MAF maxed at 5volts? Lets fix that!
06-24-2014, 01:07 AM
#1
MAF maxed at 5volts? Lets fix that!
MAF maxed at 5volts? Lets fix that!
Edit: Here is a great link that explains it even better:
http://www.focusfanatics.com/forum/s...d.php?t=202112
Thanks TeamRx8!
To do this we are using a voltage divider.
http://en.wikipedia.org/wiki/Voltage_divider
The actual resistances don't mater for the purposes of the voltage divider, the only thing that maters is the ratio of the two. If 75% of the total resistance is from the ECU signal to ground then the signal to the ecu is 75% of it's original value.
So if you go MAF → 10kohm ->Signal to ecu → 10kohm → ground.
5 volts from the maf turns into 2.5 volts to the ecu.
Or if you go MAF → 3.3kohm ->Signal to ecu → 10kohm → ground.
5 volts turns into about 3.75 volts to the ecu. (Which is .75 of the original value.)
That said, don't go much under 10k total or you may end up drawing enough power of the MAF that strange things happen, also if you bring the total resistance up high enough there might not be enough current available for the ecu to read voltage. You could also use a 10K+ potentiometer and dial it in where you want. I've read about people using closer to 100k total resistance and this might be a better plan. Regardless, the ratio is what's important!
Ok, so once you've installed the voltage divider now what? Well if you divide by .75 as in the example above then you can take your entire MAF table and multiply all the voltages by .75 That then brings your max voltage to 3.75 Squish the table to make room for more values on the end, and start calibrating at higher voltages. When I did it I ended up with about a 5-10% STFT that I had to calibrate out. Now it's close enough that I'm using LTFT to get the rest of the way there. There is a trick to approximating higher voltages, and I'll share it eventually.
If the Diablosport MAFIA is any indication then we should be able to double the capacity of our MAFs with limited negative effects.
Edit: Here is a great link that explains it even better:
http://www.focusfanatics.com/forum/s...d.php?t=202112
Thanks TeamRx8!
To do this we are using a voltage divider.
http://en.wikipedia.org/wiki/Voltage_divider
The actual resistances don't mater for the purposes of the voltage divider, the only thing that maters is the ratio of the two. If 75% of the total resistance is from the ECU signal to ground then the signal to the ecu is 75% of it's original value.
So if you go MAF → 10kohm ->Signal to ecu → 10kohm → ground.
5 volts from the maf turns into 2.5 volts to the ecu.
Or if you go MAF → 3.3kohm ->Signal to ecu → 10kohm → ground.
5 volts turns into about 3.75 volts to the ecu. (Which is .75 of the original value.)
That said, don't go much under 10k total or you may end up drawing enough power of the MAF that strange things happen, also if you bring the total resistance up high enough there might not be enough current available for the ecu to read voltage. You could also use a 10K+ potentiometer and dial it in where you want. I've read about people using closer to 100k total resistance and this might be a better plan. Regardless, the ratio is what's important!
Ok, so once you've installed the voltage divider now what? Well if you divide by .75 as in the example above then you can take your entire MAF table and multiply all the voltages by .75 That then brings your max voltage to 3.75 Squish the table to make room for more values on the end, and start calibrating at higher voltages. When I did it I ended up with about a 5-10% STFT that I had to calibrate out. Now it's close enough that I'm using LTFT to get the rest of the way there. There is a trick to approximating higher voltages, and I'll share it eventually.
If the Diablosport MAFIA is any indication then we should be able to double the capacity of our MAFs with limited negative effects.
Last edited by Harlan; 06-26-2014 at 02:33 PM. Reason: stuff
06-24-2014, 11:40 AM
#3
Haven't tested, but all hot wire MAFs can produce more than 5volts because of how they work. Eventually it will reach a max voltage (because 12volts is the high side) at which point air flow approaches infinity or it will become too restrictive to flow anymore before it hits that point. I found quite a few references to using voltage dividers to extend MAF range, but nothing was well written/explained so I didn't link anything.
06-24-2014, 12:17 PM
#4
Driving my unreliable rx8
I've played with this a bit but ran into a roadblock of the MAF output voltage.
I blew air across the MAF sensor with a compressor.
My first MAF would output as much as 6 Volts(IIRC), but the cob would never show more then 5 Volts. With the engine not running. It also showed 0g/sec
Then I swaped my MAF out with a Duralast gold from autozone, and It will not put out more then 5 Volts MAX.
