2byte airflow limit?
Dec 20, 2008, 07:02 PM
#32
Evolving Member
iTrader: (10)
Join Date: Aug 2005
Location: San Francisco, CA
Posts: 246
Likes: 0
Received 0 Likes
on
0 Posts
Glad you like the sig. lol
Wow, the HTA86 really "sucks" up top! That airflow is crazy, and at least we know the MAF is good all the way to 3500 Hz. Would you mind posting the log with time? I want to see how fast it is up top. Thanks.
Wow, the HTA86 really "sucks" up top! That airflow is crazy, and at least we know the MAF is good all the way to 3500 Hz. Would you mind posting the log with time? I want to see how fast it is up top. Thanks.
Dec 21, 2008, 01:34 AM
#33
wow Kev, that's no little drop either. I really think something else is going on...injector scale change perhaps without accounting for it in the evoscan calculation maybe? then again that's probably 2byteload...hrmmm. If the airflow truly decreased that much the power would have dropped accordingly. I assume it was dyno'd after the change to get the power back where it was.
If that where true, you could "dial in horsepower" by changing the screw plug on the bottom of the maf housing.
Dec 23, 2008, 02:29 PM
#38
Granted, this is all on 2Gs, but I wouldn't be surprised if the EVO acted similarly.
On the 2Gs with DSMLink, it was pretty apparent when you reached the limit of the airflow meter. Typically the MAF signal would be very smooth. Once you got to ~200 HZ from where it would max out, the signal would start to get rough. The higher the frequency went, the more it bounced around. Once you got to the max airflow it could read, the signal bounced around a couple hundred Hz.
In my 2G, the maf was limited to ~2900 Hz. Around 2700 Hz it would get rough. This corresponded to an airflow of ~46 pounds/min on a 70 degree day at 4200'. In 0 degree weather, it was like 52 pounds/min. The MAF sensor is a velocity sensor, so the air velocity is what will actually be the limit. At sea level, people would see similar MAF readings (2900 Hz) at around 52-54 pounds/min.
I don't know if you'll see this though with the slow datalogging rate we typically use. I think DSMLink was looking at the MAF frequency at like 100Hz. If the samples are averaged at all in the dataloggers we use, you'll never see this signal bouncing around. It should just show up as the frequency hitting a maximum limit and sticking close to it despite increases in boost.
When guys went over to the GM MAFT, I think a couple measured over 4500Hz without issue. The ECU seemed to be capable of dealing with much higher frequencies then what the MAF could produce. I don’t remember if the ECU maxed out at a certain (calculated) airflow value though despite measuring the higher MAF frequency.
Hacking the MAF seems to be very hit or miss depending on the car/setup.
On the 2Gs with DSMLink, it was pretty apparent when you reached the limit of the airflow meter. Typically the MAF signal would be very smooth. Once you got to ~200 HZ from where it would max out, the signal would start to get rough. The higher the frequency went, the more it bounced around. Once you got to the max airflow it could read, the signal bounced around a couple hundred Hz.
In my 2G, the maf was limited to ~2900 Hz. Around 2700 Hz it would get rough. This corresponded to an airflow of ~46 pounds/min on a 70 degree day at 4200'. In 0 degree weather, it was like 52 pounds/min. The MAF sensor is a velocity sensor, so the air velocity is what will actually be the limit. At sea level, people would see similar MAF readings (2900 Hz) at around 52-54 pounds/min.
I don't know if you'll see this though with the slow datalogging rate we typically use. I think DSMLink was looking at the MAF frequency at like 100Hz. If the samples are averaged at all in the dataloggers we use, you'll never see this signal bouncing around. It should just show up as the frequency hitting a maximum limit and sticking close to it despite increases in boost.
When guys went over to the GM MAFT, I think a couple measured over 4500Hz without issue. The ECU seemed to be capable of dealing with much higher frequencies then what the MAF could produce. I don’t remember if the ECU maxed out at a certain (calculated) airflow value though despite measuring the higher MAF frequency.
Hacking the MAF seems to be very hit or miss depending on the car/setup.
Dec 23, 2008, 02:49 PM
#39
And the subject of the GM MAF brings a good point, as I've seen conflicting opinions as to the actual metering capacity of the GM 3" MAF. Perhaps it depends upon placement of the sensor.
Why?
If the Karmann Vortex design meters velocity without respect to air density (which is measured elsewhere), that implies a blow-through setup would be capable of supporting greater power, due to the increased air density being the primary empowering factor as opposed to increased velocity.
Why?
If the Karmann Vortex design meters velocity without respect to air density (which is measured elsewhere), that implies a blow-through setup would be capable of supporting greater power, due to the increased air density being the primary empowering factor as opposed to increased velocity.
