Freescan vs MAP values

[Jonathan]

From

http://www.jonathanedwards.pwp.blueyonder....depth/index.htm

[Bibs]

Very good Jono, it's good to dispel the myth and have a straight piece of information to deal with. I'll add this to the Freescan FAQ

[Jonathan]

No worries - the graph is not exact but it should be within 0.05 bar accuracy.

If I can get a much more accurate compressor I'll redo the graph - it's also worth noting this is MAP boost i suppose as opposed to turbo boost which could be higher as you get losses / leaks etc.

I'll update the info later so there is no confusion but the graph shows pressure at the inlet manifold so I guess it should be refered to as usable boost.

[Vulcan Grey]

I think the thing to remember is that the fueling and turbo, that are supplied with the car, cannot really supply enough fuel or boost past 1.13 bar anyways. Some sort of upgrade is needed anyways.

Most people tend to increase the size of the secondary injectors... Which does not sacrifice street (low boost) drive ability, but when the ECU goes into open Loop the fuel injectors are supplying more fuel than was assumed by the tables in the ECU anyways... Therefore you are still able to take advantage of increased boost (up to cutoff)

Some of Marcus' chips allow up to 1.25 bar correct? So how do they do this if the MAP would be over the 5V limit at 1.13?

If you've noticed the Baro pressure reads low... Yours probably reads .85 at sea level. Mine reads .65bar at 5300 ft, and even lower when I go up to the mountains (~10,000ft). So a 2.24bar max in freescan with a .85bar baro allows 1.39bar headroom roughly for boost.

Lotus actually built in that offset to get away with a 2Bar sensor. Btw a 3Bar sensor will not work without a custom memcal program with new tables... ask me how I know;)

In fact my SE with a stock chip would boost up to 2.24 bar (seen in Freescan) often at my 5300ft altitude. With my .65bar baro that yields a nice 1.59bar of boost minus Lotus' offset from your chart...

Notice I could probably push the limits of my stock SE turbo pretty easily... Now with a RED Race or Blue Race chip from Marcus...

No myth btw, just ask Marcus.

[AJay]

Jon

Slightly off topic, but you couldn't do a boost vs timing advance graph could you?

Andy

[Vulcan Grey]

Those already exist.

http://users.ox.ac.uk/~ohare/ecu_tables.htm

Edited March 17, 2009 by Vulcan Grey

[Jonathan]

To address some points.

The BARO reading on my motor is ~0.85 - the freescan MAP at atmospheric pressure (1011millibars) was 0.91.

When you xref that on my graph it works out spot on.

The myth comment was with regards to the MAP sensor not being linear, and the old chesnut that if you take the freescan MAP reading and -1 off of the value you get boost pressure in BAR - which you dont.

On the 1.25 bar issue - I'd need to know how the ECU works it out to really make a comment. I assume it uses the BARO reading as a datum (as you say Travis, the higher altitude you go the lower that datum will be) in which case I can see how the MAP sensor can allow large amounts of boost.

However with 99% of people that live ~ 500ft above sea level or where baromic pressure is usually around 1 bar it's not going to make a huge difference for that %of people. That offset (0.91-0.85 = 0.06) also only equates to 0.05 bar or 0.73 psi.

For most users a CODE 33 will flag (MAP over voltage) and car in open loop mode guessing at the figures.

Without having more access to chipped cars I can't compare more results.

Of course if anyone wants to lend me some to do some tests on......

[AJay]

Those already exist.

Er Thanks just what i need, should have looked first.......

Andy

[eriks4]

Hi, I was trying to figure out the relation between the ECU output and the true baro and MAP values and found this thread. Unfortunately the figure referred to has disappeared and I cannot locate the FAQ.

I have also been trying to figure out the correct conversion.

I have used three different sources to try to figure it out. The first one were the various GM ALDL protocols that are similar to the one used in the Esprit.

Here, the information is VOLTS = N*5/256, and kpa = (N + 28.06)/2.71.

This is also the equation used in the GM P-4 documentation.

It clearly does not give the correct answer.

The second source I used was the Freescan sourcecode:

m_pSupervisor->m_fBaro = (((float)buffer[24] - (float)130.0)/ (float)100) + (float) 1.0;

hence, Baro=1+(ECU_value-130)/100 (in some documents it still says -128 instead of -130, but only yields a small difference.

And the same equation is used by Freescan for MAP.

This yields a value of about 0.85 when the true pressure is about 1 bar. I know that there has been a previous discussion on this forum regarding a fudge factor but the graph in this thread that shows a measured relation is no longer there.

When looking on the web for the specification of the GM sensors that are used in the Esprit for baro (GM 16006833) and MAP (GM 16009886), I found an Excel file with a graph and an equation for the MAP sensor: Vout=Vref*(0.005*P-0.04).

