Different bias resistor voltage supply calculations?

[katman]

Here's my first post, and my question!

I'm thinking that it would be easier to use MS in a 'piggy back' mode to monitor the sensor information (voltage) supplied to the existing ECU, rather than fit separate sensors for MS.

I've removed the bias resistors for the air and coolant temerature in MS, so those inputs are basically now only functioning voltmeters, with the existing ECU supplying the bias resistor, bias voltage and sensor.

I've measured the voltage applied to the sensors, and calculated the internal 'bias resistor' that the ecu is appearing to use.

Measured open circuit voltage is 4.6 volts. Air temp sensor measures 529 Ohms (out of circuit), and voltage across the sensor wit ECU connected is .893V, so calculated ECU resistance works out to 2184 ohms.

I can do similar with the water temp. sensor.

Figuring out the hot and cold temps is only a matter of setting up the temps, and measuring. THat shouldn't affect my question.


My question, is how the derived tables will be different. I'm assuming that all the EasyTherm calculations are based on 5v applied to the bias resistor.

I'm thinking that all that has to happen is that the DP entries would be multiplied by 0.92?

(That's from 5v/4.6v=0.92 correction factor.)

So...in the below table, calculated at 5v bias voltage,

; ADC Temp, degF
DB 210T ; 0 Sensor Failure - use 170
DB 255T ; 1 301.51
DB 255T ; 2 257.99
DB 255T ; 3 234.14
DB 255T ; 4 217.82

Would I want to multiply everything in the "DB" entry field by 0.92?

Or not? I'm thinking I don't quite have the concept right, as values that are closer to 0v will not change as much as values near the 255 ADC will.

Maybe I have to multiply the "position" of each value, so that it's 0.92 closer to the '0' ADC count location. THat would make more sense I think.

I'm thinking of it as compressing all the 255 ADC entries into only 234 spots, as anthing above 234 (4.6v) won't ever be seen .

I'd think that I'd also have to disregard any count over 234 (234 comes from =0.92 * 255), as the voltage as seen by MS will never get past 4.6 volts in my case. I'd treat 4.6v (ADC 234& up) as indicating failure.

Next, how would I get the new data into the MS?

Would it be as simple as editing the .inc files in notepad, after creating them, but before sending to the MS controller?

I seem to have read though, that if the .inc files are changed by had, recompiling is required. IF so, is that a part of EasyTherm, or is it now a much more complex problem?

Or even better, is the difference in Bias resistor voltage not enough to worry about?

(FYI: This is to go in an '83 Audi urq, running the MC1 motor and MAC 11 ecu)

Thanks for any thoughs!

[efahl]

My question, is how the derived tables will be different. I'm assuming that all the EasyTherm calculations are based on 5v applied to the bias resistor.

Frank,

Actually, ET doesn't care what voltage is applied, since the values are all differentials. In your case you need to actually measure Vref of your MS and use that in the calcs where you have assumed 5.00 v.

Would I want to multiply everything in the "DB" entry field by 0.92?

Yes, your logic is correct.

Next, how would I get the new data into the MS?

Now that's the hard part. If you have the P&E assembler installed, it's easy, just overwrite the existing therm- and matfactor files and re-compile...

Would it be as simple as editing the .inc files in notepad, after creating them, but before sending to the MS controller?

No, EasyTherm re-generates the .s19 file, encoding the inc files in the form that is required by the bootloader on the MS chip...

Or even better, is the difference in Bias resistor voltage not enough to worry about?

I'd say it's worth trying as the actual temperature values are not particularly important, it's just a scale error that you can tune around quite easily.

(FYI: This is to go in an '83 Audi urq, running the MC1 motor and MAC 11 ecu)

This will really bring that MC motor to life, ditching the CIS flapper almost doubles your HP right off the line.

Eric

[keithmac]

Best thing I did to mine was removing the "crap flap" .

[katman]

Ok. I'm getting what the Eric says. Makes sense to get pretty 'close' with easy therm, then tuna around the slight error.

Now to the best part: Looks like the Audi MAC 11 ECU uses a linear, positive temperature coefficient sensor for air temperature measurement.

If I plug in the values from the sensor specs, I don't suppose easytherm will sort that thing out, now will it?

I guess I'll *have to* install an additional air temp sensor..

[renns]

Frank,

If you get three data points spread nicely across the expected operating range of the sensor, EasyTherm will generate a precise table for you with any bias resistor you choose. To keep step size small across the operating range, pick a bias resistor equal to sensor resistance about mid-span.

Sorry, but the PTC type thermistor you've got is NOT supported by ET. ET assumes NTC thermistors, which have a logarithmic characteristic curve. The PTC sensors are generally quite linear in nature. You could hack through the theory in the 'How MS Measures Temp' post in this forum, and create your own custom .inc file to suit your PTC sensor. For most it'll be much faster to swap in a more typical NTC type.

Roger.

[katman]

Hello Roger:

I read your expalnation of how MS figures temp. It took a while, bit I finally figured it out. Makes sense.

It wouldn't be to hard to make a new table for a PTC sensor, but from what Eric mentions, the problem would be to compile the .inc table into the .s19 file.

That's the part that I don't quite understand...

As you surmise, it's probable easier to install a NTC air temp sensor. Good thing it's the air temp sensor, as that's way easier to install.

Thanks,

[renns]

Frank,

You'll need to generate two .inc files:

airdenfactor.inc, which is an air density correction factor based on air temp (actually raw ADC mat input). This file must be incorporated in the .s19 file, and downloaded to MS in order for the air density correction to function properly.

matfactor.inc, which is a file that correlates the raw ADC mat input to degrees F. This file is not needed by MS, but is used by MegaTune to allow laptop display of air temperature in suitable units.

BTW, if you need a hand with the config/setup steps, drop me a line via email. I'm only about 40 minutes drive away from you.

Roger.

[efahl]

Frank,

If I were you, I'd put in an NTC air temp sensor, it's just way easier all around. I put mine in the hole vacated by the cold start injector, wanting to see if heat soak would cause problems on my engine (it has not), but if I were doing it for someone else, I'd relocate it to a more conventional place, out in the post-intercooler/pre-throttle body part of the plumbing.

As for coolant sensors, that's actually easier than the air sensor. Look on the back of the head, down near the deck on the driver's side. You should see an allen head plug in a (from memory) M12x1.5 hole. If you pull that plug an screw in a Bosch 0 280 150 037, then you can have a dedicated coolant sensor, too. Dig around here for more:

http://www.not2fast.com/audi/5ktqa/ms-2004/
http://www.not2fast.com/audi/maxq/max2002/

Eric