Push Button Start

I have an installer friend who has contacted me and asked that I design a push button start system for a customers 2007 Mitshubishi Eclipse.

I won't be doing the installation, just (possibly) providing the part.  The owner plans to physically remove the mechanical wheel lock (assuming one is present) and is going to use the push button start exclusively to start the vehicle (no key at all will be needed).

The installers idea was to use the OEM keyfob to trigger a 3 minute window that would allow the car to start.when the button was pushed.  This would retain some security, as you would need the OEM keyfob, but at the same time would leave the car useless if the keyfob died, was lost, or destroyed, so not a very practical option.

So, the next option would be to use the 'unlock' actuator signal.  This would trigger the window any time the doors were unlocked (remotely or with the interior switch).  This, to me, creates quite a security risk, as anyone could drive away with the car at pretty much any given time.  If he installs an aftermarket alarm with a starter kill the starter kill could add a layer of protection, but I'm not sure if that is in his plans.

The installer also mentioned using a 'ghost switch' to add a bit more security. 

From a design stand point, I don't want to provide something that isn't secure.  Anyone have any other ideas of how to make a full function push button start secure?

Thank you,

hi

here's a piece https://bulletproofelectronics.com/arseHOME.htm.... looks a bit pricey, but does appear to have ample security provided.

shouldn't need too many modifications to convert to automotive use.

just a thought

mark

Interesting...

The way the Honda S2000 did it for a while (don't know if it still works like this) was you had to insert your key into the center console and just rotate it clockwise to a set position and then you could press your push button to start the car up.  I believe the immobilizer was linked to this center console key barrel to provide the antitheft component.

Could you rig something like that?  Could you rig somethere where the immobilizer system detected the actual key (with transponder chip) in a plastic dish somewhere in the center console --> this would then activate the immobilizer to deactivate and allow the car to start.

Thank you all for the ideas so far, some really cool stuff that I have never seen has been mentioned.

Keep the ideas coming!

The Nissans have several recievers placed throughout the car to actually detect key position (sensors in each door, and in the trunk and I believe one n the middle of the car).

Depending on where the car calculates the actual key to be the push button start will work. 

I'm guessing its both - convenience and 'cool' factor.

The problem I have with simplifying it is if someone knows about the system (word of mouth or whatever) I would hate for someone to smash his window and drive off.

I need to get a hold of the guy and see exactly what the guy has in mind.

Update:

I got the RFID stuff, and now even have keyfob RFID tags that should work great. 

I've also decided on the button that will be controlling the system - it is the stock button out of the new Altimas.  It has a great feel and look to it, and has multiple LEDs that can dictate system status.

Sweet -- did you purchase a bunch of Alty start buttons from nissan or a nissan parts site?

Can you take a pic when you get a chance of your work re: this push button project -- i'm very interested to see it.  How does the RFID work?

I bought two from the local dealership - about $30 a piece, not a bad price considering the quality.  The unfortunate side is that I can't find the plug for it (it is custom) so I had to solder the wires directly to the back.  There are 8 wires total, so I used a cat5 cable.  I'm going to crimp a cat5 connector on the other end and then use a cat5 plug to make it easy to remove.

I've also already ordered new LEDs to convert it from amber to blue lighting.  Hopefully that will go well.

Here is a picture of the push button.  I've replaced the amber LEDs with blue LEDs and I'm pretty happy with the results.

That is EXTREMELY nice work --

Does the button have a starter kill built into the button circuitry --if not, how did you rig the starter kill? (i.e. without remote starting the car, when you get in, register the RFID, and push start .. how does is THEN kill the starter during ignition_on)?

Also .. do you have it so that you just press the button once and it starts the car automatically (i.e. employing the remote starter brain to start the vehicle) or do you have to hold down the start button until the engine has started appropriately?

How does ACC, and IGN work .. do you press the button sequentially? And the hold down for start?

I haven't started on the controls integration yet, so I don't have concrete answers to most of your questions.

