I get bored easily.
01-31-2012, 02:55 AM
#81
Elite Member
Join Date: Jul 2005
Posts: 6,420
Total Cats: 84
If the gate driver ground pin is gonna be separated from the MOSFET Source pin by 2 long ground wires that go to the engine block, the MOSFET gate resistor better be increased to more like 1 kohm. And it has to physically sit close to the MOSFET with short trace to the gate pin. If not you might have yourself a very energetic 20 MHz oscillator and the magic smoke might come out.
On the datasheet of the IXDI414, and "clean analog ground" is relative to its VCC. It doesn't care if it sits on a ground plane that is bouncing around relative to the engine block. What matters then is the noise it sees on its input, and how tight the connection to its MOSFET is.
And that's why I would use a 74C14.
On the datasheet of the IXDI414, and "clean analog ground" is relative to its VCC. It doesn't care if it sits on a ground plane that is bouncing around relative to the engine block. What matters then is the noise it sees on its input, and how tight the connection to its MOSFET is.
And that's why I would use a 74C14.
Reply
0
0
02-15-2012, 04:27 AM
#90
Boost Pope
Thread Starter
iTrader: (8)
Join Date: Sep 2005
Location: Chicago. (The less-murder part.)
Posts: 33,485
Total Cats: 6,898
For what it's worth, I've seen a lot of people ask "can I tune with a narrowband sensor", and I'm officially changing my answer from "Well, probably, but it'd be sub-optimal" to "you'd have to be an idiot to try."
With a fresh ECU and an untuned VE table, trying to use a narrowband sensor to get things whipped into shape is like trying to un-bend a connecting rod using a PowerCard.
Fortunately, I still have my old AEM UEGO lying around which I pulled out of the red car in disgust and replaced with an LC-1 several years ago. It's not the most accurate thing on the planet, but it's "good enough" to get the VE table roughed in. I had actually planned on this possible contingency when building the Hellasquirt, and brought out an additional analog input from the CPU's AD6, so that I can connect the NB02 sensor to it as a calibration reference for the wideband. I also split the wiring in the harness to allow swapping this wire with the O2 wire, and additionally brought out power, ground, and the aux analog input on a separate 3-pin Molex connector at the harness.
So, long story short, with the AEM connected to O2 and the narrowband connected to AD6, I can see that the AEM is, as usual, reading about 0.5AFR lean at stoich, and have corrected the AFR table to accommodate this. With that done, it's running well enough to get the VE table sort of roughed-in by hand and then let TunerStudio do its magic.
In keeping with the low-budget theme of this build, behold my ghetto-fabulous gauge mounting solution:
(Yes, that's a Fedex box.)
The ultimate goal is to return to a narrowband sensor once the VE table is more or less beaten into shape.
With a fresh ECU and an untuned VE table, trying to use a narrowband sensor to get things whipped into shape is like trying to un-bend a connecting rod using a PowerCard.
Fortunately, I still have my old AEM UEGO lying around which I pulled out of the red car in disgust and replaced with an LC-1 several years ago. It's not the most accurate thing on the planet, but it's "good enough" to get the VE table roughed in. I had actually planned on this possible contingency when building the Hellasquirt, and brought out an additional analog input from the CPU's AD6, so that I can connect the NB02 sensor to it as a calibration reference for the wideband. I also split the wiring in the harness to allow swapping this wire with the O2 wire, and additionally brought out power, ground, and the aux analog input on a separate 3-pin Molex connector at the harness.
So, long story short, with the AEM connected to O2 and the narrowband connected to AD6, I can see that the AEM is, as usual, reading about 0.5AFR lean at stoich, and have corrected the AFR table to accommodate this. With that done, it's running well enough to get the VE table sort of roughed-in by hand and then let TunerStudio do its magic.
In keeping with the low-budget theme of this build, behold my ghetto-fabulous gauge mounting solution:
(Yes, that's a Fedex box.)
The ultimate goal is to return to a narrowband sensor once the VE table is more or less beaten into shape.
Reply
0
0
02-15-2012, 01:13 PM
#92
Boost Pope
Thread Starter
iTrader: (8)
Join Date: Sep 2005
Location: Chicago. (The less-murder part.)
Posts: 33,485
Total Cats: 6,898
First, the narrowband gauge is inherently more accurate. This AEM unit I have is known to have some offset in the analog output, and I don't really trust its internal calibration to begin with. By comparison, I have a brand new 4-wire narrowband installed, and the "calibration" of narrowband sensors, by definition, is more or less inherently perfect. There is some temperature-related drift in the voltage, but regardless of the absolute "rail" voltages, it will always undergo a sharp swing in output right at the stoichiometric point.
As I need to pass CA emissions, accuracy of the O2 sensor at the stoichiometric point is critical.
