john banks

The three pin MAP sensor (Denso) on the inside of the driver's side wing - what sort of output does it give? If it is a voltage that is proportional to manifold pressure could this be a means of making an electronic boost gauge - ie attach a voltmeter? The port under the steering wheel into which the select monitor goes has only seven connections. Is data sent as encoded signals down here to give error codes/revs/MAP etc or do individual pins supply this info? I expect RPM would be pulsatile.

StephenDone

Hi John,

The under dash connector is a serial port... but not as we know it. 4800 baud, 12v/0v on pin 7 - for 99/00MY anyway.

What model year is your car ?

RPM output is not available on this port, though you can splice into the loom. The RPM output is a pulse waveform. Rate of pulses is proportional to revs.

I'm not sure of the direct MAP output yet, but I could check it with a DSO if you like. Likewise, if you find out, let me know.

You sound techie. Perhaps you'd like to join forces...

Please see my thread here:

john banks

It looks like the middle pin - pin 2 of the MAP sensor is the business part. I couldn't get easy access to the actual connection with it plugged in, and the socket seems to be a well sealed unit. Not too keen on splicing the wire unless I know it is a goer. The other two are presumably supply and ground, and I know these things usually give a voltage proportional to boost, but I'm not sure if there is oscillation involved which would need some decoding to be useful. Certainly on the select monitor (which is only plugged into the diagnostic port) you get RPM, and RPM and Speed pulses are easy to find in the wiring loom going to the speedo - I investigated as I was considering a speed-volume control headunit at one stage. The Possum Link manual ECU wiring diagram is my only source of info so far, but all I have is my printed copy and a dead link to it (printed from the dead link a while back).

Presumably the easiest way to try the "serial" port would be with a lap top. Does RS232 run 12V or 5V I can't remember?

I think for this sort of task we really need detailed info on the port and the pinouts and data format/commands. I didn't realise the port was so sophisticated, and thought it might just present some handy voltages/frequencies on individual pins just ripe for logging on a laptop through an analogue to digital converter.

MY00 PPP is my car. And part of the reason is a capability of possibly logging boost curves so I'm getting the optimum, as well as an insane fascination with these things. I might be easier buying a boost gauge and watching it carefully whilst someone else drives.

Anyone know of any technical manuals on this available to the public?

john banks

Steve - just looked through your website - it is blimmin brilliant. Packed full of info - should be one of the Scooby classics it really is that good. Your technical competence re the port is way beyond what I am capable of, but I would love to discuss further with you your ideas/progress.

StephenDone

Hi John,

Sorry, I could have made my mail a bit clearer. Maybe I should have read it.

What I meant to say was that the physical RPM voltage/waveform is not presented on the OBD connector. You _can_ access RPM via a select monitor attached to this connector, though all data is sent and received on pin 7.

In case you are interested:
Pins 4&5 are Ground
12 & 13 are more subaru specific pins.
16 is Vbat.

Sadly voltage levels aren't compatible with RS-232, though you can make an adaptor.

If anyone out there knows how to send 4800 baud from a PC, 8 bits with no start, stop or parity bits, then I'd like to buy them a drink.

Mail me off line if you like, John. Steve@ukmail.org.

Cheers

Steve

P.S. Thanks for the compliments on the site.

john banks

I understand now. None of the options in control panel I can find allow no stop bits - is that the problem or am I simplifying it? Someone somewhere must have written a simple terminal program that allows full control over these properties even if it runs under DOS?

Bob Rawle

With regard to the map sensor the output is linear with pressure/vac from minus 1 to plus 1.8 bar, supply voltage is circa 4.98, output in the range of 0.8(8) to 4.8(4) ish.

Atmospheric is 2.24 cvolts dependant on the day

Hope that helps

dowser

Steve - nice project

I investigated something similar....but gave up - too much to learn, too little time!

These *may* be useful;

Sniffing Software;

john banks

Thanks Bob - looks like my starter project is a goer then. Just need a suitable gauge.
What do the figures after the brackets mean Bob or anyone? 0.8 to 4.8V bit I understand.
"output in the range of 0.8(8) to 4.8(4) ish"

Thanks for the LINK link, and as ever Richard you are a mine of info. This will be an interesting project.

StephenDone

Hi John,

Your missing Link link is:

john banks

Oh I am being stupid. I think Bob mean 0.88 to 4.84V - I thought the figure in brackets might be a reference to something else, but perhaps it is just to say approximate. Correct me if wrong guys, and is it AC? Guess I'll find out when I put the meter on!

Bob Rawle

You're right, the figs in brackets were the second decimal place, will probably vary a bit sensor to sensor but should be in that ball park

Bob

john banks

Thanks again. Pin 2 is indeed the business and the figures are pretty close. At rest just under 2.3V, peak and held boost appear to be the same in gears 2-5 at about 1.25bar (=4.1V). I think the mechanical voltmeter I am using is pretty quick to move when you put a voltage across it.

john banks

Couldn't find a decent connector so I used a dreaded snap lock after trimming just a tiny bit of the outer casing around the three wires. If the wire fails at least it is close to the sensor so I could put a new piece in - I think that's unlikely as the insulation is just nicked in a tiny area as the wire is quite chunky compared with the snap lock. It won't be a permanent arrangement, but at the moment the wire goes in through a small grommet near the clutch pedal mount. The voltages Bob gave correlate with DC measurements I made. If I select AC it goes up to an indicated 6.6V, so I'm using DC for my calculations.

I wish I had a digi scope, then I could plot the boost curve, freeze the screen and see run a pointer along it afterwards. Maplin have a Velleman HPS5 which takes my fancy, not just for this but for other fiddling - similar price to a decent voltmeter.

I'm also thinking at some point about getting into doing some PIC stuff, or maybe using some of these BASIC stamps.

StephenDone

How did you get at the signal, John ?

Did you get some connectors to intercept the sensor cabling or chop into the loom ? If so, where was easiest ?

Cheers

Steve

StephenDone

Hi John,

If you want to come over to Maidenhead, we can use my 2 channel digital storage scope, no problem. You can then go away with a log file of what you want.

Try before you buy !

I nearly bought the Maplin jobby. But got one of these...

StephenDone

Hi John,

What did you find the supply voltage to be for the MAP sensor ? Was it 5v or 12v ?

Cheers

Steve

Bob Rawle

Supply volts typically 4.98 (ish)

StephenDone

Thanks Bob.

That's a nice regulated supply from the ECU then ? I need to steal some clean power for an op-amp circuit.

Cheers

Steve

john banks

Didn't actually check the supply voltage. I think signal correlated well with a mechanical boost gauge once I took errors into account.

Unfortunately I'm in Scotland!

StephenDone

Scotland !

Oh. Right then. It would cost you less to buy the scope than drive down to me.

Have a look at the PicoTech ADC-212. They have several examples on their site for automotive applications.

Cheers

Steve