jimmyblue

iTrader: (11)

Looked at a Delphi 85mm MAF on the bench,
varying the +5V reference supply and looking
at output frequency (dead air). Varying the
supply voltage +/-10% appears to make an
output frequency drift +/-20%. This I believe
is showing that the power chopper is working
with the raw unfiltered / unregulated +5V that
is presented to the MAF, the heat power to the
sense wires is fixed pulse width * frequency *
(V ** 2). Square law dependence on the PCM-
sourced reference supply value. This is head-
end regulated but who ever checks it?

Now, a person who cared to make their own
"+5V" off, say, IGN with an adjustable regulator
could therefore "bend" this MAF high or low
simply by tweaking the created power supply.
And there's no reason why you couldn't throw
the tach, TPS, or other voltages (or images,
buffering would be smart) into the mix by fairly
straightforward analog means.

The 5V supply/ref and GND are both straight wired
to PCM in the stock setup, presumably to eliminate
ground loop issues. Any cobbled-up regulator would
want to still reference the black wire as its ground.

Of course if you have tuning capability all of this
is relatively pointless, except that any deviation
from ideal in the +5V reference goes straight to
your MAF's match to the table, and pretty strongly -
if you want 1% MAF table fidelity, you'd best
have 0.5% reference supply accuracy (and no
internal error in the MAF electronics).

Just today's bit o' science, FWIW.

patSS/00

iTrader: (6)

Not sure which pin you're using, but the + supply to the MAF is 12V not 5V. But I think the MAF logic circuit (that generates the frequency output) also draws power from the PCM via the freq. output wire, since that wire is pulled high (+5) by the PCM but using a very low resistance. (The reason I know this is that I've got a home-made box with a microcontroller that acts as a kind of fancy MAF translator, but with more inputs than just MAF freq). Anyway putting a regulator on the MAF + supply might not make it any more accurate, because the +5 on the other wire would still be from the PCM.

jimmyblue

iTrader: (11)

Thanks... I saw the 5V pullup & missed the 12V supply.
I will have to see if 12V produces a more stable / less
supply sensitive operation. Do you know what the PCM
pullup resistor value is? I am using 1Kohm but this
has a kinda-slow risetime on the 'scope; not knowing,
I haven't wanted to push the load value down, and not
that keen on forcing the PCM pin to measure it either.

patSS/00

iTrader: (6)

I don't know the actual value - I know that 470 ohms works (that's what I use), whereas > 10K ohms doesn't work. Also if you measure resistance between the PCM pin (from MAF), and gnd, with the MAF disconnected, it's 1.7K. These values seems pretty strange to me - why would there be so much load on that pin inside the PCM? Anyway my very rough calculation based on V & R measurements I could make was that the pull-up must be really low, like 200 ohms, so I tried 470 which I had lying around and it worked.