Alternative MAFs?

[manga_blue]

I should've done this years ago.

We scoured the Batemans Bay wreckers yesterday looking for anything suitable, preferably Nissan. The best fit was out of a '97 1.6 Pulsar. Certainly the donor looked pretty unpromising so the wrecker lent it to me, together with the airbox, the IAT and the two pigtails. At least the throat area was marginally bigger than the NB8A's MAF, meaning it was more than double the NA8's. The general design looked cleaner than both as well.

Bench testing back in the shed with the vacuum cleaners showed the Pulsar MAF had the same output curve as the NA8's, while the NB one seemed to work backwards Hair dryer testing showed the IAT met Mazda workshop test specs too. So it looked like it was worth a go. I cut the old MAF plug off, taped on the pigtails to the stubs, rammed it into the air intake and tied it in with cable ties. Started and idled first pop - all good - so I took it out to a quiet place and just pottered around until it was warmed up.

I stopped at the bottom of a long steep hill, turned on the logger, idled it into 2nd and planted it at 1,700. Freakin awesome acceleration building all the way through until it slammed the redline at 7,500. It used to start fading at around 6,900 and wasn't worth going past 7,200. Dodgy Mazda tacho was actually reading 8,000

Log curves for the MAF(green) suggest it's sucking about an extra 10-15% of air - definitely feels like it too. AFR's (purple) remain the same as they were with the NA8 MAF, apart from that little dip below 12.0 above 7,000. Check how the acceleration (black RPM) doesn't taper at all until fuel cut just after 7,453. X-axis is seconds.

N15 MAF 2nd gear uphill.jpg

Thanks Paul for the Nissan prompts, thanks Fatty for throwing in the hairdryer suggestions (wouldn't be a hairdresser's car otherwise, would it? ), thanks Matty, NB MAF not needed.

[broady]

Whoa, so this is a viable upgrade?

[manga_blue]

Maybe it is, maybe it isn't. All I have now is a Pulsar MAF held in a fairly modded MX5 by a collection of cable ties and electrical tape. It's been tested to prove that it pulls really well at full throttle but I haven't explored what it's like to live with yet. In any case this is only a goer for an NA8, which is probably the least common MX5 model.

The original objective was to find something that gave higher airflow and use the tuning capabilities of this engine setup to overcome any running/tuning problems that arose. This engine setup includes a ported head, re-shaped chambers, performance valve springs, full 4-2-1 exhaust system, CAI and stock cams and stock throttle body. The ECU is a cracked OEM that allows remapping fuel, timing and rev limits. What this upgrade hopefully does is release potential in the motor which had been locked up by an undersized MAF.

What's promising about this trial was just how plug and play it was. Only 5 wires in one bundle at the top of the air intake system are cut and re-soldered. It does not appear to materially alter AFRs. Pulsar MAFs should be cheap and plentiful. The mounting from MAF to airbox will need to be fabricated. It would be interesting to see what it does to a more stockish setup.

[Matty]

So were you running the same CAI/intake setup? comparison is meaningless otherwise...

Also, from what I vaguely recall of the NA8 MAF, it actually opens up internally as the air goes past the central "bullet", so the cross section isn't as compromised as it first appears if you just stare through it.

The other thing you need to do is a manometer test to see if you have actually improved the airflow or not. Dead cheap to make (you can borrow mine as it is now redundant to me), and very meaningful results.

[manga_blue]

Matty, I could go forever with the direct empirical approach. I did remeasure the NA8 MAF - not easy but my best guess is a minimum throat of 1535mm2, which is still 28% less than the throttle body.

What I'm actually after is improvement in acceleration and throttle response. So the easiest way of knowing that I'm getting there is to log speed against time. 0.3 secs in a 0-100 would be very nice but 4500-7000 is where I most want the improvement for track work.

So I took the car out again today and re-did a series of 2nd gear pulls, on the flat, no wind, same road, CAI, fuel, tyres, pressures, temperature, load, fuel map, timing map, etc, etc.
3000-7000 NA8 MAF 5.57 Nissan MAF 5.33
4500-7000 NA8 MAF 3.53 Nissan MAF 3.40
4500-7400 NA8 MAF 4.34 Nissan MAF 4.17

Small measurable improvements but throttle response seems better and intake noise is surprisingly a lot quieter. Only downside so far is a slight stumble when backing off from full throttle to cruise throttle - might be something to do with transition from open loop to closed loop.

[Fatty]

good stuff phil, very interesting. seems like a worthwhile experiment .

