let`s take control of the SWAS

VoodooJef

VoodooJef

My friends call me Captain Zen
Greenie N00B Member
I did a lot of reading yesterday about SWAS and the various ways people have been defeating it, and it all ends with dash lights on and potential abs/dsc defeat. I want neither of those. SO I dug into it a little more and got the sensor voltage defaults (ie; the voltage that tells the system where the wheel is pointed). Centered puts out .25 to .75 volts, and anything left or right does .75 to 4.75 volts. The very first thing I thought was "well f**k, why not just use a simple integrated voltage regulator to cap the output voltage at .75 volts?". Yes, it`ll work. It`ll work like a champ and completely solve all of the headaches associated with SWAS.

"So, Jef, what`s the catch? If there wasn`t one you`d have surely put that into action."

There is a catch, but only a little one. I can`t find anything lower than a 1.5 volt regulator. Given how sensitive the system is I don`t think it`d do any good allowing more than the .75 volts through. So I need one of two things to move forward: either a .75 volt regulator source OR evidence that the neutering is triggered at a higher voltage (the system giving allowance for X degrees off center). Again, referencing back to how sensitive the system is I`m betting that there is no wiggle room, but we can hope.

With that no on the table, what say you? Can we source the correct regulator?
 
my only concern is current. Any time we drop voltage we raise current, and that could go badly. The signal from the SWAS goes to the PCM and sending too much current....RIP. But how much is too much in this case, and how much are we dealing with, and how much can the PCM read?
 
My guess is considering the center input is designed to sense 0-5v input anything in that range would not cause damage due to internal resistance of the pcm

Sent from my SM-G920V using Tapatalk
 
I did a lot of research/experimentation on this and did a long write-up with lots of updates. It can be found here, if people want to know why disabling it is really desirable for lots of folks. BTW, SWAS unplugging doesn't disable ABS in my experience; DSC is a different story: https://mazdaspeeds.org/index.php?t...swas-disable-features-on-2010-test-mule.4004/

Having said this, a "fix" to remove the two warning lights would be very nice.
[doublepost=1493484916][/doublepost]If we're increasing resistance, it'll actually lower the current of the system (V/R) = I (see below), so long as we're using resistors. Resistors have been used to "trick" the ECU to think the VTCS is still operational after it's been removed on ported IMs, so it could work here too:

upload_2017-4-29_11-52-0.png

The trick is to keep the voltage in a specific range so that the system never thinks the steering wheel isn't pointing straight ahead. I'm thinking that just strong enough resistor(s) would do the trick because voltage is a function of ohms (resistance) x current (amps). Personally, I'd aim for right in the middle of the .25v-.75v range to account for voltage fluctuations in the car's overall power system, so what about .50v instead?

The question is, are we only talking about one wire this will need to be done to? Also, does the ECU/PCM have a failsafe function that'll trip a CEL if this system is messed with in this manner, especially if a resistor drops the voltage too low? And what about heat dissipation? I agree a regulator seems the way to go. Hunting for the right one too...
[doublepost=1493492631][/doublepost]Dang, it's hard to find a regulator with a 0-5v input and a 500mv output! Argh...
 
Last edited:
Yep, the absolute simplest solution that would require nearly no effort at all would be a 5v - 0.5v step down regulator. I`ll stop bt Mouser on monday and see what they suggest
 
VoodooJef said:
Yep, the absolute simplest solution that would require nearly no effort at all would be a 5v - 0.5v step down regulator. I`ll stop bt Mouser on monday and see what they suggest
Click to expand...
Grabbed what I hope to be a useful page from the factory service manual:
upload_2017-4-29_15-24-22.png
It appears that Terminals A, C, E, and D should be limited to .50v. I say .50v because electrical components have variances of a few percent most of the time. So if we shoot for .25v or .75v and the variance takes it out of the range, SWAS may still interfere. That's my rationale regarding regulating the voltage to .50v (500mv).

But according to this, we may only have to worry about Terminal A (the light blue wire - the only one you address in installing a switch):
http://www.mazdaspeedforums.org/forum/f429/how-install-switch-swas-enable-disable-102406/
upload_2017-4-29_15-39-50.png
 
Last edited:
I always thought cutting one of the wires was to cause a fault and totally disable the SWAS.
From what I see above, I think to truly fool it and see that there is no angle change at all you need A & E to be 0.25-0.75V and C & D to be at 4-4.75V constantly.
 
Yes, cutting the wire disables the whole thing and puts up a fault code. The diagram seems to imply that each terminal is responsible for detecting progressive degrees of wheel angle. A seemingly more reliable system, but also more sensitive if they all trigger the same reaction (which I suspect they do, given how touchy the system seems to be for some folks). And yep, looks like A/E need to register the .5v and C/D need 4.5v. It also appears that we could leave terminal D intact and keep full swas function at a true steering angle over 3 degrees should we want that. A and E would be easy to wire in an inline regulator. C and D would be easiest to feed them a steady regulated 4.5v rather than try to step up the .25v. I`ll go by mouser tomorrow and see what they can offer :)
 
Actually...what do the other two wires on the connector do? Power and ground?

I'm kinda thinking you can just fool it with a properly connected voltage divider.
 
