Air flow meter tests

[august]

There's one test shown in the N9T/N9TE/N9TEA for checking the AFM:

Connect an ohmmeter between pin 5 and pin 7 of the AFM. Then move the

flap door slowly to full open.

It says it should be between 60 and 1000 ohms.

Now, I've done this test on 4 AFM's. You'd think that whatever the

value the book was calling for would be a clean sweep from low (full

shut), to high (full open). Not the case, on all FOUR that I have.

They seem to jump around a lot. As the doors opens, the value rises,

but can go up and down, and full open on all isn't the peak value.

(more like, 60(door closed), 100, 300, 150, 500, 800, 200, 900, 400

(full open))

Is this normal, or do I have a bunch of bunk AFM's?

Here's a sample reading:

full shut: 80

peak: 1000

full open: 450

Do any of you have experience with this? Or would any of you be

willing to test the AFM's you have? I've been looking around the net

to try to educate myself on exactley how these things work, but

haven't found anything on super bouncy readings.

thanks!

this was also posted to peugeot-l

[Bean]

I think I may have a spare AFM in the garage I can check out.

Any chance you can post the pin out diagram? I can find it for myself - but if you want others to check as well it might be handy for them to have. (Or is it labelled on the connector?)

It makes sense that it *should* be linear - but I'm guessing after all these years there may be some dead spots or "dips" in the resistance readings.

It'll be good to hear other results - but I'd guess that as long as there's a trend of linear progression they should still work.

Rabin

[august]

Any chance you can post the pin out diagram? I can find it for myself - but if you want others to check as well it might be handy for them to have. (Or is it labelled on the connector?)

It's labeled on the connector

Bill Bruce posted something to Peugeot-L about it being "logarithmic". I did a little reasearch and found out more info.. here's part of my return post:

From the Bosch book of Automotive Fuel Injection Systems Technical Guide:

Page 100

"The relationship between the mass airflow and the flap angle is

logarithmic, which ensures a non linear characteristic in the

potentiometer."

i take it logarithmic means its not linear (obviously), but it will

fluctuate during the complete sweep of the flap.

I asked another friend / peugeot homie who said:

"it should read the highest value when the flapper reaches to most

deflection it can attain with airflow pushing on it"

(post on Peugeot-L can be found here: http://www.505turbo.com/peugeotlist2/12/2005/5670.html)

[Bean]

I guess the thing to test is that all 4 AFM's have the same reading at max flap angle - one would assume that they should read identically at a given apeture - so I'd also test with holding the flap open with various objects to see if the AFM gives the same readings...

This is a great topic so I'm vaery curious to find out what the results are!

Rabin...

[august]

yep, that's a good point Bean, but unfortunately they all seem to have quite a range. And being the AFM is such a critical point to N9T** performance, its important to have a 'good' one.

I know that blackie's AFM peaked around 1300 ohms, which could partly be why it was so hot damn fast.

each one i have has completley different readings. one peaks at about 700, one 950, one 1100, and one i saw peak at 1247 (!!!), so you can imagine which one I used. Supposedly, its also important to have a good 'range', its not all about the peak. i.e. better to have one that ranges from 50-1000 rather then 170-1200.

as you can imagine, this value is a critical one when it comes to BID (basic injection duration), another being the NTC, TPS, etc etc.

[Bean]

Looks like the AFM itself is a reason to upgrade the ECU! My initial reason was because I see the vane style AFM as a choke point - but now we know they're flakey too...

Hopefully once some of us get to experiment with alternative engine management - I have a feeling that we'll really wake these cars up and see some of the potential they are capable of.

300 HP/ 300 ft-lbs should be a cake walk for these motors. 400 HP shouldn't be out of reach either...

As a side note - a buddy of mine that is going to help be do the swap ECU swap onto my car took a look at the N9TE head and said the Dodge guys would KILL to have a head that looked like ours. I'm tempted to drop it off at the machinist that ported and polished his 2.5L Turbo Dodge head just to see what it flows stock out of the box. The TD limitation on the stock head is about 300 HP - so if the Pug flows more than that we should be good to 350+... Now to see how much he'd charge just to flow the head...