[Matty]

Matty, I could go forever with the direct empirical approach. I did remeasure the NA8 MAF - not easy but my best guess is a minimum throat of 1535mm2, which is still 28% less than the throttle body.

What I'm actually after is improvement in acceleration and throttle response. So the easiest way of knowing that I'm getting there is to log speed against time. 0.3 secs in a 0-100 would be very nice but 4500-7000 is where I most want the improvement for track work.

So I took the car out again today and re-did a series of 2nd gear pulls, on the flat, no wind, same road, CAI, fuel, tyres, pressures, temperature, load, fuel map, timing map, etc, etc.
3000-7000 NA8 MAF 5.57 Nissan MAF 5.33
4500-7000 NA8 MAF 3.53 Nissan MAF 3.40
4500-7400 NA8 MAF 4.34 Nissan MAF 4.17

Small measurable improvements but throttle response seems better and intake noise is surprisingly a lot quieter. Only downside so far is a slight stumble when backing off from full throttle to cruise throttle - might be something to do with transition from open loop to closed loop.

Good data, small but noticeable improvements. The trick now is determining whether the difference is due to better air flow, or the effects on fuel and ignition.

You can watch open-closed loop effects with a $15 AFR gauge.

Your best tool for seeing whether the air flow is actually improved is a manometer. (like I said, you're welcome to borrow mine)

I remember when I was stuffing around with an RX-7 AFM on my 1.6 - the top end power was better but the throttle response and midrange was so badly affected that it wasn't worthwhile

[Fatty]

matty, have you still got that rx7 afm?

[93_Clubman]

Phil, a good result for the effort & cost - tks for sharing.

James, if you recall since Matty first tried it, the best result with the RX7 AFM was to transplant the NA6 internals into it to over come at least the idle problem.

[Fatty]

yeah i remember that paul, ta. i'd still need to housing tho to try that tho eh. but even then, i probably wouldn't need to bother with a transplant, i could just use the dfa to acheive the same result?

[93_Clubman]

transplating the na6 afm internals into the rx7 afm body solved the flapper door spring tension problem, which from memory was causing an elevated idle. also might be benefits else where, but as you say dfa could possibly sort those.

[Matty]

No I don't have it any more... not that I recommend it.

The problem wasn't with idle per se. The first issue was that the standard spring is much stiffer, because the '7 unit is designed for a much higher peak air flow. With the original spring in it, it idles perfectly (the vane barely moves at idle on either meter). However the vane doesn't open enough once on the power, so it runs lean. To correct the midrange leanness, you can loosen off the vane so that it swings to the correct position under power. However this then doesn't provide the tension to make it close at idle, so it stalls or runs pig rich and stutters.

By swapping in the MX-5 spring, it runs a little better, but it is still not enough to be anywhere near as good as the standard meter. Perhaps the DFA could do some further correction, but I'm not sure. There were two problems still - one is that even with the original spring it would run lean under power - the taller design means that to flow the same amount of air it simply doesn't need to open as far - ideally you would need an even softer spring than the MX-5 one to compensate. The other is that it suffered from poor tip-in response - perhaps because the larger heavier vane is slower to move, and the ECU isn't calibrated for this appropriately, but that's a guess.

[Fatty]

yeah sounds like more trouble than it's worth then, thanks matty.

anyway... sorry for going off topic phil

[manga_blue]

yeah sounds like more trouble than it's worth then, thanks matty.

anyway... sorry for going off topic phil

I'd be totally amazed if anything stayed on topic for 2 pages on this forum There's nobody without ADHD on it .

How does the response curve go for an AFM? The reason I ask is that the NA8 and Nissan MAFs behaved as I expected, i.e. you feed in 12V and the output varies from 1.1V at idle up to 4.5V flat out. That NB MAF that I initially said didn't work was different: feed in 12V and 5V comes back out with zero airflow, put the vacuum cleaner on it and the voltage starts decreasing. I didn't give it much time or thought but ... might it work in an NA6? or was it just a aberrant MAF?

[Fatty]

yeah, that goes against my previous reading / understanding of how mafs work. as far as i knew, all mafs worked in the same direction, whereas what you're describing is more like an afm. maybe that maf has some sort of voltage inverter circuit in it

it could be an interesting science project to whack your aberrant maf on a na6 and see what happens. altho i'm pretty keen to stick with an afm, as i'm planning to fit some cams to the b6 and apparently it's easier to tune wilder cams with a afm system compared to a maf or map system... (i think this was mentioned in that thread that got you started on this whole idea in the first place)