Miata Turbo Forum - Boost cars, acquire cats.

Miata Turbo Forum - Boost cars, acquire cats. (https://www.miataturbo.net/)
-   MEGAsquirt (https://www.miataturbo.net/megasquirt-18/)
-   -   Another wideband to tuner studio problem (https://www.miataturbo.net/megasquirt-18/another-wideband-tuner-studio-problem-91181/)

Zammish24 11-11-2016 06:36 AM

Another wideband to tuner studio problem
 
3 Attachment(s)
hello everybody,
Im hoping someone can help me as my AFR gauge on tunerstudio is very messy and dosen't match my innovate gauge, ive already tried a few solutions on other posts but not had any luck
Ive connected the wideband from the yellow wire directly to the wiring inside the ms1pnp, the wideband uses the megasquirt power supply (white with red stripe) and the ground is attached to the engine block (next to injectors). i took it to get mapped thinking the difference is little however when the mapping company tried to tune it told me the wideband is completely out.
on the attached file the dotted line shows the AFR ratio fluctuating during boost. Is there anything i can do? please help its my daily driver :(

my setup:
MS1PNP v3.57 board
550cc High impedence RX8 injectors
Required fuel is 5.9

Chiburbian 11-11-2016 08:13 AM


Originally Posted by Zammish24 (Post 1373683)
hello everybody,
Im hoping someone can help me as my AFR gauge on tunerstudio is very messy and dosen't match my innovate gauge, ive already tried a few solutions on other posts but not had any luck
Ive connected the wideband from the yellow wire directly to the wiring inside the ms1pnp, the wideband uses the megasquirt power supply (white with red stripe) and the ground is attached to the engine block (next to injectors). i took it to get mapped thinking the difference is little however when the mapping company tried to tune it told me the wideband is completely out.

Out as in non-functional or out as in not functioning properly?


on the attached file the dotted line shows the AFR ratio fluctuating during boost. Is there anything i can do? please help its my daily driver :(

my setup:
MS1PNP v3.57 board
550cc High impedence RX8 injectors
Required fuel is 5.9
Change your ground point to the same point that the ECU is grounded OR directly to the negative battery terminal. Scott Clark from DIYautotune recommends a direct ground to the negative battery terminal for the ECU and the AFR gauge. His source for his opinion is the Bosch automotive EFI book.

ALSO - here is your problem:

The analog outputs can be programmed, see chapter titled “Programming the Analog Outputs”. The default analog outputs are as follows: Analog output one (yellow) is 1.1V = 14 AFR and .1V = 15 AFR. Analog output two (brown) is setup as 0V = 7.35 AFR and 5V = 22.39 AFR. The analog outputs can be programmed with the provided software
Change your wire to brown for step 1.

Zammish24 11-11-2016 09:44 AM

https://youtu.be/ZG9BZnaobEc
it's out but only just and it's not steady at all. I'm not sure why the video quality is bad I'll try upload a higher res one shortly

Chiburbian 11-11-2016 09:46 AM

Disregard... The MTX-L uses backwards wiring setup... here from the manual for clarity:


4 Optionally, the YELLOW (Analog out 1) and/or BROWN (Analog out 2) can be connected to the analog inputs of other devices such as data loggers or ECUs. If either one or both of these wires are not being used isolate and tape the wire(s) out of the way. The default analog outputs are as follows: Analog output one (yellow) is 0V = 7.35 AFR and 5V = 22.39 AFR. Analog output two (brown) is 1.1V = 14 AFR and .1V = 15 AFR.
Check your grounds.

Also- looking at your video your afrs are pretty close to what the gauge says. What happens if you lean out your mixture? Can you pull fuel and get the AFR gauge and tuner studio to give you about 14.7:1 indicated?

You are not that far off.

Zammish24 11-11-2016 09:51 AM

I was told that I should connect the ground to sensor ground on the ecu but was also told it can damage the ms1,

deezums 11-11-2016 10:03 AM

It's impossible to entirely remove the voltage offset issue from a MTX-L, it has horribly shitty grounding.

