've gotta assume that the ECU changes fuel flows every .25 seconds or so. Will a vertical "bump" affect flow at this adjustment rate? I dunno...If anyone knows exactly what the ECU's injector cycle adjustment speeds are, chime in (I'm not talking duty cycle here, but "changes" to the duty cycles per second).

I'll go ahead and leave the AFM level for now, to keep you happy, NZ

 

since we're on afms =)
the afl level on my safc2 reads 1.0 - 1.1 on startup and idle.. is this normal?
sometimes it will be .4 on startup.
always have been wondering if this is normal or if it could be because my afm is at 15 deg down.

 

I'll go ahead and leave the AFM level for now, to keep you happy, NZ

Bah, humbug. Makes me want to mount it in a downward direction and cocked to one side and loose in its mounts so it'll jiggle around some.

 

That's the ECU going into closed loop and calculating fuel requirements using the O2 sensor instead of the internal fuel maps. The ECU doesn't really read the specific mixture, it's just looking for a number that's richer or leaner than stoichiometric. If the reading is "high" (rich) it'll cut back on fuel, and if it's "low" (lean) it'll add fuel. It performs this cycle many times per second (the healthier the sensor the higher the frequency), and if you have an LED A/F gauge you'll see the display rapidly oscillating back and forth around the center of the scale. Despite what many people seem to think, this "light show" is actually telling you your engine is running properly. It only happens under cruise conditions.
Closed loop is totally sperate from the AFM's reading, so it should have no effect. In fact I'm not even sure the AFM's readings are used during closed loop.

 

Yes, NZ I know all the "closed loop" stuff stated above; good point on the AFM inputs during that time, never thought about that before..."Many times a second" doesn't answer my earlier question, though. Can we assume that the rate of change displayed on the meter (or the pretty little LED's) during stoich is also the rate of change made to fuel flow during all other transient events? If so, then we're looking at about 4 or 5 times a second...Another interesting observation I made while looking at all the ECU inputs while driving is that the BAC valve actually energizes during decel (throttle lift) at cruise, at a 10% to 20% duty cycle or so. You can also see the results of this on the O2 sensor inputs- she'll go way lean...In any case, back to the original topic at hand- all sensor numbers were the same with the AFM at 90 or 0, there's just one tiny difference I can see- she idles about 50 rpm lower with it at 0, as compared to the 90 position. Figure that one out...

 

Ok, now I've been playing with that damned AFM, and I'll get it almost perfect, it'll pull like a banshee and then the next day it don't pull worth crap.... the only inconsistant variable that I'm using is fuel.... Sometimes I use 89 oct, sometimes 87... But it's gettign to the point where I'm having trouble keeping the AFm at that "sweet spot" for optimal flow, but at the same time, since it's a S5, and the ecu should filter out everything, could it be that the AFM is going bad at only 80,000 miles?