Rabin

[JunktionFET]

Indeed those AFMs are weird. For some reason, when they get a lot of time on them they tend to read a bit low at their peak. Younger ones or ones that have seen less use than typical seem to read better. Needless to say, the peak value is probably the most important, though that doesn't mean the other measured points aren't important.

I have considered adding a tuning feature to the AFMs to tweak their range into something closer to what the shop manual specifies. I've thought of placing a pair of multi-turn trim pots on the AFM's cover to adjust both the closed reading as well as the peak reading, which would naturally affect the curve in between.

I'll have to look at my old experimental AFM to see if it's even possible, but that would be a great modification even for those who just want to tune their car stock.

As far as AFM restriction is concerned--more and more I believe this is hype. Though you see it on websites and discussed amongst so called experts regarding high performance cars, I think the biggest gain (if any) seen with an upgraded AFM is the modified fuel curve and not the restriction imposed by the original AFM.

To satisfy this curiosity a few years ago, I measured the vacuum in the accordian hose between the turbocharger and AFM on a Turbo we had running on 17 psi of boost, and it was barely over 2in Hg. Removing the air filter and airbox lid makes a bigger difference than that in pressure, and we all know how miniscule the power gains are from doing that.

Big AFMs are undoubtedly necessary for heavily modified engines running loads of boost, or as an alternative way to get a more desireable fuel curve. So are big AFMs good for gains in performance? Yes. Is it because of less intake restriction? Probably not in most cases.

[august]

interesting JunktionFET! good info for sure..

speaking of restriction, i think the 550hp 505 'PRODUCTION' race car used K-Jetronic (or some other form of CIS), and the fuel distributor has a similar type of method for measuring air right? its like a dish which the bottom of drops with the more air enters the system?

[Bean]

I totally agree that losing the AFM isn't the holy grail of power production - but to me it's definitely one of the areas for improvement. The simple fact that the air path has a door that is pushed aside for the air to flow tells me that it's not necessarily the most ideal flow path. The size alone likely doesn't affect peak power production - it's just the way it measures the airflow. Spring loaded doors just seem to be the wrong way to do it.

But it's a moot point if your just want a well running stock 505 Turbo - getting the car dialed in like Joe is famous for is going to net fantastic gains. I think the stock motor and set up (with tweaks) is going to easily hit the 250+HP mark. For me however - I'm pretty serious of my goal to get 300+ HP from this motor - and I want it to still be dead reliable. Since guys are getting this kind of power out of little Honda's, and Turbo Dodge guys are getting this out of their engines - I'm thinking it's entirely possible with the Peugeot. To me - In my mind - this has to be done with a more sophisticated engine management set up.

I could be wrong - but I really want to try it and see for myself...

[Guest]

I think I'll measure my AFM too. It's easy now, because it isn't attached to the car.

I've read that losing the AFM should lead to better throttle response, because the pulse wave generated from the twitch of the throttle plate doesn't have to travel upstream through the pipes, the ic, turbo and finally to the AFM.

Megasquirt uses MAP, so it'll be interesting if I one day manage to get the car working with that.

The head should be fairly good for a 8-valver. The shape of the combustion chamber and the ports are quite nice. The only drawbacks are that it is prone to cracking, and it has just 8 valves... I guess that the dodge heads have the traditional wedge-style combustion chamber, which doesn't flow as well as the peugeot head should.

It will be very interesting if you'd get those flow figures for the stock head. The maximum flow very much dictates the possible power output from the engine. I've read that a decent engine produces 1hp per 1,5cfm of air. Of course it can be compensated to some extent by turbocharging.

[Bean]

The only thing I don't like about the stock head is how shielded the spark plug is. Danielson found a way around it - but I still don't know how it was changed to get the 10mm spark plug in there. (I took a look to see if I could get mine changed like V-M did - but I had no idea how!)

I'm thinking that if I can do a custom piston (Volvo?) and get some more displacement and compression I can afford to open up the combustion chamber a bit to expose the spark plug a bit more. Even indexing the plug would help a bit. (Making sure the electrode is open to the chamber - if the regular plugs are being used.)