What you can do is tune it a bit, using the custom wideband config settings. You'll find the voltage offset isn't linear though, it's exponential. What that means is you'll have to choose a small range to remain mostly accurate, everything outside this range will quickly become more and more incorrect the further from the accurate range it gets. You'd like to be able to measure from 11-16AFR to tune a turbo car, everything outside that range is pretty much useless. Try and center the accurate readings on these values, or as close as you can get.

There's a piece of software from innovate called lmprogrammer. You can use that to output some fixed voltages from the wideband to the megasquirt, then you can do math to get the voltage offset on the low and high side. Much easier than just shifting the values and watching the results.

Zammish24 11-11-2016 10:17 AM

I was playing aroung with LM programmer this morning and tried to follow the instructions on how to calibrate the AFR table on MX-5 Unleashed - Mega Squirt PNP installation instruction for Eunos or miata
i followed this and still there was alot of fluctuation between the gauge, if anything it made it worse so I did a factory reset on LMprogrammer and set the EGO switch point to 2.509, its still fluctuating alot but the gauge seems to be getting closer.
Wheres the best place to ground the MTX-l?
Also my gauge is mounted right beside the radio, would that affect it at all?
Also on my dyno printout, is there any reason why the afrs are fluctuating very badly?

deezums 11-11-2016 10:22 AM

1 Attachment(s)
I found the best place to ground it, and power it, was from the battery with a relay. Pretty much like the manual suggests.

Is your car a 1.6? Can you unplug the megasquirt and check the integrity of the signal/sensor ground circuit by measuring continuity between it and the frame/engine block?

Attachment 230615

Zammish24 11-11-2016 11:11 AM

i have just done a continuity check and there is continuity between the sensor grounds and framework

Zammish24 11-11-2016 11:30 AM

2 Attachment(s)
this is where my wideband is grounded

deezums 11-11-2016 11:31 AM

1 Attachment(s)
My 1.6 had that too.

There's a group of four wires near the cas on the back of the head, for some reason mazda grounded the sensor ground circuit here. You can see it on page 4 here. Edit: Pretty much right where you've got the wideband grounded, lol
http://www.mellens.net/mazda/Mazda-M...991_wiring.pdf

Attachment 230614

It's a black and green striped wire in along with three black wires. I suggest taking them all off the block and keep checking continuity so you know you're clipping the right one, but I'm pretty sure your car will be the same. Once you remove this wire from the head you should not see continuity between ground circuits, so long as the megasquirt is unplugged.

When I had a botched sensor ground circuit my coolant, tps and other inputs would jump around like mad. With your grounding the wideband so close to this sensor ground, I wouldn't be surprised to find it's responsible for a lot of that noise.

Zammish24 11-11-2016 11:39 AM

thank you for your reply, ive put the car back together as i need it for work tomorrow. originally the sensor was grounded on the chassis (dashboard framework) and powered by the cigarette power supply but after some research i got the power directly from the ecu and moved the ground to the engine head and it had the same amount of fluctuation back then as it has now so really is hasnt helped at all.
the mapping company told me it fluctuates its worst when its boosted, and advised i try a rising rate fuel pressure regulator as im running a factory fuel pump, does this sound right?

Chiburbian 11-11-2016 11:48 AM

Last iteration of my build I grounded the ECU and the wideband to the following points: (hopefully the embed works)
https://www.miataturbo.net/attachmen...ound-point-jpg
https://www.miataturbo.net/attachmen...round-back-jpg

Zammish24 11-13-2016 08:51 AM

1 Attachment(s)
okay so i got another data log, it looks like Bcduty3 is causing some kind of problem. Can someone please check this for me?

oops wrong post! my bad lol


All times are GMT -4. The time now is 01:19 PM.


© 2024 MH Sub I, LLC dba Internet Brands