I'm planning on building the 'ultimate' system in stages.  Stage 1 is to get the switch wired and working, which I've pretty much done.

Stage 2 will be getting all the lights and functions of the switch to work while still using the OEM key.  This would involve putting the key in the ignition and turning it to the 'ON' position, and then pushing the button to crank.  There won't be any engine monitoring so as long as you hold the button down it will crank.  The control system will be microprocessor controlled and will not only start the engine, but will stop it also (after all, the button is a start/stop button).  I plan to include a digital 'engine running' signal to allow the processor to decide when to start and when to stop.  My thoughts are, on modern fuel injected vehicles, you should be able to reference the fuel pump signal to determine if the car is running. 

Stage 3 will include the addition of the RFID circuitry and a tach reference.  At that point, you'll simply press the button and let it go and it will do the rest.  With the RFID I'm planning on turning on ACC with the RFID signal, and then turning on ignition power and starting the car when you push the button within the allotted time.

I think the timeout will be around 10 minutes, or until the drivers door is opened.  This will give people the ability to listen to the radio, without risking draining the battery.  After the 10 minute period a new 10 minute period can be started by simply removing the RFID tag from the antennas range and then bringing the tag back in to range.  This will allow poeple to simply listen to the radio, and perform other ACC tasks with the vehicle off.

peterubers, I missed your question about the RFID.  The RFID part is actually fairly simple, you can purchase RFID modules that are pretty much self contained readers.  You then interface them in to your project.  You can get them with digital or serial outputs, depending on your goals.  Then, you attach an antenna and learn the tags to the system and you are ready to go.  For this project, I plan to send a 1S signal every time a card is read (whether the car is read for 0.25S or 10,000S only one 1S signal is sent).  The RFID stuff is pretty cool!

KP --

I just drove a buddy's M35x that has the exact same Altima start/stop button.  I realized that you had to depress the brake in order to activate the starter -- i thought this was a very cool feature -- what are your thoughts?

The button looks the same, but it has some small differences (color and function).

I never thought about that, but it could be incorporated.

I was planning on setting it up so that you had to hold the button down for longer then just a 'push' to make it impossible to activate the system by accidently bumping the button.  Requiring the brake to be pushed down would only help eliminate that scenario.

are the LED's inside the button ring going to light up correctly to show acc, on, and lock?     and where are you going to mount this thing at?

as a side note...this is the stuff that dreams are made of,  it makes me wanna quit best buy and open a custom shop only doing crazy stuff like this all day.  it would be weird to actually use my brain instead of just sliding ford radios in and out or windsheild mount sat installs.

just kidding, we get to do some cool stuff there I just love the puzzle of getting one off systems like this to work and become a reality.

Yes, all the lights and functions will work just like OEM.  The controller is based around a 10mhz microprocessor.  I almost have the circuit board layout done that will control the button.

I believe, for now, I'm going to replace my cigarette lighter with the button.  They are very close in size, and my cigarette lighter would be the ideal place.

KPierson wrote:

---

I plan to include a digital 'engine running' signal to allow the processor to decide when to start and when to stop.  My thoughts are, on modern fuel injected vehicles, you should be able to reference the fuel pump signal to determine if the car is running. 

---

Keep in mind that usually there's a 2 second prime when the key is first turned to the 'on' position, and that the fuel pump runs during cranking.. so you could run into a situation where the pump is already running (but the engine is off) when you hit the button.

Try using a ground from one of the injectors; the ECM usually pulses the ground to fire the injector, and you could look for that pulse to determine 'start or stop'. Start would be switch closed without injector pulse, and stop would be injector pulse present and then switch closed. You could also use a trigger from a coil/coil pack.

I use Siemens PLCs.. bit of overkill for this situation, but it's nice having inputs and outputs and making stuff work by changing the software, isn't it?

