1
\$\begingroup\$

Please excuse what is likely to be a very basic question, but I am an ME, and this stuff is somewhat magic to me. First, I'll provide a minimal, generic question, and then, for those interested, I'll provide all of the application details.

I have a signal wire, and when a switch is flipped from its normally closed position (along with a normally closed safety switch), I want the signal to be broken. So, any condition where any switch is in its resting position should allow signal flow.

I intend to spit the signal line into 2 parallel lines, with a switch in the middle of each branch.

My question is this, can I reasonably expect the signal integrity to remain adequate after doing this, even if the branches aren't the same length, and/or if the length is nominally the same? Also, can I expect the signal integrity to remain adequate when signal is flowing through both switches, and one switch gets opened, or when signal is flowing through 1 switch, and the other gets closed?

I am not concerned about the transient response going from no signal flow to one or both switches being engaged, which will make more sense for those who care to read the application, described below.

The application is for a Jeep (2015 Wrangler). The newer Jeeps have an ESC system that remains intrusive offroad, even when "off". This is less true for a stock Jeep, but some aftermarket mods make the system a problem. There is a common mod that involves a momentary, NC switch which is placed in an accelerometer signal wire. Interrupting that signal kills the ESC system. Earlier, I said I wasnt worried about the transients that could occur from signal restoration, and that is because restoration of ESC function requires the ignition to be cycled after signal is restored. What I want to do differently from most Jeep owners is to have a main safety switch, which will need to be opened in order for other switches to do anything, to prevent accidental engagement. My concern is in the signal integrity at steady-state single and dual switch closure, considering a potential difference in lead lengths, and also in the potential for there to be noise in a 2-1 or 1-2 switch engagement.

Also noteworthy is that I intend to make the main safety switch a DPDT switch, which will also be used to switch ground (NO configuration) for other nearby circuits. I'm not sure what cross-contamination I may experience from this, but I think it will be minimal, as the other circuit is just signal ground for 2 relays.

Edit...more details:

I've added some drawings and photos to help clarify. The diagrams are my attempt at illustrating the factory circuit, the most common modification, and what I want to do. I've also attached a photo of the actual switch panel in my Jeep. The switch labeled main will be a DPDT switch, where the normally closed terminals will be used for the circuit in the diagram. There is a blank spot for a switch (where the little plastic cap is), where I will put an SPDT switch, and again, I'll use the normally closed terminals. The idea is that ESC will be disabled by "engaging" (opening) both switches, as seen with 2 switches in one of the attached photos (where we pretend the top switch is the one I'll be adding). My concern is that this will introduce noise in the signal, and cause the stability control system to behave erratically when it's on, and/or when only one switch is being opened/closed.

enter image description hereenter image description hereenter image description hereenter image description here

\$\endgroup\$
3
  • 1
    \$\begingroup\$ It's unclear what you are trying to describe re your requirements. Try a little harder. In particular "I want the signal to be broken" is not registering and neither is the safety switch you mentioned. \$\endgroup\$ – Andy aka Oct 11 '20 at 16:51
  • \$\begingroup\$ Hey Andy. I've added some photos and additional clarification to the question. I hope this helps. \$\endgroup\$ – jdavis Oct 11 '20 at 17:43
  • \$\begingroup\$ There’s a good chance of signal to noise ratio being degraded if you significantly modify the type of shielding on the wire to unshielded hookup wire. Is this coax or twin ax? Most Accelerometers are in console For COG reasons. 3 wires means +5,0,signal 0 to 5V. what do you really need? external g monitor? Or dual sensor switch? There’s no need to break ground. Show whole schematic of This part . Noise on accelerometer can be filtered by cap but may add delay,. Open circuit with a pull-up R on ESC detects an open as a collision >5g. I see problems with your plan. \$\endgroup\$ – Tony Stewart Sunnyskyguy EE75 Oct 11 '20 at 18:59
0
\$\begingroup\$

If I understand correctly you're trying to use a DPDT switch a single wire signal from the accelerometer between 2 paths: one which goes directly to the ESC input and one which goes through a momentary NC switch before going to the input.

Given that the current aftermarket mods work through 1 switch, it's reasonable to assume the signal quality through a second switch would be adequate for the ESC to receive. However, based on the description of the current aftermarket method for disabling the ESC, you'll likely still kill the ESC when you throw your safety switch as this will still cause a temporary disconnect from the ESC, and thus shut it down. If you were to consider a make-before-break switch, you'll likely run into signal errors if the the accelerometer output is digital (which it likely is) due to propagation delay differences if the two branches aren't EXACTLY equal which will also likely kill your ESC.

My recommendation would be to find a mechanical safety method, such as using a push-button or toggle switch with a cover.

\$\endgroup\$
1
  • \$\begingroup\$ Thanks Phy51x. That's what I was afraid of. I think I will make it a standalone switch, the way most people do. I was driven mostly by OCD to make the main switch disable the ESC killswitch, since it disables the 2 locker switches on the same panel. I suppose I'll relabel it "L.MAIN" to appease my OCD. \$\endgroup\$ – jdavis Oct 11 '20 at 18:07

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.