I did source a MAF extenstion cable for 29 dollarsUnderdog Racing Development and have a handfull of 15KOHM Multi turn potentiometer's.
A potentiometer is simply a adjustable voltage divider. All you do is adjust it to your desired range then use a little nail polish to lock it in place. There is no need to play with various resistors. and It would allow you to lower your MAF just as much as you need no more.
I think I was looking at a total cost of 35-40 dollars for a DIY Plug and play MAF extender.
I blew air across the MAF sensor with a compressor.
My first MAF would output as much as 6 Volts(IIRC), but the cob would never show more then 5 Volts. With the engine not running. It also showed 0g/sec
Then I swaped my MAF out with a Duralast gold from autozone, and It will not put out more then 5 Volts MAX.
I did source a MAF extenstion cable for 29 dollarsUnderdog Racing Development and have a handfull of 15KOHM Multi turn potentiometer's.
A potentiometer is simply a adjustable voltage divider. All you do is adjust it to your desired range then use a little nail polish to lock it in place. There is no need to play with various resistors. and It would allow you to lower your MAF just as much as you need no more.
I think I was looking at a total cost of 35-40 dollars for a DIY Plug and play MAF extender.
Last edited by logalinipoo; 06-24-2014 at 12:25 PM.
06-24-2014, 12:28 PM
#5
But how often would you really need to alter the voltage range. With a simply voltage divider you just set it and forget it based on the requirements of the system. A pot just throws in an extra failure point... not to mention a headache and a half if someone or something mucks with the wiper
06-24-2014, 12:45 PM
#6
Driving my unreliable rx8
Yes, good point. Just sharing what I was thinking. Once POt's are set they normally aren't a issue IMHO.
On the upper end the Air flow raises Exponentialy when compared to the Voltage. I think 25% is a very extreme drop. I wanted to dial mine in directly to what I needed. with a 25% drop that basically raises your MAF to 6.66 Volts output. or 1.66 Volts extra. By my estimate that's about 350 G/sec over the 430 that you get at 5 Volts for a total of 780 g/sec. I don't thinnk the renesis will ever get to those power levels. and you loose a lot of definition in your MAF curve. By dialing it just as low as needed or a much smaller percent like 10 I think you'll have plenty of room and better definition in your MAF table.
at 5.5 Volts that should be about 530g/sec
Plus the Divider allowed me to prove the concept by letting the car idle and rolling it up or down then watching the g/sec roll up and down equally.
On the upper end the Air flow raises Exponentialy when compared to the Voltage. I think 25% is a very extreme drop. I wanted to dial mine in directly to what I needed. with a 25% drop that basically raises your MAF to 6.66 Volts output. or 1.66 Volts extra. By my estimate that's about 350 G/sec over the 430 that you get at 5 Volts for a total of 780 g/sec. I don't thinnk the renesis will ever get to those power levels. and you loose a lot of definition in your MAF curve. By dialing it just as low as needed or a much smaller percent like 10 I think you'll have plenty of room and better definition in your MAF table.
at 5.5 Volts that should be about 530g/sec
Plus the Divider allowed me to prove the concept by letting the car idle and rolling it up or down then watching the g/sec roll up and down equally.
Last edited by logalinipoo; 06-24-2014 at 12:50 PM.
06-24-2014, 12:47 PM
#7
Yeah that was my thought as well. You could use a pot for testing and then actual resistors for a permanent solution, but a pot for a permanent solution is a BAD idea. Imagine if it gets bumped, next thing you know you have to either reset it to known resistance values or retune.
Kinda surprised at the duralast MAF not going above 5volts. A MAF is just a 12volt to 5 volt voltage divider that happens to use heated wires for the resistance and flows air over them. If it maxes out at 5volts that means 5volts is where airflow goes infinite. It also means the slope of the curve went away from stock a bit before that. If so it could be a higher flow MAF already.
Sure not many people will need this, but lets just say I'm an optimist. There may come a day when people are routinely maxing the maf with turbo setups, and running for more than 20k.
Kinda surprised at the duralast MAF not going above 5volts. A MAF is just a 12volt to 5 volt voltage divider that happens to use heated wires for the resistance and flows air over them. If it maxes out at 5volts that means 5volts is where airflow goes infinite. It also means the slope of the curve went away from stock a bit before that. If so it could be a higher flow MAF already.
