Can I use a MAP sensor module to measure barometric pressure?
Kevin Mouton
sensor works by varying the voltage out on the signal circuit. I do know that MAP
sensors in general are notoriously inaccurate. I don't think the resolution of the
signal would be narrow enough for your proposal, either. The map can change the
signal voltage across a five volt range with a varying absolute pressure from 30
hg to 0 hg, but if you are monitoring barometric pressure you would want to be
able to resolve very minute changes in pressure. As far as I know, the typical MAP
sensor circuit in automotive computers only needs a resolution of about 1.0 hg
change in pressure. For a barometer you would need a resolution of 0.01 I really
don't think you will get satisfactory results using an automotive MAP sensor for
that application. I do seem to recall seeing a very inexpensive digital barometric
pressure gauge at a gift shop one time. Perhaps that sort of thing might lend
itself to your needs.
Good luck
Kev
eric...@my-dejanews.com wrote:
> As inexpensively as possible I need to measure barometric pressure for use by
> a computer. There are lots of expensive, instrument grade sensors out there,
> but I will gladly trade resolution, accuracy, linearity etc. for as
> inexpensive as possible that will give rough values (within a few percent).
>
> After rummaging around the Internet and this newsgroup I have learned that
> there is a common automotive part that may do the job. Please bear with me as
> I know very little about automobiles (but quite a bit about electronics and
> computers :-)
>
> The facts (I am uncertain about a points marked ?):
>
> - There is a "MAP" (Manifold Absolute Pressure) module in most (all?) new cars
> - MAP output is proportional to barometric pressure minus manifold pressure
> (vacuum)
> - Barometric pressure is measured by the MAP sensor at Key On Engine Off or
> wide open throttle
> - MAP is a 3 terminal device: signal gnd, +5 and output
> - Ford MAP modules output a 0 to +5 volt square wave that varies frequency
> about 92 to 162 Hz
> - Some (all?) other MAP modules output a DC between about 1 and +4.5 voltage
>
> Conclusion:
>
> If I get a non Ford MAP module, hook the right pins to +5 and ground then the
> output will be a voltage somewhere between about +1 and +4.5 that will vary
> from low to high barometric pressure, tracking my barometer on the wall
> (assuming my altitude stays fixed). I can easily convert the voltage to
> numbers for the computer with an Analog to Digital converter and handle
> offset / calibration via simple arithmetic in my software.
>
> Questions:
>
> - Is my conclusion correct? - Is there any particular MAP module that anyone
> would recommend? It has to be as cheap as possible at a wrecker or parts
> store, be the voltage output type and I also need to be able to easily get
> the "pin out" diagram of the module to connect it up. - Any other comments?
>
> Thanks very much in advance for any assistance.
>
> Eric Pierce
> eric....@ZZZaltavista.net (non spammers, remove ZZZ)
>
> -----------== Posted via Deja News, The Discussion Network ==----------
> http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own
--
To reply, replace "NOSPAM" in return address with "eatel"
http://www.eatel.net/~kevinm/homepage.htm
************************************************
Kevin Mouton - Automotive Technology Instructor
************************************************
"If women don't find you handsome they
should at least find you handy!"
Red Green of Possum Lodge
************************************************
eric...@my-dejanews.com
Mike Brown
Eric,
This is what I measured from a couple GM MAP modules. It is pretty
coarse and I don't think it will work for what you want. Also MAP
sensors measure the differential presure between outside air pressure
and what is in the manifold so if you took one to 1000 feet and both
ports where exposed to atmosphere you would still read 0". Hope this
helps.
mike
Vacumn Voltage
(in)
0 4.61
1 4.45
2 4.29
3 4.13
4 3.96
5 3.80
6 3.63
7 3.475
8 3.30
9 3.12
10 2.971
11 2.85
12 2.656
13 2.444
14 2.301
15 2.123
16 1.890
17 1.713
18 1.574
19 1.368
20 1.201
21 1.000
22 .825
23 .637
24 .460
25 .285
26 .100
27 .026
28 .026
29 .026
none...@execpc.com
> This is what I measured from a couple GM MAP modules. It is pretty
> coarse and I don't think it will work for what you want. Also MAP
> sensors measure the differential presure between outside air pressure
> and what is in the manifold so if you took one to 1000 feet and both
> ports where exposed to atmosphere you would still read 0". Hope this
> helps.
You would still read 0" correct, but at the higher altitude the 0" voltage
output would read lower. It's all relative to the 5 volt input voltage.
In other words, the 4.61 volts you measured today would read different at
a different barometer reading.
Just wanted to clarify.....