Last edited:
VoodooJef said:
Yes, cutting the wire disables the whole thing and puts up a fault code. The diagram seems to imply that each terminal is responsible for detecting progressive degrees of wheel angle. A seemingly more reliable system, but also more sensitive if they all trigger the same reaction (which I suspect they do, given how touchy the system seems to be for some folks). And yep, looks like A/E need to register the .5v and C/D need 4.5v. It also appears that we could leave terminal D intact and keep full swas function at a true steering angle over 3 degrees should we want that. A and E would be easy to wire in an inline regulator. C and D would be easiest to feed them a steady regulated 4.5v rather than try to step up the .25v. I`ll go by mouser tomorrow and see what they can offer :)
Click to expand...
I hear ya. My thing is that if a switch can entirely disable SWAS interference by placing it on Terminal A alone (light blue wire), then wouldn't controlling the voltage on said wire to, say 500mv (nice, middle number to account for voltage/hardware variance) be sufficient? Seems like only 1 wire needs messed with.

I also agree a regulator is the way to go. We just need to know which wires need a regulator (if more than 1), and which wire(s) needs to be in the middle of what voltage range(s) always for SWAS to not interfere.
 
Last edited:
Hmm...I looked at the service manual more there is a steering initialization process the BCM goes through when there is a reset/loss of power. It seems that BCM is looking at a series a pulses depending on the angle and direction the wheel is turned. Any erratic pulses will cause a fault. Someone prove me wrong. lol
 
Ok so SWAS... It is used by a lot of shit. You cannot do anything to disable one system and leave the others fine ELECTRICALLY. It can only be done thru code. SO the SWAS is a multi position sensor that takes it data and brodcasts it on the CAN network when requested. Each system DSC,ABS,TORQUE LIMITING ECT.... when it needs this data it simply looks on the can network. Disableing the SWAS sensor kills the data feed period. This affects all systems. the only way to do this without code is to get a computer and put it between the can network and the SWAS can driver. The can bus triple rings a bell. it could be programmed to feed bogus data to the pid the torque limiting programs used to request Steering Wheel Angle.

After all of this TBH a switch under the Colum cover and a light one the dash for no swas is fine for me. I switch it on when it rains or family is in the car. I would be more interested in a EGR or O2 sim electronics to pass emissions!
 
Se now, there is some info I was looking for. It had just barely crossed my mind that SWAS could also be on the Canbus, in which case it is correct that theres no simple solution other than complete disabling and fault lights.
 
Had my hopes up. But, we're back to exactly what I've done for years: unplugging and just getting so used to the lights being on that you don't even really notice them any more. And for me, the driving improvement is such a night/day difference, I'll always run unplugged.
 
I cut the signal wire and spliced it into a switch that i mounted on the side of the column. I leave it off most of the time, but its nice to leave on when its slippery out or family in the car. Best of both worlds. When you go SWAS off to on, you have to restart the car for it to take effect and make the light go out.

Sent from my SM-G935V using Tapatalk
 
msayre91 said:
I cut the signal wire and spliced it into a switch that i mounted on the side of the column. I leave it off most of the time, but its nice to leave on when its slippery out or family in the car. Best of both worlds. When you go SWAS off to on, you have to restart the car for it to take effect and make the light go out.

Sent from my SM-G935V using Tapatalk
Click to expand...

Anyone have pics or a video of the switch install process? Seems common enough that it should be out there somewhere. Looks like there were some threads in mazdaspeedforums but they're all dead..

Edit: a helpful link, seems like it doesn't matter which wire is cut for the switch.

http://www.torontomazda3.ca/forum/archive/index.php/t-78634.html
 
Last edited:
This is an old thread but for discussion why didnt anyone bring up using a potentiometer (variable resistor)

I used to run in my Chrysler Conquest/Mitsubishi Starion a MAF Translator and SAFC 2 combo. Basically the MAFT allowed me to use a GM MAF sensor and then converted the GM signal to a Karman Vortex Style signal. The MAFT sent that into the SAFC as a KV signal and the SAFC allowed for more fine tuning over all RPM ranges vs the 3 settings I had for the MAFT.

The reason I bring that up is an SAFC required a tach signal but I couldn't ever get it to work right unless I used a potentiometer to "bleed" tach voltage off going to the SAFC.

It seems to me that if a VR is used in the SWAS signal circuit then it could trick the ECU into seeing less voltage. I get that people talked about using a resistor but trying to use a set 1k, 10k, 100k ohm resistor may not get the voltage right where it needs to be. It would have to be dialed to nail that perfect voltage. Also unless running a basic circuit board with resistors soldered onto it to achieve the set value needed then you'll have a long wire with resistors stacked in line.

I picked up a VR from RadioShack years ago for about 3 dollars. It was a 3 terminal resistor with a dial for adjustment.

If someone here has an image of the SWAS wiring diagram I could figure out if that would work. I dont know how many wires go to the ECU and how many go to ABS etc. It was stated it talks on a CAN network and I am not sure if anything like this could work but it would be nice to defeat the signal at the ECU but retain SWAS control for ABS.
 
You must log in or register to reply here.
Back
Top