Sure not many people will need this, but lets just say I'm an optimist. There may come a day when people are routinely maxing the maf with turbo setups, and running for more than 20k.
06-24-2014, 01:00 PM
#10
Driving my unreliable rx8
ok, you lost me a little on the % change per hundreth of a volt.
Have you considered how you would run your MAF scale? I was thinking of adjusting the voltage not the MAF for the first 103 g/sec. That way all the IDLE and crusing is at the same resolution. Then squeezing all the rest of the Air flow into the upper end of the MAF scale.
Have you considered how you would run your MAF scale? I was thinking of adjusting the voltage not the MAF for the first 103 g/sec. That way all the IDLE and crusing is at the same resolution. Then squeezing all the rest of the Air flow into the upper end of the MAF scale.
06-24-2014, 01:28 PM
#11
this is neat but what is the max current the ecu will put out it before it considers the MAF to be faulty?
I am assuming your N/A, in principle this would work out great for a boosted engine, but what will the ecu do on the upper limit, looking at your first diagram, if your boosting say 20psi and you reach 3.25 volts, then that's 6.5 original,
I would imagine that it would require more current then the ecu is expecting to send and you would get a High maf input CEL.
this same condition crops up if the MAF is wet.
I am assuming your N/A, in principle this would work out great for a boosted engine, but what will the ecu do on the upper limit, looking at your first diagram, if your boosting say 20psi and you reach 3.25 volts, then that's 6.5 original,
I would imagine that it would require more current then the ecu is expecting to send and you would get a High maf input CEL.
this same condition crops up if the MAF is wet.
06-24-2014, 01:34 PM
#12
Driving my unreliable rx8
I've never seen that error. We have a couple of people who have maxed out the 5 Volts and I've never seen them say anything about it.
By maxed out I am not saying they are only producing 5 volts. They are producing that or more. But the ECU puts out the 5 volt restriction.
All this is doing is reducing the voltage going out from the MAF into the ECU.
By maxed out I am not saying they are only producing 5 volts. They are producing that or more. But the ECU puts out the 5 volt restriction.
All this is doing is reducing the voltage going out from the MAF into the ECU.
06-24-2014, 01:52 PM
#13
Look at the stock MAF table. From 5.45 to 6.00 g/sec there is a 10% change in airflow, but only a .04volt change. That is a resolution of 2.5% per hundredth of a volt. On the top end from 355 to 365 g/sec there is a 2.8% change in air flow for annother .04 volt change, or about .7% per hundredth of a volt.
So, until you aproach the actual max flow of the maf the low end is more restrictive. If it still idles you're fine.
MAF scale is not that important. The ECU linearly interpolates values between points on the table, you could remove half the table and still run just fine. Your engine doesn't care about the exact number, it cares about getting close enough percentage wise to get a good fuel air mix with the voltage it sees from the MAF. The important thing is to provide enough data points that the interpolation is still good enough. When I went to .75 voltage I threw out some data points and did a little math with the rest to end up with a voltage scale near stock all the way to 3.75volts. I then modified an existing table in a spreadsheet to estimate the maf flow the rest of the way up. When I pretty up the spreadsheet I'm using I'll share it. This is not to go in the place of calibrating your MAF, it's just to give you an educated guess on where to start. I was less than 10% off the entire band before cal.
MAF current doesn't change much with this mod, that's the point of keeping resistance above 10kohm. I don't know about the associated CELs, I haven't experienced them. I know is that some cars have high and low volts CELs, but
haven't seen ours. Oh and if you change the first two cells of the table you must be careful on the slope to not let the ECU interpolate a negative g/s while the engine is stopped or coasting. I think that's the origin of that CEL. They can be modified, but if voltage gets low enough that the ecu thinks g/s should be negative it gets pissy.
So, until you aproach the actual max flow of the maf the low end is more restrictive. If it still idles you're fine.
MAF scale is not that important. The ECU linearly interpolates values between points on the table, you could remove half the table and still run just fine. Your engine doesn't care about the exact number, it cares about getting close enough percentage wise to get a good fuel air mix with the voltage it sees from the MAF. The important thing is to provide enough data points that the interpolation is still good enough. When I went to .75 voltage I threw out some data points and did a little math with the rest to end up with a voltage scale near stock all the way to 3.75volts. I then modified an existing table in a spreadsheet to estimate the maf flow the rest of the way up. When I pretty up the spreadsheet I'm using I'll share it. This is not to go in the place of calibrating your MAF, it's just to give you an educated guess on where to start. I was less than 10% off the entire band before cal.