--
Neil Nelson
If you don't believe in something.....
You'll fall for anything....
Member of i-ATN....international Automotive Technicians Network
summer toys; 70 AAR 'Cuda #s match original not restored
64 Valiant 340 4bbl
winter toys; 95 Dakota SLT 4X4
88 Toyota Camry AWD
Pete Barnes
sensitive. Try this web site http://www.sensym.com/ they make all ranges
of sensors.
In article <73kvln$vuh$1...@nnrp1.dejanews.com>, eric...@my-dejanews.com
writes
--
Pete Barnes
Mike Brown
I don't believe so. When you hook up two vacumn source to it, one to
the port into the manifold and the other to the outside vent, and pulled
5" thru the manifold then pulled 5" at the outside vent then voltage
will read 4.61 volts. If you drove your car to a 5000 foot elevation
and measured the output voltage (motor off) it would still read 4.6
volts. I have been working on dis-assembling and commenting the code
that runs the ECU in my hot rod ('34 ford runing the Webber/Edelbrock
EFI) and when it sees 4.6 volts it computes and displays 0". All we are
trying to do here is to figure the pressure difference between the
intake manifold and outside pressure to determine how much air is
entering the motor.
mike
none...@execpc.com
> Neil,
> I don't believe so. When you hook up two vacumn source to it, one to
> the port into the manifold and the other to the outside vent, and pulled
> 5" thru the manifold then pulled 5" at the outside vent then voltage
> will read 4.61 volts. If you drove your car to a 5000 foot elevation
> and measured the output voltage (motor off) it would still read 4.6
> volts. I have been working on dis-assembling and commenting the code
> that runs the ECU in my hot rod ('34 ford runing the Webber/Edelbrock
> EFI) and when it sees 4.6 volts it computes and displays 0". All we are
> trying to do here is to figure the pressure difference between the
> intake manifold and outside pressure to determine how much air is
> entering the motor.
>
> mike
Hi Mike,
Going to skip the part about the vacuum being applied to manifold vent and
the outside vent as I really don't understand the application of it here.
The thing to remember is how the algorithms for the MAP circuit treat the
information. When the ECU first powers up, that is the first second or so
that the vehicles key is switched on, the MAP sensor reports a voltage
reading to the ECU...that voltage reading could be 4.61 as you measured,
it could be 4.0, it could be 3.5....they all represent possible barometric
readings, information the ECU look-up tables need (desperately) in order
to deliver the correct mixture of fuel based on air density. You need to
fully understand that it is not just the fixed value of 4.61 volts that
represents 0" of vacuum....any map sensor voltage value sent to the
computer at initial key on power up WILL represent 0" of vacuum.
Below is the chart* GM Training publishes for MAP output vs. Altitude...
Altitude in Ft Voltage range
below 1000 3.8-5.5
1000-2000 3.6-5.3
2000-3000 3.5-5.1
3000-4000 3.3-5.0
4000-5000 3.2-4.8
5000-6000 3.0-4.6
6000-7000 2.9-4.5
7000-8000 2.8-4.3
8000-9000 2.6-4.2
9000-10000 2.5-4.0
*= GM fuel Injection training manual #16009.10-1
The fact that the MAP sensor outputs different voltages at different
altitudes (and barometric pressures) is what is responsible for the engine
managements ability for altitude compensation.
At 5000 ft altitude, the MAP ~might~ output the 4.6 volts as you
mentioned, depending on the slew of the sensor, it certainly does fall
close to the acceptable range....but if it did and you took it back down
closer to sea level, the output would be significantly higher (assuming a
properly functioning sensor).
Just about the time you make some progress on your 34 Ford Hot Rod.....
Along comes that Nelson guy......8^)
HTH
Mike Brown
I understand what you are saying. It would make sense that the output
of the sensor would change as altitude increased (air density
decreased). The air is less dense and the mixture would need to be
leaned to keep the AFR correct. I was thinking in terms of the IC that
are used to measure pressure. They are usually two port devices but if
you blocked off the reference port with 14.7 lbs/sq" then the device
would measure an absolute preassure relative to sea level. You are
correct as that is what is happening here. If I understand what you are
saying about how the algorithm works, it sound like the ECU takes the
first reading and uses it to normalize all the other readings so it
knows where 0" is. If it did that then the what would be the point of
having absolute pressure avaiable to it. What it is doing is taking the
MAP reading correcting it for MAT, further correcting for battery
voltage and coolant temp(warm up stuff), factoring in rpm dependent
volumetric effeciency of the motor to come up with the pulse width
required. I went back and looked at the code (remember I have a
commented source code) and there is nothing significant about the first
reading taken. If there was then if you were pushing a car with the
clutch out and ignition off and throttle plates closed you would be
generating a fair amount of vacumn in the manifold. If you then turned
on the key the ECU would have taken the first reading as a baseline and
the car would have run like hell because you confused it with a lower
pressure as the first reading. This is not what happens.