Jim

The more I think about it, the less likely I'll even need an engine running signal.  I think I'm going to include it as an option, but at this point I think I'm going to build in a manual override to the on/off status of the system.  The button has an 'ON' LED that can be used to indicate to the driver the status that the system 'thinks' the car is in.  If the car is running, but the LED is off the driver can then quickly press and release the button and the state will change.  It would then require a longer press to start and stop the car (Pt >0.5 seconds for instance).

Once the digital input is set I have the ability to look for pulses or a constant signal - it's all in the programming.

My PLC background consists of AB and GE Fanuc 90-30s.  I've never had the opportunity to work with any Siemens stuff.  If you like PLC programming you should check out microcontrollers.  They offer quite a few advantages over PLCs - in size, price, and I/O flexibility.  The programming is a bit more difficult, as it is code based and not ladder logic, but it's pretty basic.

I 'stole' a microprocessor system from my little brother.. haven't had time to work with it yet.

Here's a description of one of the facilities I support..

There's a total of six buildings on the PLC network. Each building runs a GEM-80 PLC; five of the buildings are set up with control panels, and the sixth has a master control panel. No difference in the equipment, just configured differently. On the control panels, the PLC controls CCTV, Intercom, and door access. Four of the five control panels operate about 250 doors, 10 cameras, and around 20 intercom nodes.  The fifth control panel operates approximately 50 doors, 5 cameras, and maybe 30 intercom nodes. The master control panel has group control over each control panel (operate doors as a group), controls all cameras on each control panel, and selected intercom nodes on each control panel. From any control panel you can transfer control to the master control panel (which will also disable that control panel), or you can sieze control from any control panel using the master control panel.

The fire alarm system is also tied into the PLC system.. locally at each control panel the fire panel will signal the control panel information on where the fire is in the building; on the master control panel the resolution is limited to what building has a trouble or alarm on the fire panel system. The master control panel also monitors the fire panels in every building on the facility, regardless if they have  a PLC or not.

There's a bit of wiring involved.

I'd still encourage an 'engine running' signal on the microprocessor you're doing.. it'll help with anti-grind, and makes it better for the driver. Also continue to use the neutral safety circuit.

Regards..

Jim

Interesting stuff.  Our little facility has roughly 15 seperate machines utilizing different AB processors.  Some of the PLCs are about as old as me, and some of them are brand new, so it's interesting keeping them all working.  At the facility I work at we make paper caulking tubes (like the Liquid Nails people use to make speaker enclosures).  We're one of only two factories in the USA that make paper caulk tubes, so uptime is a very high priority.

I just finished the initial circuit board layout.  I included some flexibility on the running input.  I included space to make an optically isolated tach reading or a simple transistorized digital input.  Keep in mind, that in this revision the start button won't actually start the vehicle automatically - I'll have to hold the button down to crank.  In the next stage a simple push of the button (after RFID permission is granted) will completely start the vehicle using a programmable tach signal.

What kind of microprossor did you steal?  I'm a big fan of the Atmel AVR series.  They are cheap, reliable, powerful, and readily accessible.

It's been a while, but I've made a little progress.  I've got a pushbutton start system installed in my 1995 Eagle Talon.  This is the 'test' car before it goes in my '04 G35 twin turbod coupe.

Before:

After:

Hide away control module:

All and all it works pretty good.  There are 2 or 3 bugs in the firmware that need to be ironed out, but I have acomplished the first part of my goal.  The next step is to integrate the RFID portion in the security and adapt the module to act more like a 'remote start' then a 'relay controlled by a pushbutton'.

Update!

I finally acomplished my goal - OEM style pushbutton start in my G35.

If anyone wants to see a video of this working here is a link:

Pushbutton Start System - ~10meg AVI

This system still needs the key to be in the 'ON' position, but I may take it a step farther and implement the RFID 'keyless' start.  Unfortunately, the RFID version would be much more complicated and require almost starting from scratch to design the system to be compeletely independent.

Ya know dude..... I'm not much into that kinda of "don't fix it  if it ain't broke" kinda thing....  BUT THATS REALLY F'n COOL!

Nice work Kevin!

Whats the total investment, $ (minus your time) THEN, what kind of time do you have into this?

