Miata Turbo Forum - Boost cars, acquire cats.

Miata Turbo Forum - Boost cars, acquire cats. (https://www.miataturbo.net/)
-   Insert BS here (https://www.miataturbo.net/insert-bs-here-4/)
-   -   How (and why) to Ramble on your goat sideways (https://www.miataturbo.net/insert-bs-here-4/how-why-ramble-your-goat-sideways-46882/)

stratosteve 06-26-2018 04:40 PM


Originally Posted by y8s (Post 1488471)
do you have EGR installed and working?

are you running the factory ECU to control the fuel and satisfy the O2 sensors?

tricking the system is incredibly hard. The ECU wants to see a very specific lean-rich pattern for each O2 sensor (they are different) to show that both sensors and the catalyst are working properly. If either is out of balance, it will either never become ready or throw a code.

Should have listed current setup....
factory ecu/injectors, wastegate disabled, and egr installed with a block off plate. Front o2 sensor requires an extension to reach the factory plug. Rear o2 sensor is installed but removed from exhaust in case high flow cat would cause issues. I ran the rear in the exhaust stream also, with no difference.

Joe Perez 06-26-2018 08:54 PM

I just envisioned an alternate past in which all police cars were Karmann Ghias.

y8s 06-27-2018 10:32 AM


Originally Posted by stratosteve (Post 1488473)
Should have listed current setup....
factory ecu/injectors, wastegate disabled, and egr installed with a block off plate. Front o2 sensor requires an extension to reach the factory plug. Rear o2 sensor is installed but removed from exhaust in case high flow cat would cause issues. I ran the rear in the exhaust stream also, with no difference.

EGR: There's a boost sensor that may need to see a change in manifold pressure when the EGR opens. A block off will prevent that. I'm not sure if this is an issue. Why not just run regular EGR if you're already carrying the weight? It's not all that bad.

O2 sensors: rear needs to be installed downstream of the cat. It should work if the cat works.

Can you plug in an OBD2 monitoring device and verify the signals from the two O2 sensors? Graphing both outputs at the same time is very helpful.

You'll probably want to refer to this:


https://cimg4.ibsrv.net/gimg/www.mia...299b716f8a.png

Once you know what the signals look like, you can start to diagnose.

stratosteve 06-27-2018 11:45 AM


Originally Posted by y8s (Post 1488586)
EGR: There's a boost sensor that may need to see a change in manifold pressure when the EGR opens. A block off will prevent that. I'm not sure if this is an issue. Why not just run regular EGR if you're already carrying the weight? It's not all that bad.

O2 sensors: rear needs to be installed downstream of the cat. It should work if the cat works.

Can you plug in an OBD2 monitoring device and verify the signals from the two O2 sensors? Graphing both outputs at the same time is very helpful.

You'll probably want to refer to this:


https://cimg4.ibsrv.net/gimg/www.mia...299b716f8a.png

Once you know what the signals look like, you can start to diagnose.

Before I started this build, i researched the egr portion. I was under the impression, as long as the egr sensor was connected, that satisfied what the ECU needed.

As far as the rear o2, the ecu needs to see less fuel than the upstream sensor sees, to conclude that the cat is doing its thing.

I do have an obd2 scanner and I have looked at the live data. I havent seen anything that looks wrong. I am getting different voltages from front/rear o2 sensors. I guess I need a tech 2 to see what is holding up the process.

B6Tfastiva 06-27-2018 12:38 PM

Ecu needs to see the rear o2 voltage switch between lean/rich just like the front so it need to be in the exhaust stream. Ummmm tech 2 is for gm vehicles unless I misunderstood what you're working on.

stratosteve 06-27-2018 01:59 PM


Originally Posted by B6Tfastiva (Post 1488614)
Ecu needs to see the rear o2 voltage switch between lean/rich just like the front so it need to be in the exhaust stream. Ummmm tech 2 is for gm vehicles unless I misunderstood what you're working on.


I was referencing a more advanced scan tool when mentioning the tech 2. My error.

I have been running a rear o2 in two spark plug defoulers (out of the exhaust stream) on a 01 protege for about 100k. It's been through battery changes etc, so the memory has been wiped and had to start over as far as drive cycles. The pre cat is sensitive in that car and triggers a p0420 code easily. That fixed it and I havent had an issue with readiness codes, ever in that car.

y8s 06-27-2018 04:02 PM

rear o2 needs to see a 2x or more slower cycling rate than front. and both need to cycle within a specific duty cycle range near 50% or else readiness will not be set. If they are too far off 50%, you'll get a lean or rich code.

Godless Commie 06-27-2018 07:50 PM

Wait..
I was under the impression the rear O2 sensor should not oscillate if the cat is doing its job.
Front O2 sensor's activity manifests itself in a sinusoidal graph because of the instantaneous corrections in fuel mixture.
What "drives" an O2 sensor is the amount of oxygen in the exhaust stream. Once those gases go through the cat, they have been processed and broken down.
So, it would not be feasible for the rear sensor to display a similar pattern as the front one.
Rear sensor should display a "flattish" line when graphed.
The ECU throws a code if the rear sensor starts oscillating, because that pretty much means the cat is not functioning as designed.

y8s 06-28-2018 09:23 AM

Keep in mind this is during closed loop, light load operation only.