MAF current doesn't change much with this mod, that's the point of keeping resistance above 10kohm. I don't know about the associated CELs, I haven't experienced them. I know is that some cars have high and low volts CELs, but
haven't seen ours. Oh and if you change the first two cells of the table you must be careful on the slope to not let the ECU interpolate a negative g/s while the engine is stopped or coasting. I think that's the origin of that CEL. They can be modified, but if voltage gets low enough that the ecu thinks g/s should be negative it gets pissy.
06-24-2014, 02:01 PM
#14
Look at the stock MAF table. From 5.45 to 6.00 g/sec there is a 10% change in airflow, but only a .04volt change. That is a resolution of 2.5% per hundredth of a volt. On the top end from 355 to 365 g/sec there is a 2.8% change in air flow for annother .04 volt change, or about .7% per hundredth of a volt.
So, until you aproach the actual max flow of the maf the low end is more restrictive. If it still idles you're fine.
MAF scale is not that important. The ECU linearly interpolates values between points on the table, you could remove half the table and still run just fine. Your engine doesn't care about the exact number, it cares about getting close enough percentage wise to get a good fuel air mix with the voltage it sees from the MAF. The important thing is to provide enough data points that the interpolation is still good enough. When I went to .75 voltage I threw out some data points and did a little math with the rest to end up with a voltage scale near stock all the way to 3.75volts. I then modified an existing table in a spreadsheet to estimate the maf flow the rest of the way up. When I pretty up the spreadsheet I'm using I'll share it. This is not to go in the place of calibrating your MAF, it's just to give you an educated guess on where to start. I was less than 10% off the entire band before cal.
MAF current doesn't change much with this mod, that's the point of keeping resistance above 10kohm. I don't know about the associated CELs, I haven't experienced them. I know is that some cars have high and low volts CELs, but
haven't seen ours. Oh and if you change the first two cells of the table you must be careful on the slope to not let the ECU interpolate a negative g/s while the engine is stopped or coasting. I think that's the origin of that CEL. They can be modified, but if voltage gets low enough that the ecu thinks g/s should be negative it gets pissy.
So, until you aproach the actual max flow of the maf the low end is more restrictive. If it still idles you're fine.
MAF scale is not that important. The ECU linearly interpolates values between points on the table, you could remove half the table and still run just fine. Your engine doesn't care about the exact number, it cares about getting close enough percentage wise to get a good fuel air mix with the voltage it sees from the MAF. The important thing is to provide enough data points that the interpolation is still good enough. When I went to .75 voltage I threw out some data points and did a little math with the rest to end up with a voltage scale near stock all the way to 3.75volts. I then modified an existing table in a spreadsheet to estimate the maf flow the rest of the way up. When I pretty up the spreadsheet I'm using I'll share it. This is not to go in the place of calibrating your MAF, it's just to give you an educated guess on where to start. I was less than 10% off the entire band before cal.
MAF current doesn't change much with this mod, that's the point of keeping resistance above 10kohm. I don't know about the associated CELs, I haven't experienced them. I know is that some cars have high and low volts CELs, but
haven't seen ours. Oh and if you change the first two cells of the table you must be careful on the slope to not let the ECU interpolate a negative g/s while the engine is stopped or coasting. I think that's the origin of that CEL. They can be modified, but if voltage gets low enough that the ecu thinks g/s should be negative it gets pissy.
06-25-2014, 10:08 PM
#15
Congratulations on your basic google-fu skills 
https://www.rx8club.com/series-i-maj...4/#post4597383
https://www.rx8club.com/series-i-maj...4/#post4597383
06-26-2014, 02:48 PM
#18
https://www.diablosport.com/forum/vi...hp?f=12&t=5359
https://www.diablosport.com/forum/vi...p?f=12&t=35159
.
https://www.diablosport.com/forum/vi...p?f=12&t=35159
.
Last edited by TeamRX8; 06-26-2014 at 02:52 PM.
06-26-2014, 03:10 PM
#21
vout = vin x (lowerR / (upperR + lowerR))
if vin = 5, with a lowerR of 10,000ohm and upperR of 3300ohm, vout here = 3.759...
Mind you this is with 0 load.
You are correct in that the resistance (impedance) of your load will define the actual output
Last edited by paimon.soror; 06-26-2014 at 03:28 PM.
.