I'm not sure on how much it cost me total.  The button itself was $32 or so from Nissan.  The circuit boards aren't cheap, as they are custom made (I did in fact design them myself).  If you order them in low quantities like here they are $6-7 a piece.  The rest of the components weren't that expensive - definately under $100 in parts total (not counting tooling required to make the two harnesses).

Time wise, that is another story.  I designed everything myself, so there was very significant time invested.  I started this project with a ton of research before I came up with a basic hardware layout, I then designed the circuit board around my hardware design.  Once I had the hardware I started to write the program for the 8mhz processor.  I'm still not actually done with the programming, but the program is to the point that it works extremely reliably and provides the basic functions I was looking for.  The next step is to add 'tach learning' and 'push and release' start which will allow the vehicle to start without holding the button down - basically like a remote start.  In the very long run, I would like to integrate the OEM keyfob to the circuit so I can make it peroform remote start functions as well.

Howie, thanks for the compliments.  I will try to post some pictures of the 'finished' circuit board that has the customary green solder mask and is silk screened.  They look much more 'professional' when they are finished.

There is no 'value' in it.  This is purely an aesthetic mod.

The original plan was to design and build a completely keyless system.  This is a step in that direction.

usabuilt wrote:

---

whats the reason for the pushbutton start? Its seems highly illogical to me as to why you would want to push a button to start your car, especially since you can still start it by using your key?

Nice work, but I see no value in it.

---

profuse007 wrote:

---

Its part of the car culture. I am not into those things but a lot of people(young crowd) do like those modification. There is a sense of value if you enter your "show car" into an event and points maybe added toward the electronic category.

Whats more practical of this whole push-button start idea is to use a DEI RS add-on w/ the nissan, toyota, or honda button. Not trying to bash KPeirson's genuine accomplishment but thats the most effective way, both cost and time, for anyone out there.

usabuilt wrote:

---

whats the reason for the pushbutton start? Its seems highly illogical to me as to why you would want to push a button to start your car, especially since you can still start it by using your key?

Nice work, but I see no value in it.

---

---

I thought about the DEI remote start with a (-) activation input.  However, the 5 second delay would be a bit annoying and it lacks any ability to control the status LEDs.  It also doesn't provide a way to shut the car off.  So, even using a remote start platform there would be more to add to make it work.  From my perspective, with my abilities, it made more sense to just build the entire thing from scratch.  There isn't much complexity in a remote start system, especially compared to everything else required to make this work.

It's kind of funny, I started this project for an installer friend to install in a customers car, and ended up putting it in my car instead.  My friend is no longer installing, so he didn't have a need for it.  I guess it's more then just an aesthic mod to me, its a sign of acomplishment and a 'trophey' for all the hours I put in to it.

Once the car is drivable (ie remote start disabled) it provides no way of shutting the car off.  You can not drive a DEI remote start car with the remote start engaged (tach cut off and time cut off).  Therefore, once you are driving the car and the timer expires there is then no way to use the added button to shut the vehicle off.

If you are using the Honda S2000 button the DEI part would work great (engine start only, no indicators).  But, when using an engine start/stop button with indicators I don't think it really serves a purpose in the overall picture.

Here are the pictures of the final module right before installation:

fbird08 wrote:

---

Where do you go to get modules like that built for you? It has a great finished look to it.

---

The enclosure is made by Hammond Manufacturing - a very high quality on the shelf enclosure.  I purchase them through Mouser.com (good guess profuse!)

I use two different companies for circuit boards, one place called www.PCBPro.com and the other is called www.4PCB.com.  I believe that 4PCB made the prototype board in the first pictures and PCBPro made the finished board above.  In my experiance, 4PCB is a great place for prototype low quantity boards (<10).  PCBPro seems to be a bit better with bigger production runs.

You can get circuit board layout software for free on the 'net, or you can spend a ton of money and get a nice package.  I use a free program from www.freepcb.com.  It doesn't have any advanced features, but the price is right and my boards are small enough that I can manually route everything with minimal issues.

