PDA

View Full Version : MAP sensor?



Confused
28-11-2011, 03:27 PM
Is there any chance it's possible, or any thoughts of doing so if it is, of the standard ECU being modified to work using a MAP sensor, rather than the MAF?

Wodjno
28-11-2011, 03:37 PM
Not sure if with all the latest revelations that Kenneth and Co are finding possible the the standard ecu, whether it's possible or not ?
The only other way to keep the OE ECU and go MAP, is with a piggyback ECU.

Confused
28-11-2011, 03:40 PM
I know that prior to the ability to flash the ECUs, the only way was with a piggy-back ECU. Now that there's the ability to flash and do some funky stuff with it, I thought I'd pose the question as a discussion point :)

Wodjno
28-11-2011, 03:41 PM
I know that prior to the ability to flash the ECUs, the only way was with a piggy-back ECU. Now that there's the ability to flash and do some funky stuff with it, I thought I'd pose the question as a discussion point :)

Tis certainly worth a pose /yes

Davezj
28-11-2011, 06:25 PM
I have been thinking about this recently.
And I don't see why you can't use a voltage to frequency converter in line between the MAP sensor and the factory ECU and not change the ecu at all.
You would just have to match the boost pressure to frequency.
MAP output 0 - 5V if remember rightly.
And the MAF has something like 0 - 5Khz output.
I am not sure it is a linear relationship though.

If some can tell me what the actual output of the maf will be.

Nick Mann
28-11-2011, 07:19 PM
The MAF will not give figures with a simple proportional relationship with boost. MAF is measuring flow, so when you go to MAP you need density, not pressure, and also engine air throughput. 6000 revs, cold air at .5 bar is a LOT more than 2500 revs .5 bar, hot air.

The best way is potentially what the MAP ECU guys came up with - log the MAF output against MAP sensor combined with revs and air temp. Then you know how to program something to output the right figures for the MAF replacement.

Also, as I understand it, the ECU is fuelling for flow at the MAF, so compensates for a small lag in the system as the pressure builds or lessens, whereas a MAP sensor sits at the manifold, so needs no compensation. However, the ECU will carry on compensating the fake MAF output which can lead to rich/lean spikes in extreme changes in air flow.

foxdie
28-11-2011, 08:24 PM
Just to add to this, our ECUs do actually have a 0-5V input spare, it's reserved for a secondary o2 sensor on other cars but on ours it's unused and the ECU can see the value (you can log it in EvoScan)..

I've been doing some work with Kenneths ECU Mods V2.0 ROM (http://www.clubvr4.com/forum/showthread.php?61071-Kenneth-s-ECU-Mods-V2.0) today, although it doesn't officially support MAP sensors to determine load, you should in theory be able to select one or more of the eight maps based on the value of that input.

Strictly as an example (voltages won't necessarily match up, each MAP sensor is different), if the perceived manifold pressure was over 0.5 bar (secondary o2 line goes above 1.5 volts as an example), a different set of ignition / fuelling tables (aka map) can be activated, if the pressure then goes above 1.0 bar (2 volts) then another map can be activated and so on.

Kenneth
28-11-2011, 09:23 PM
It is in the pipeline. Can't give you an estimate on time frame sorry, I have a fairly busy few months ahead.


Is there any chance it's possible, or any thoughts of doing so if it is, of the standard ECU being modified to work using a MAP sensor, rather than the MAF?

Confused
28-11-2011, 10:22 PM
That's great to know that someone's at least thinking about it :)

Means I might be able to somewhat simplify the pipework for my Anglia if I don't have to think about suitable location for a good filter and the standard MAF :)

scott.mohekey
28-11-2011, 10:30 PM
Means I might be able to somewhat simplify the pipework for my Anglia if I don't have to think about suitable location for a good filter and the standard MAF :)

Have it sticking out of the bonnet! j/k

Davezj
28-11-2011, 10:54 PM
The MAF will not give figures with a simple proportional relationship with boost. MAF is measuring flow, so when you go to MAP you need density, not pressure, and also engine air throughput. 6000 revs, cold air at .5 bar is a LOT more than 2500 revs .5 bar, hot air.

The best way is potentially what the MAP ECU guys came up with - log the MAF output against MAP sensor combined with revs and air temp. Then you know how to program something to output the right figures for the MAF replacement.

Also, as I understand it, the ECU is fuelling for flow at the MAF, so compensates for a small lag in the system as the pressure builds or lessens, whereas a MAP sensor sits at the manifold, so needs no compensation. However, the ECU will carry on compensating the fake MAF output which can lead to rich/lean spikes in extreme changes in air flow.

Cheers nick, I thought I was missing something, or someone would have done this already.

Reconstructor
29-11-2011, 06:42 PM
The best way is potentially what the MAP ECU guys came up with - log the MAF output against MAP sensor combined with revs and air temp. Then you know how to program something to output the right figures for the MAF replacement.

Correct. However, air temp may not be needed. Since I'm planning to split the intake and to put separate air filter to each turbo, such piggyback is a must for me. It is not hard to develop such a device, I think the hardest task will be the logger.

Adam.Findlay
30-11-2011, 12:38 AM
pretty sure you still need the air temp. no matter how well you log the map sensor against the maf without air temperature there isnt any way to compensate for air density which changes with temperature. colder day = denser air = more air for the same pressure reading on the MAP. both the MAP and IAT go inside the plenum I think so you can still have your individual pod filters.

Reconstructor
30-11-2011, 11:27 AM
That is correct logic. However, there is already temp sensor, paired with MAF, so I assume that the MAF itself can't handle the air density. Anyway, when I start developing, it will be clear. :)

foxdie
30-11-2011, 12:04 PM
Here's an idea;
MAP sensor doing its thing taking the air pressure from the manifold outputting a 0-5V analog signal
Digital air temp sensor (such as a DS18B20 (http://proto-pic.co.uk/one-wire-digital-temperature-sensor-ds18b20/), good for -55C to 125C +/-0.5C) located inside the elbow joint connecting to the throttle body
Amtel ATtiny85 (http://proto-pic.co.uk/avr-8-pin-20mhz-8k-4a-d-attiny85/) running the Arduino-Tiny (http://code.google.com/p/arduino-tiny/) codebase, it can then be easily programmed to read the two above sensors (it has 6 IO lines, 4 of which are analog with 10-bit ADC resolution), calculate everything that needs to be calculated and output a signal identical to the MAF signal
Cheap to implement, compact and tuneable later down the line. Chances are it'll probably be more accurate than a 14 year old MAF sensor anyway :)

foxdie
30-11-2011, 01:40 PM
Just to add a little more to this in the hope it'll help someone, the following image shows the Air HZ readings between a slow moving start, 1st and 2nd gear pulls (each hitting about 1.4 bar boost), followed by no load (engine running on overrun so negative manifold air pressure)...

49964

In terms of the Air HZ signal, I've observed the minimum it reads is 25Hz and the maximum is 1475Hz. As you can see, the signal does seem to indeed lag behind actual manifold pressure (using 2-byte load as an loose estimate for MAP) for increases in pressure, but seems to be pretty on the ball for loss in pressure.

Reconstructor
30-11-2011, 02:00 PM
ATtiny85

Tiny is not enough, since too little memory. It will need a map with many cells. :) 2 or 3 dimensional, depending on will or not air temp is needed. The Mega series with at least 4K RAM are more suitable.

Reconstructor
30-11-2011, 02:01 PM
Air HZ readings

So, it works on below 1KHz range? :)