I may pony up for a Danielson head eventually (if mine cracks) - but It'd be a lot cheaper if I could pull a

V-M and just have mine changed...

I'm definitely curious on the flow characteristics - might have to get the head flow tested right after I get back from my trip...

Rabin

[JunktionFET]

I believe the placement of the spark plug is to reduce the tendency for knocking without the use of an excessively cold spark plug or wide gap (which would increase the tendency for misfire under light load, affecting economy, emissions, etc).

The intake valve is by far the coolest part in the combustion chamber because it is constantly chilled by evaporating fuel moving past it. So placing the spark plug near the intake valve ensures that the fuel will ignite slower and in a more controlled fashion, and without the need to retard the timing too much.

For a purpose-built or exclusively high performance motor, an idle misfire isn't much of a concern (hell, an idle misfire is actually sort of nice sounding with open exhaust), so in such an application it may be better to use a cold spark plug with a more central orientation.

At any rate, it is something to consider when looking at cylinder head modifications--the stock placement is definitely a crucial engineering practice and plug relocation should be carefully weighed based on the car's intended purpose.

[Guest adegnes]

Do you aply any voltage to the afm when testing it, I've tried, but the only reading I get is "1" (infinite resistance, no conection at all...) I did the test with two afm's disconected from the car's harness, both read the same...

[august]

Do you aply any voltage to the afm when testing it, I've tried, but the only reading I get is "1" (infinite resistance, no conection at all...) I did the test with two afm's disconected from the car's harness, both read the same...

i don't think i applied any voltage, i was just doing the tests on the bench if i remember right.

strange your getting infinate ohms, are you sure you have the leads to your DVOM hooked up right?

[Johnny]

I measured my AFM today, and I found that it doesn't have 7 pins, but 5. Anyone know why that is?

However, I got a comparable result if I measured resistance between pin 5 and 2 (5 beeing the pin closest to the airbox).

Here is what I got:

Fully closed: 107 ohm

Fully open: 440 ohm

Peak: 990 ohm

I also cleaned it because it had alot of grease and dirt in it, and after that I sprayed some 5-56 in it. I thought that maybe a drop or two would get to the potensiometer and maybe it could clean it a bit.

Was this stupid?

[august]

I also cleaned it because it had alot of grease and dirt in it, and after that I sprayed some 5-56 in it. I thought that maybe a drop or two would get to the potensiometer and maybe it could clean it a bit.

Was this stupid?

nah i don't think it was stupid. i've heard mixed things about cleaning out AFM's. some have said not to spray ANYTHING in there, and to just use a rag or something to clean it out. when i do it, i'll try to use an electrical friendly cleaner with a shop towel, and if i spray any inside, i make sure to keep the potentiometer side UP.

the potentiometer is sealed from the outside, so i don't think there's much dirt getting in there. every AFM i've taken apart the potentiometer is still really clean.

that's my $.02 , i'd be curious to hear what the L-Jet bimmer guys do.

[Johnny]

ok, I just thought that if it wasn't properly sealed (due to age) then the potentiometer would oxidise, and would work better if it was exposed to some anti-oxidiser.

But why does my afm only have five pins when yours have seven?

Is it supposed to be that way? (us versions and eu versions)

The number on my afm is: 0 280 202 057

[Guest adegnes]

i don't think i applied any voltage, i was just doing the tests on the bench if i remember right.

strange your getting infinate ohms, are you sure you have the leads to your DVOM hooked up right?

Turns out my DVOM was toast...

my AFM goes from 134 to 1280, full shut at 460.

[Guest adegnes]

ok, I just thought that if it wasn't properly sealed (due to age) then the potentiometer would oxidise, and would work better if it was exposed to some anti-oxidiser.

But why does my afm only have five pins when yours have seven?

Is it supposed to be that way? (us versions and eu versions)

The number on my afm is: 0 280 202 057

If I remember right, the number on my AFM is 0 280 202 044, it has five pins (I belive all of them do)

pin number seven isn't the seventh pin, it's just the number of the pin...