With a Vref of 5V this works nicely for the baro sensor too (my barometer here says 0.997 at the moment, the Esprit baro says 0.98 and the MAP says 1.02, engine is off).

The difference between the baro and the MAP reading can be caused by sensor inaccuracy. However, I used the equation provided for the GM sensor 16009886, for both the baro and the MAP, so if the equation for the baro (GM 16006833) is slightly different, this can be another explanation.

I would like to apply the correct corrections in the next version of Espritmon.

My question: does anyone have more accurate information on the GM 16006833 sensor used to measure baro pressure in the Esprit? Can the graph be reposted?

ah, I just found the spec of the sensor so the graph would be enough. The equation for the GM baro sensor is:

Vout = Vref ( ( 0.01059 * P ) - 0.1518 ).

Indeed according to the Lotus manual the voltage measured at the baro sensor should be approximately twice the value of the voltage at the output of the MAP sensor with the engine off.

However, unlike the ADC value for the MAP, the ALDL data stream clearly does not contain the ADC value (it is far too low) but a computed value (according to GM specs).

So my guess is that given the similarity between the the baro and the MAP values in the ALDL data (with the engine off), the baro value is converted to the same scale as the MAP by the ECU. This would allow me to use the earlier mentioned equation.

If Jonathan or someone else could repost the graph I could use this to verify this assumption.

Thank,

Erik

[sailorbob]

eriks4, one thing to watch out for is whether the Vref is 5.0v or slightly different (i.e. 5.12v). This caught me out when I first started reverse engineering Ford eec-iv ecu's.

[Jonathan]

PM me if I dont get on it within a day - so much to do (quitting work etc) I've only just got back off holidays as well.

Best thing to do is PM me then I clear my messages etc.

I do get around to doing stuff - just a little slow.

All my stuff goes to flickR now - much easier to work with.

[sailorbob]

Jonathan, do you have the Freescan datalog from that graph please?

[Jonathan]

Not really - all I did was attach a compressor (which was checked with an analouge and digital gauge for accuracy) to the MAP sensor.

I adjusted it back and forth at a known pressure and took the readings from freescan on paper - what you see from that is the graph.

Also I did voltage checks on the device to determine it's max saturation voltage (which is at 2.25 freescan iirc).

The confirmation for me was the fact my car always max boost ~2 which is very close to the 0.85bar max boost it should achieve.

It's meant as a guide really - a better guide that the old "take the freescan value and minus 1 off" to get real boost.

[sailorbob]

Oh well, not to worry. I was just curious about comparing the raw data the ecu was outputting rather than the Freescan converted data against the pressure you supplied to the sensor.

[Jonathan]

The only way I can read it is via freescan - to confirm the reading I also took voltage measurements from the sensor.

The idea was to basically give an idea of what real pressure is represented on freescan - I also wanted to see what the MAP sensor actually reads upto before I decided not to chip my car.

[eriks4]

Jonathan, this is of great help. I converted the Freescan values on your x-axis (1, 1.1, 1.2 etc.) to their original ECU value.

Freescan converts them like: FREESCAN_MAP=1+100*(ECU_value-130), hence ECU_value=(FREESCAN_MAP-1)/100+130

Next I converted the ECU value into the MAP it should have been using the equation for the GM 2 bar sensor: MAP=ECU_value*2/255+0.08.

This yields the following results:

Freescan value,ECU_value,TRUE_MAP

1,130,1.10

1.1,140,1.18

1.2,150,1.26

1.3,160,1.33

1.4,170,1.41

1.5,180,1.49

1.6,190,1.57

1.7,200,1.65

1.8,210,1.73

1.9,220,1.81

2,230,1.88

2.1,240,1.96

2.2,250,2.04

If you subtract 1 bar for baro and plot this in your graph you will see that the resulting plot of the corrected MAP matches vary closely to what you measured. Hence, using the GM equation to converted the ECU MAP value to MAP yields the correct results. I will soon post an update of Espritmon on my website with the new equations implemented.

Just made the changes, the new version of espritmon with corrected baro and MAP readings (GM equation) can be downloaded from my website. Would be great if you can test whether the pressure is indeed correct Jonathan. Now we can use the Esprit to predict the weather.

I also updated the manual regarding the control of the quad drivers, the modification of AFR and desired idle, and the communication with the ABS and SIR computers. Still haven't figured out how the alarm can be seduced to say something over the ALDL plug.

Have fun,

Erik

[Jonathan]

I onlyunderstood 50% of that LOL (me being thick)

Can't wait to try espritmon again and check it out.

Great job.

[BRGEsprit]

Am I reading this correctly?

Does this mean that all along people have been using Freescan and thinking that they are running higher boost than they really are?