The basic steps in a process like this is to first find an enclosure you want to use for the final product.  You can then create the circuit board layout in the software package that will fit the enclosure.  You then add all the components you need and make all the copper traces.  Next, you Email the file created by the software package to the company that you pick to make the boards.  For a small prototype run you can expect to have the boards within a week.  For a bigger production run it can take up to 5 weeks (depends on how much $$$ you want to spend.)

No Ph.D here, I only have a bachelors degree (in electronics) and a lot of motivation and willingness to learn about new things.  I learned all about making circuit boards the hard, expensive way - trial and error.  The first batch I ever ordered - 100 boards - were pretty worthless because I didn't lay my ground paths out correctly.  Making them manually truely is an art, and you can spend hours upon hours upon hours perfecting a board.

Yeah, the actual starting part isn't hard.

The challanges from this particular module were:

Stopping the car with the push button
Only allowing the starter to crank when the vehicle was off
Allow all three status LEDs to work as designed
Use a short delay to prevent accidental start/stops

The actual starter output is just a 500mA output driving a relay, just like a typical remote start.  Since the key is needed to be turned in the ignition to defeat the mechanical wheel lock I left out the ignition output.  I could have added it and then the key wouldn't need to be in the 'on' position, but you are already putting it in and turning it so there really isn't a point that I could see.

Er, it's Kevin!

Not sure why, but many, many people call me Ken, which is actually my Father's name!

I'm now in the process of updated the system (already).

I got to test drive an '07 G35 coupe and that was the first time I ever started an OEM pushbutton vehicle.

To make my system more OEM like I am adding a brake wire input and am going to set the system up to automatically start the car when you press and hold the brake and press and release the start button.  It will use a tach monitoring system for now to start the car.  If that doesn't go well I might just make it voltage monitoring.

Kevin, have you looked into the way the mechanical lock is overridden on the late-model Nissans? I can't imagine that it's much more than a solenoid which is actuated by the RF starting system.

Incidentally, where'd you go to school? I'm looking at master's programs in EE/ECE and would love to go somewhere that does a lot of automotive tech.

Cool beans...I'm basically self-taught as far as electronics goes...have a bachelor's in Business/Economics, so it's a leap to convince some schools that I should be in their graduate EE/ECE programs. Anyway...

What have you looked into as far as an RF system on your project here? Would something like a building access system work? I checked into those but it seems like overkill to have a system designed with 8 receivers and hundreds of potential transmitters just to enable car starting.

I'm assuming you're talking RFID (building access).

I was planning on going with RFID but I can't find a system with enough range at a decent price to make it feasable.

There are many, many ways to acomplish this, including bluetooth, zigbee, RFID, RF, and the list goes on.

The last system I was looking at was RF based.  It has great range, but is battery based.  That seems to be the biggest drawback so far - after 5 years or so you won't be able to start your car.  Each of the above systems have pros and cons, and I just haven't committed to anything yet.

KPierson wrote:

---

I'm assuming you're talking RFID (building access).

I was planning on going with RFID but I can't find a system with enough range at a decent price to make it feasable.

There are many, many ways to acomplish this, including bluetooth, zigbee, RFID, RF, and the list goes on.

The last system I was looking at was RF based. It has great range, but is battery based. That seems to be the biggest drawback so far - after 5 years or so you won't be able to start your car. Each of the above systems have pros and cons, and I just haven't committed to anything yet.

---

I was able to finish another update this weekend.  The car now starts itself upon pushing and releasing the button.  There is a "tach learn" routine like most remote start units have and then the module monitors the engine RPM.  After it has reached x% of the programmed RPM value it kills the starter output. 

Here is a video of it in action:

https://www.kptechnologies.com/movies/KPtechPBStart.mp4

Note that the key is in the ignition and switched to the "ON" position already

It will take a few more days to get it working perfectly. It is having some cold start issues, but I should be able to smooth that out pretty easily, it will just take some time.