Dec 23, 2008, 03:07 PM
#40
On DSMs (the only really testbed I have seen) going to a blow through design is worth power just because the intake isnt being restricted. On the one Evo I have tuned with MAFT and a GM 3" the only thing I noticed about metering is how dependent it is on air temp and humidity (since there is no correction in the MAFT for that).
The car in question made 438whp on pumpgas (24-25psi) with a 2.0L ETS 61mm kit, blow through, stock computer.
We did another car that was very similar, 2.0L, ETS 61mm kit, MAFT Pro in speed density, stock computer, etc. This car made 436whp at the same boost level, which to me would indicate that the GM MAF is basically not a restriction. The first car was a stock Cam IX and the second was an VIII with HKS272s so to me its about a wash as far as the cam situation.
If it does in fact become a restriction it wouldnt be until much later in the ultimate airflow game but I honestly havent watched what the highest blow through DSM is currently so I have no real numbers to back up.
I can however say I know of one 35R first gen that made 550whp on E85, and another one on a 3065 that made 542whp both in blow through at about 33-34psi. The 3065 car had a sheetmetal intake manifold whereas the 35R was on a stocker at the time.
The car in question made 438whp on pumpgas (24-25psi) with a 2.0L ETS 61mm kit, blow through, stock computer.
We did another car that was very similar, 2.0L, ETS 61mm kit, MAFT Pro in speed density, stock computer, etc. This car made 436whp at the same boost level, which to me would indicate that the GM MAF is basically not a restriction. The first car was a stock Cam IX and the second was an VIII with HKS272s so to me its about a wash as far as the cam situation.
If it does in fact become a restriction it wouldnt be until much later in the ultimate airflow game but I honestly havent watched what the highest blow through DSM is currently so I have no real numbers to back up.
I can however say I know of one 35R first gen that made 550whp on E85, and another one on a 3065 that made 542whp both in blow through at about 33-34psi. The 3065 car had a sheetmetal intake manifold whereas the 35R was on a stocker at the time.
Dec 23, 2008, 03:21 PM
#41
I see, but while I do not claim to be an engineer in fluid dynamics, a MAF at atmospheric pressure (on the intake) is subjected to far greater air volume and velocity than is the same meter in an area of increased air density (blow through). In other words, I can't help but wonder if the factory MAF may function with increased, improved capacity and response in a blow through situation due to reduced restriction - and without the need for an MAFT.
I may have to try it.
I may have to try it.
Dec 23, 2008, 03:46 PM
#44
Newbie
iTrader: (11)
Join Date: Aug 2006
Location: Ca
Posts: 85
Likes: 0
Received 0 Likes
on
0 Posts
when i tried blowthrough with an evo maf an evo rom(3g eclipse) it just doesnt work once the pressure builds. The airflow Hz maxes at about 8-900 and the car wouldnt go over 200 load or 4k rpms, until i went to suckthrough w/ evo rom. Karmen vortex just wont work with pressure. The 3g eclipse an evo8/9 mafs are the same in respect to dimensions. If i flash stock evo scaling settings it works in a suckthrough setup. If i remember correctly, its somethin about the baro sensor cant see/compensate the pressure changes. I dont have those logs anymore but at 3.5-4k with a evo8 turbo the spoolup would start an the car would just lean out an bog.
Didnt wanna fool with the maft anymore so i had to change it to suckthrough.
Didnt wanna fool with the maft anymore so i had to change it to suckthrough.
Last edited by Dyrexion; Dec 23, 2008 at 03:51 PM.
Dec 23, 2008, 03:48 PM
#45
Well, I forgot a lot of facts about the Karmen vortex MAS over the years, but for one, our MAS does not have a pressure sensor capable of measuring the pressures in blow-through. I'm not positive about the temperature sensor.
As you may or may not know, our MAS houses it's own baro and temp sensors, so that data can be used by the ECU to calculate a mass airflow.
Besides that, if the Karmen MAS could actually meter airflow properly in a pressurized environment (not sure how the airflow vortices would be affected), I would assume the scaling would be drastically different. That MAF scaling table in the ECU is basically a Hz vs L/Hz table.
mrfred has done a lot of disassembly in the airflow routines, but I would imagine that many tables would need to be adjusted to new calibrations.
Of course, all of that is a moot point, since the MAS sensor doesn't have the proper pressure and temp sensors incorporated anyway.
As you may or may not know, our MAS houses it's own baro and temp sensors, so that data can be used by the ECU to calculate a mass airflow.
Besides that, if the Karmen MAS could actually meter airflow properly in a pressurized environment (not sure how the airflow vortices would be affected), I would assume the scaling would be drastically different. That MAF scaling table in the ECU is basically a Hz vs L/Hz table.
mrfred has done a lot of disassembly in the airflow routines, but I would imagine that many tables would need to be adjusted to new calibrations.
Of course, all of that is a moot point, since the MAS sensor doesn't have the proper pressure and temp sensors incorporated anyway.