On my AFM there is no pin number 5, I have to do the measurings between pin six and seven... wierd...

I didnt get any reading before (my dvom was toast) with my new dvom I measured:

fully shut: ca. 130

Peak: ca. 1280 (highest reading 1287)

max open: ca. 460

Isnt that fully shut reading a little to high?

[Johnny]

I didn't find any numbers on the afm or the connector, I just assumed that they were numbered 1-5 because there were only 5 pins.

[august]

i'm going to try to remember to have a look tonight when i get home regarding part numbers, and pin's. this might be a good thing to do a HOWTO on..

[august]

alright, let me see if i can help clear this up:

But why does my afm only have five pins when yours have seven?

As far as the pin's go, see:

Is it supposed to be that way? (us versions and eu versions)

The number on my afm is: 0 280 202 057

If I remember right, the number on my AFM is 0 280 202 044, it has five pins (I belive all of them do)

The number on my US spec N9TEA (0 280 202 057), which is obviously the same as yours Johnny. I'm assuming your 88 N9TEA has a lamda sensor, eh? On the other hand, adegnes has a different number (0 280 202 044). I'm assuming that's a pre 88 505 turbo (N9T/N9TE)?

see:

I didnt get any reading before (my dvom was toast) with my new dvom I measured:

fully shut: ca. 130

Peak: ca. 1280 (highest reading 1287)

max open: ca. 460

Isnt that fully shut reading a little to high?

adegnes: 130 is a little high, but not considering your peak is 1280 (which is really high). check out this picture

As for the actual testing procedures, here's what the manual has to say:

(test 3)

III Electrical operation.

- Hook up an ohm meter between pin 5 and pin 7 of the air flow sensor.

- Using a screw driver (hah!), vary the position of the sensor flap

The resistance will vary with the position of the sensor flap

Correct reading: 60 to 1000 ohms

...of course what the manual doesn't tell you is the varying isn't liner due to the test being logrhythmic. the ohms will jump all over the place, which of course is what started this thread.

the peak ohms seem to be when the flap is about 3/4 depressed. the only thing i / we can come up with is this is the position where the flap would be under full load conditions. i think this is one of the reasons that causes the ecu to throw errors under much higher then normal boost conditions. the flap is open more then it should, yet things arent matching up and the ecu's are freaking out. this seems to start happening on the N9TEA around 20psi. now maybe if we used a AFM that didn't peak at 1000 ohms, we could have managed 23psi without codes (get what i'm saying?). in any case, for more info on HP and AFM's read some of the posts in the beginning of this thread.

are y'all a little more clear now? or did i just ad a whole new layer of confusion.

[Peder]

Did not know if anyone answered the question about 5 or 7pins on the AFM? Here it goes anyway;

The old ones with 7 pins the two last pins are for the fuelpump-shutoff, during motorbraking. The newer ones that is built in in the tps and the Ecu.

//Peder

[Guest adegnes]

alright, let me see if i can help clear this up:

As far as the pin's go, see:

The number on my US spec N9TEA (0 280 202 057), which is obviously the same as yours Johnny. I'm assuming your 88 N9TEA has a lamda sensor, eh? On the other hand, adegnes has a different number (0 280 202 044). I'm assuming that's a pre 88 505 turbo (N9T/N9TE)?

The car's an '88, the engine is a n9te ('86), don't know about the AFM (the one I'm not using is from '86, with the same numbers on it, so the one I'm using is probably '86 to.)

The spare one's readings are:

shut: 90

peak: 970

open: 440

Maybe I'll try it out, see if I notice any power-difference...

[Toni Lindroos]

I measured a Bosch 0 280 202 044 AFM from my Turbo #2, European 505 Turbo Injection 1985 (N9TE). The AFM has a blue plastic cover and the pin numbering is 6 9 8 7 27. So, I measured between pins 6 and 7.

Fully closed: 140 ohm

Fully open: 480 ohm

Peak: 1100 ohm

I haven't measured AFM from my white Turbo #1 or from the Turbo #3 (parts car) yet.