Mike
none...@execpc.com
> Neil,
> I understand what you are saying. It would make sense that the output
> of the sensor would change as altitude increased (air density
> decreased). The air is less dense and the mixture would need to be
> leaned to keep the AFR correct. I was thinking in terms of the IC that
> are used to measure pressure. They are usually two port devices but if
> you blocked off the reference port with 14.7 lbs/sq" then the device
> would measure an absolute preassure relative to sea level. You are
> correct as that is what is happening here. If I understand what you are
> saying about how the algorithm works, it sound like the ECU takes the
> first reading and uses it to normalize all the other readings so it
> knows where 0" is. If it did that then the what would be the point of
> having absolute pressure avaiable to it. What it is doing is taking the
> MAP reading correcting it for MAT, further correcting for battery
> voltage and coolant temp(warm up stuff), factoring in rpm dependent
> volumetric effeciency of the motor to come up with the pulse width
> required. I went back and looked at the code (remember I have a
> commented source code) and there is nothing significant about the first
> reading taken. If there was then if you were pushing a car with the
> clutch out and ignition off and throttle plates closed you would be
> generating a fair amount of vacumn in the manifold. If you then turned
> on the key the ECU would have taken the first reading as a baseline and
> the car would have run like hell because you confused it with a lower
> pressure as the first reading. This is not what happens.
>
> Mike
Hi Mike,
I don't know from first hand experience what would happen in the push
start scenerio you describe, whether the false baro reading would effect
driveability would largely depend on the internal programing of the
particular ECU in question. Yours is based on an aftermarket system
correct??
GM and Chrysler put high authority on both pressure signals (Baro & MAP)
for their speed density systems.
My first experience with a baro sensor failure was back in 80 or
81...Chevy something or other, car ran like crap, wound up calling the
closest airport in Milwaukee that had a manned control tower to verify the
altitude at the small grass airstrip about 3 blocks from the
dealership....the baro voltage wasn't off by much, but it definately was
out of the range of the published GM chart for the time, maybe 1/10 of a
volt, but what a difference a correct reading sensor made!! This was
followed by about 8 years of trying to get the rustproof kid to understand
why he can't spray gobs of Rusty Jones(tm) all over the little black boxes
under the hoods of the new cars.
E. William Lawrence III
GM MAP and BARO sensors are almost identical except that the old BARO sensor
had a filter on the nipple to keep dirt away.
Both sides of the diaphragm are not exposed to atmosphere.
The upper side is exposed to a small chamber which contains a partial vacuum.
The action of the pressure on the other side causes the wein bridge to deform and
varies the return voltage to the computer.
In fact modern vehicles use the key on reading of the MAP as a BARO reading.
If a BARO sensor was exposed to atmosphere on both sides how would you get
a reading? The wein bridge would remain flat if the pressure was equal on both
sides.
Bill
By the way the answer to the original question is yes, obviously.
none...@execpc.com wrote:
E. William Lawrence III
The measurement taken by a MAP sensor, or a BARO sensor is not manifold vacuum
versus ambient air. It is either atmospheric or manifold pressure against a
calibrated
value maintained inside the sensor in a small sealed vacuum chamber with a flexible
element that changes resistance with shape to indicate the pressure acting against
it, period.
The MAP sensor on modern cars is used as a BARO sensor and will function in
this capacity.
Bill
Mike Brown wrote:
> Neil,
> I don't believe so. When you hook up two vacumn source to it, one to
> the port into the manifold and the other to the outside vent, and pulled
> 5" thru the manifold then pulled 5" at the outside vent then voltage
> will read 4.61 volts. If you drove your car to a 5000 foot elevation
> and measured the output voltage (motor off) it would still read 4.6
> volts. I have been working on dis-assembling and commenting the code
> that runs the ECU in my hot rod ('34 ford runing the Webber/Edelbrock
> EFI) and when it sees 4.6 volts it computes and displays 0". All we are
> trying to do here is to figure the pressure difference between the
> intake manifold and outside pressure to determine how much air is
> entering the motor.
>
> mike
>
Bruce Hamilton
>As inexpensively as possible I need to measure barometric pressure for use by
>a computer. There are lots of expensive, instrument grade sensors out there,
>but I will gladly trade resolution, accuracy, linearity etc. for as
>inexpensive as possible that will give rough values (within a few percent).