Second, my ultimate goal for the car is to have a 100% stock appearance, which again goes back to the CA emissions thing. So that means only one O2 sensor installed in the downpipe. I should mention at this point that some previous owner, for reasons unknown, had a second O2 sensor bung welded into the downpipe. When I bought the car, there was an old 4-wire sensor still in the second hole, with the wires cut off. (I found no wiring in the car which corresponded to it, so this pipe may have come from a different car.) This was convenient for me as it allows me to run the wideband along with a narrowband reference sensor, but ultimately I will be removing the wideband and plugging the hole, thus returning to an apparently stock configuration.
Obviously this means I will be ultimately be running open-loop at the very top of the VE table, so I'll just use the wideband to get that part of the map dialed-in where I want it, and then lock it off and leave it alone.
Reply
0
0
02-15-2012, 01:42 PM
#93
Elite Member
iTrader: (3)
Join Date: Apr 2008
Location: Outside Portland Maine
Posts: 2,023
Total Cats: 19
Well, a couple of reasons.
First, the narrowband gauge is inherently more accurate. This AEM unit I have is known to have some offset in the analog output, and I don't really trust its internal calibration to begin with. By comparison, I have a brand new 4-wire narrowband installed, and the "calibration" of narrowband sensors, by definition, is more or less inherently perfect. There is some temperature-related drift in the voltage, but regardless of the absolute "rail" voltages, it will always undergo a sharp swing in output right at the stoichiometric point.
As I need to pass CA emissions, accuracy of the O2 sensor at the stoichiometric point is critical.
Second, my ultimate goal for the car is to have a 100% stock appearance, which again goes back to the CA emissions thing. So that means only one O2 sensor installed in the downpipe. I should mention at this point that some previous owner, for reasons unknown, had a second O2 sensor bung welded into the downpipe. When I bought the car, there was an old 4-wire sensor still in the second hole, with the wires cut off. (I found no wiring in the car which corresponded to it, so this pipe may have come from a different car.) This was convenient for me as it allows me to run the wideband along with a narrowband reference sensor, but ultimately I will be removing the wideband and plugging the hole, thus returning to an apparently stock configuration.
Obviously this means I will be ultimately be running open-loop at the very top of the VE table, so I'll just use the wideband to get that part of the map dialed-in where I want it, and then lock it off and leave it alone.
First, the narrowband gauge is inherently more accurate. This AEM unit I have is known to have some offset in the analog output, and I don't really trust its internal calibration to begin with. By comparison, I have a brand new 4-wire narrowband installed, and the "calibration" of narrowband sensors, by definition, is more or less inherently perfect. There is some temperature-related drift in the voltage, but regardless of the absolute "rail" voltages, it will always undergo a sharp swing in output right at the stoichiometric point.
As I need to pass CA emissions, accuracy of the O2 sensor at the stoichiometric point is critical.
Second, my ultimate goal for the car is to have a 100% stock appearance, which again goes back to the CA emissions thing. So that means only one O2 sensor installed in the downpipe. I should mention at this point that some previous owner, for reasons unknown, had a second O2 sensor bung welded into the downpipe. When I bought the car, there was an old 4-wire sensor still in the second hole, with the wires cut off. (I found no wiring in the car which corresponded to it, so this pipe may have come from a different car.) This was convenient for me as it allows me to run the wideband along with a narrowband reference sensor, but ultimately I will be removing the wideband and plugging the hole, thus returning to an apparently stock configuration.
Obviously this means I will be ultimately be running open-loop at the very top of the VE table, so I'll just use the wideband to get that part of the map dialed-in where I want it, and then lock it off and leave it alone.
Reply
0
0
02-15-2012, 02:16 PM
#97
Boost Pope
Thread Starter
iTrader: (8)
Join Date: Sep 2005
Location: Chicago. (The less-murder part.)
Posts: 33,485
Total Cats: 6,898
In all honesty, I'm not entirely certain. The law is extremely vague on this point, and I have heard of penalties ranging from fines to seizure and crushing. It's not my intention to find out for sure.
Having a pre OBD-II vehicle makes things slightly easier in this regard, since there's no plugin test to be done- only the visual and tailpipe inspection. The tailpipe test, in theory, should not be a problem, so I'm focusing on visual.
To that end, I'm trying to keep everything as stock-like in appearance as possible. This means no abnormal-looking wiring under the hood, no gauges on the interior, etc. I have even set the system up so that the Check Engine light comes on when you power up the car, and then goes off once RPM rises above 500.
Had.
I sold the red car late last year. Still not entirely sure what the next project is going to be.
Reply
0
0
02-15-2012, 02:46 PM
#100
Boost Pope
Thread Starter
iTrader: (8)
Join Date: Sep 2005
Location: Chicago. (The less-murder part.)
Posts: 33,485
Total Cats: 6,898
But no turbo for this car. It's a shortnose with poor compression numbers that I paid $800 for, and I genuinely do intend to keep this one stock just to have as the "normal" car.
Reply
0
0