Does this also mean that all of those numbered chips that have been sold do not actually make the boost levels they claim, because of an error in the equation used by Freescan?

This would explain the trouble shooting charts in the Lotus manuals stating 207kPa as a cut off point.

Tell me I'm mistaken.

[Bibs]

You're mistaken. This isn't news, it's just what we know put into a chart and explained entirely.

[eriks4]

In answer to the questions of BRGEsprit:

1. People that were unaware of the error in the equation used by Freescan may have thought they were running at a higher boost level than they were actually getting.

2. There is no relation between what the different chips can achieve in terms of boost limit and the error in Freescan

3. An important limitation is that given the range of the MAP sensor, the ECU is unable to tell the difference between any boost level above approx. 1 bar. This has its implications

[Jonathan]

With regards to point 3.

No one who has made chips for the car could tell me what the fueling is for anything above 1 bar.

Since no one was willing to talk about the ins and outs of boosting (who thing about the memcal seems to be one big secret) the turbo past 1 bar 1.1 / 1.2 bar I could only assume it was a bit of a con and a bit silly given the fact the car has no idea that there is a significant amount of air going in and keeps the fueling the same.

Mebbe I got it all wrong but I usually find when people hold info back and wont answer difficult questions, they're usually unsure themselves !

I think Erik has done some invaluable work on the ECU side of things.

[BRGEsprit]

Eric, no disrespect intended, I think you doing a knock-up job with the software you're putting together, but I would like to add some food for thought:

1. People were unaware of the error in reporting boost levels in Freescan? Which means almost everyone using it in unaware of the error.

2. The error in Freescan would be DIRECTLY related to the chips supplied by Marcus due to the fact that Andy Whittaker wrote Freescan and he also wrote the .TDF file that Marcus uses. Why would Andy use a different equation in the .TDF file than Freescan?

3. Given the limitation of the MAP sensor and the fact that it will set a code 33 if the manifold pressure is 204 kPa with the engine running is a dead give away that the chips are NOT making the boost they claim.

Reference from your chart:

2.2 BAR = 250 counts = 2.04 BAR

Along with a reference from the Service manual EMH 4 page 60

Code 33 set when:

Engine running

MAP signal greater than 90 kPa

TPS less than 0.4%

VSS less than 1 mph

Above conditions met for 20 seconds

OR

Engine running

MAP signal above 204 kPa

Also, just for people's archives, the entire Delco system is based in kPa, not BAR, there is only a very subtle difference, but it is as obvious as warm beer to an assembly language programmer.

[Jonathan]

Steve, when I did this experiment I was looking at a way to represent boost on a visual scale in the car becuase the boost gauge on mine is so horribly out at anything other than 0.5bar

I was also looking at getting the car chipped and wanted to know the relationship between the REAL boost reading and the freescan MAP value.

When I hocked the MAP sensor up to a compressor and looked at voltage readings I was suprised to see the car could not read much over 1Bar or 100KPa and wondered how chips managed to regulate pressure at this level.

Long and short is that it can't - from all I found out and worked out the top level of the MAP sensor is 214kPa (214 / 100 = 2.14 bar then take off the atmospheric pressure = ~1.14bar)

There are problems with Andy's original tdf file wrt to boost and it was never corrected aparently - and this was my issue that no-one could really answer.

All I know is that with the GM sensors the values achieved for boost pressure were not as expected so it was 'tweaked' to match.

IF the boost pressure tables were wrong, are not the fuel tables wrong as well ? (I strongly suspect they are)

IMO putting a chip into a car is just as easy to put a bleed valve in there and fudge the car like that !

Chipping over 1 Bar on a std compressor is not much use anyways but that's not the point.

My opinion of the chippers and chipping the car was lowered a lot by finding this out.

I would like to chip the car in later life but on my terms not someone elses which I havent a clue on - when I get time there are a few of us wanting to look at the definitions.

WRT to boost my conclusion was buy a good quality bleed valve and set it to just under 1.1 bar and let the engine stock code do the rest, it's a hell of a lot cheaper than some people charge for an EPORM these days.

[BRGEsprit]

I am interested to see that no one wants to comment on the FACT that the ECM will throw a code 33 at 204 kPa (2.04 BAR)

Since I am unaware of people reporting throwing a code 33 when running the various numbered versions of the chips, it must confirm the idea that the chips are not raising the true boost levels.

Additionally, the boost equation error likely explains why the numbered chips have no changes to the main fuel tables or the main ignition tables, they are most likley scaled improperly in the .TDF file.

[Vulcan Grey]

I get no codes at 2.24bar...

My car hits 2.24 bar all the time due to my altitude.

Even when triggering overboost cutoff, I get no codes.