There used to be Philips and/or Sensym pressure sensors that were about
$3/each ( a decade ago! ) that were linear and came in a range of options.
The ones I have used were the 0-30 psi absolute ones about the size of a
fingertip ( below the fingernail ).
They came in a range of models, and had outputs that were easily
calibrated. The Sensym manual had circuits etc. I believe they may
have been used as cheap pressure controls and cutoffs in appliances.
They have the advantage that being absolute, atmospheric pressure
is midscale, and they were very stable. Rather than mess about with
automotive sensors, I'd go with the cheap pressure sensors available
in larger electronics/hobbyist shops. Several electronics magazines
have offered simple barometer kits that would name suppliers - I've
no ide if SenSym still exist, but they had a huge range of pressure
sensors - from cheap to very expensive.
Bruce Hamilton
-Bodnar,B.L.
E. William Lawrence III <lawren...@uclink4.berkeley.edu> wrote:
>Again,
>The measurement taken by a MAP sensor, or a BARO sensor is not manifold vacuum
>versus ambient air. It is either atmospheric or manifold pressure against a
>calibrated
>value maintained inside the sensor in a small sealed vacuum chamber with a flexible
>
>element that changes resistance with shape to indicate the pressure acting against
>it, period.
>The MAP sensor on modern cars is used as a BARO sensor and will function in
>this capacity.
>Bill
>
[ some text deleted for brevity]
Let me just add a bit to the above information...
One of the electronics trade magazines (I think it was EDN) had an article on
MAP sensors and their construction about a year ago. It turns out that each
MAP sensor's resistive membrane is laser-trimmed to provide (1) linearity in
pressure vs. output signal and (2) high accuracy. Prototype MAP sensors built
by Hewlett-Packard (?) have an analog-to-digital converter built into them.
The output of the ADC feeds a PROM in the MAP sensor which holds a signal vs.
pressure correction table. The output of the PROM goes to an
digital-to-analog converter which then feeds the requisite buffers (or, in the
case of Ford, variable frequency oscillator). The table is calculated for
each sensor during its construction. The article stated that much of the cost
of a MAP sensor goes into the customized trimming of the resistive membrane
and that the downloadable table approach would substantially reduce the
manufacturing cost.
Taking into consideration the manufacturing processes required to produce a
MAP sensor, I'd expect very good accuracy in measuring pressure. Indeed, when
one looks at pressure vs. output signal data (I have data for Ford's MAP
sensor and used to have it for Chrysler stuff), one also sees *high precision*
-- typically, three sig figs.
Comment pertaining to the above table obtained by Mike: service-grade
vacuum gauges are not accurate enough for relative pressure measurements.
Furthermore, the accuracy varies with pressure changes. I've run experiments
(strictly to satisfy my curiosity) where I had a vacuum pump feed a known good
MAP sensor and two different vacuum (service grade) gauges. All three
disagreed at low and high pressures and had sufficient service-grade matching
at moderate pressures (around 18" Hg vacuum).
For the person who wanted to use a MAP sensor to measure barometric pressure:
I'd use a Ford MAP sensor and measure interpulse time. This time can then be
used as an index to a pressure lookup table (send me e-mail if you want the
frequency vs. pressure data). This way, you reduce the parts count, simplify
the design, and have programmable control over the measurement's precision
(you're not constrained by the precision of the ADC and you don't have to
worry about signal-to-noise ratio problems at low pressures).
Cordially,
Bohdan Bodnar
bbo...@lucent.com
[blanks added to keep silly posting software happy...]
Don Stauffer
>
> Taking into consideration the manufacturing processes required to produce a
> MAP sensor, I'd expect very good accuracy in measuring pressure. Indeed, when
> one looks at pressure vs. output signal data (I have data for Ford's MAP
> sensor and used to have it for Chrysler stuff), one also sees *high precision*
> -- typically, three sig figs.
>
> Cordially,
>
> Bohdan Bodnar
Having an accurate and precise sensor is not all it takes to produce
measurement of something like barometric pressure. One must read out
the sensor with some electronic system. All electronic systems have
some noise level. If the noise level in the system is greater than the
output signal change of the sensor over the range you are interested in
observing, you won't get very good results. Also, in addition to the
accuracy and resolution of the sensor, you need to know ITS noise
output. Now, some definitions of precision include noise, others
don't. In looking at sensor specs you need to know whether noise was
factored into the definition of spec resolution/precision.
--
Don Stauffer in Minneapolis
home web site- http://home1.gte.net/stauffer/
home email- stau...@gte.net
work email- stau...@htc.honeywell.com