And that I didn't write the FSM, I only used it to design a clever piece of electronics that plugged into the stock ECU when I was running another ECU in piggyback.

The circuit did a fantastic job of providing the correct O2 signals for both sensors and had the ability to adjust for both duty cycle and relative frequency. The car managed to set all the ready codes. And as long as I didn't drive too long, it didn't throw any codes. The thread about this process is around here somewhere...

(time passes)

hnng: https://www.miataturbo.net/adaptroni...82/#post476173

stratosteve 06-28-2018 11:10 AM

https://cimg3.ibsrv.net/gimg/www.mia...e49994724d.jpg
Using the above procedure to satisfy the drive cycle needs.....I cant legally do this right off my driveway. I have to drive several miles, so that I can drive 55mph. If there is a delay between steps, will that cause a problem?

(downstream o2 is installed in it's proper location now)

Guardiola 06-28-2018 11:54 AM

Get a temp tag from your DMV. They are made for exactly this kind of situation.
As long as you have proof of insurance, a temp tag will get you on the road so you can get the car ready for inspection.

stratosteve 06-28-2018 01:08 PM


Originally Posted by Guardiola (Post 1488790)
Get a temp tag from your DMV. They are made for exactly this kind of situation.
As long as you have proof of insurance, a temp tag will get you on the road so you can get the car ready for inspection.

Car is legal for the road, but doing 55 in my neighborhood is not.

y8s 06-28-2018 03:46 PM

you can do the 55 portion on a different day

rleete 06-28-2018 07:49 PM

How do you break up two blind guys fighting?


Yell, “I got 10 bucks on the guy with the knife!”

kenzo42 06-30-2018 11:25 PM


Originally Posted by gooflophaze (Post 1487705)
555 timer in monostable mode. 555 (NE555) Monostable Circuit Calculator

Basically, with your switch as a trigger, the 555 will perform an action for x amount of time. So plug in a 1 uf capacitor and 500-ish ohms resistor, and you'll get a .55 second "beep" on your output (being driven by the mosfet.

Actually, now that I think about it, I'm not sure if the 555 will one shot in the situation you described above. http://www.n5dux.com/ham/files/pdf/F...20Circuits.pdf is how I started with 555's though. https://m8051.blogspot.com/2013/02/r...g-circuit.html looks closer to what you want.

I built this monostable 555 circuit, but if I hold the switch down, it continues to beep. It beeps .55sec only if I do a push and then release. Is the link you posted to correct for this?

gooflophaze 07-01-2018 12:35 AM

Yeap. So what you need is known as one-shot mode. Right now it does a fixed-length output on a single trigger - but when it's holding on an closed trigger it continues to output - which is what I realized when I put the 2nd link in ( https://m8051.blogspot.com/2013/02/r...g-circuit.html ). That's known as one-shot mode - so on multiple trigger inputs, it only outputs once. But one shot mode only works once - so for the next time the switch closes, it needs to be reset. The easiest/hacky way to do this is run your VCC across the hall switch and run your trigger to ground so it's always "triggered" - but the circuit is getting turned on and off so the beep will only be .55 second maximum length.

http://www.falstad.com/circuit/e-555monostable.html has a neat simulation on one shot - though that one appears to not need retriggering. Are you sure you have a pullup resistor on the trigger? Might want to make that 1/10th the value of your R for the time constant - so if you used a 1k, make it 100 ohm. And toss a 10uf cap inline. Basically, copy the simulation.

kenzo42 07-01-2018 02:44 AM


Originally Posted by gooflophaze (Post 1489216)
Yeap. So what you need is known as one-shot mode. Right now it does a fixed-length output on a single trigger - but when it's holding on an closed trigger it continues to output - which is what I realized when I put the 2nd link in ( https://m8051.blogspot.com/2013/02/r...g-circuit.html ). That's known as one-shot mode - so on multiple trigger inputs, it only outputs once. But one shot mode only works once - so for the next time the switch closes, it needs to be reset. The easiest/hacky way to do this is run your VCC across the hall switch and run your trigger to ground so it's always "triggered" - but the circuit is getting turned on and off so the beep will only be .55 second maximum length.

555 Monostable Multivibrator has a neat simulation on one shot - though that one appears to not need retriggering. Are you sure you have a pullup resistor on the trigger? Might want to make that 1/10th the value of your R for the time constant - so if you used a 1k, make it 100 ohm. And toss a 10uf cap inline. Basically, copy the simulation.

This is probably going to sound dumb but what is "ctl" connected to? Why is "reset" and "ground" missing?

rleete 07-02-2018 10:13 AM

Got my latest scan results. Still leaking.

Which means another operation, another overnight, another 11 hour round trip. Crap.

I leak more than Aidan's car.

Joe Perez 07-02-2018 10:22 AM


Originally Posted by rleete (Post 1489360)
I leak more than Aidan's car.

:bowrofl:

I'm sorry, I know you're dealing with some serious shit, but that really made me LOL.

Good luck, man. If all else fails:


https://cimg4.ibsrv.net/gimg/www.mia...3940602286.png

aidandj 07-02-2018 10:27 AM

At first I was like :sadwavey:

Then I was like :rofl:


All times are GMT -4. The time now is 10:09 PM.


© 2024 MH Sub I, LLC dba Internet Brands