Golf mk7: Auto Start/Stop disable switch - who wants to help?

[Ralfi]

Respect to those participating in this. Taking a risk in tweaking with your new baby, for the benefit of others. Good stuff.

[AJW]

AJW: just had another look at your (excellent) design. It looks like a dual mono-shot configuration with the first timer output being capacitor coupled to trigger the second mono (albeit I'm a bit confused about the need for the 10k pull-up resistor from pin 5).

I am impressed at the extent to which you have conditioned the trigger signal to ensure reliable start-up (both for pin 6 and for the reset pin 10) - and I thought that I was being ultra-conservative LOL! You have adopted a wise approach given the lack of information and the early days in our understanding of the car's electrical environment for the S/S kill switch. It might be possible to refine both our designs in subsequent versions.

The PCB screen and track layout clearly contemplate the solid state FET relay and the current limiting resistor, where as your wiring diagram suggests an electromechanical device. Will you be able to overlay the coil-relay over the PCB tracks on your PCB design, or will you put-in a second set of pads?

The resistor is 100K and I was using it as a pull up resistor, I'll have a look on my protoboard design and cut it off it mightn't be needed.
I was noticing that the second stage timer circuit (the one that drives the relay) could be triggered about 1 in 20 times when I applied a 12V Plug Pack directly to power the PCB.
If I used a switch inline with the power and turned it off/on it always worked.
The overboard design was also to help with this random triggering event and not knowing exactly how the power would come on when the car starts. In a lot of older cars starting the car also dips the supply rail enough that things will reset.

Yes there's a difference in my circuit diagram and PCB layout.
I started with a relay then decided I wanted to use SMD so went looking for a Solid State relay.
On my Protoboard I'm using a Photomos relay that's using a LED (IR?) to make a Mosfet conduct.
They're only good for low current and lower voltages but it should work in this application.
So on the PCB layout I'm using a G3VM-61G1 - normally open solid state relay.

I'm hoping to have my car this Friday so I can check some things out with my design.

Thanks Mattaus for showing how to pull apart the switch cluster, doesn't look too bad.

[DV52]

I was noticing that the second stage timer circuit (the one that drives the relay) could be triggered about 1 in 20 times when I applied a 12V Plug Pack directly to power the PCB.
If I used a switch inline with the power and turned it off/on it always worked..

AJW: Not sure what type of 12V Plug pack that you used, but I've found in the past that the performance of these packs are highly susceptible to the load on switch-on. I suspect that it has to do with the quality of the internal switched mode design, but I've found that some have a notorious high ripple that is injected onto the voltage bus if they are loaded on switch-on. I wouldn't be worried about this aspect of your design.

I considered filtering the power supply pin on my ciruit, but the problem is that for the device to work properly, both our switches need the voltage at the Vcc pin to accurately follow the Terminal 15 rail. If we beef-up or filter the power supply ciruit too much, we run the risk of missing multiple quick starts of the ignition(I think)
Don

[AJW]

Just looked I'd removed the 100K pull up resistor on my Protoboard design anyway.
I guess that's one less component, I'll need to change my PCB layout now.

I think I also saw the second stage random trigger issue on a regulated bench supply.
At least with the design I've got now it seems fairly robust.

I was also looking at supply filtering but realised like you that it stuffs up the power on triggering.
I have got a 100nF cap as on the supply rails for the NE556 - this will also reduce voltage spikes when the output transistors of the timer change states.
There's some good 555 info at this website that has some good reference material.
555 Timer Circuits

[DV52]

Just looked I'd removed the 100K pull up resistor on my Protoboard design anyway.
I guess that's one less component, I'll need to change my PCB layout now.

I think I also saw the second stage random trigger issue on a regulated bench supply.
At least with the design I've got now it seems fairly robust.

I was also looking at supply filtering but realised like you that it stuffs up the power on triggering.
I have got a 100nF cap as on the supply rails for the NE556 - this will also reduce voltage spikes when the output transistors of the timer change states.
There's some good 555 info at this website that has some good reference material.
555 Timer Circuits

AJW: Great news about the 100K resistor -my philosophy these days when designing circuits is that every component on a PCB needs to earn its place there - there is no room for lazy/idle components !!

Your design inspired me to add a couple of free wheeling diodes; to supress the possibility of negative spikes from the relay coil (getting into the monostable oscillator) and to filter out noise on the 12V rail. I'll update the circuit diagram on my first post later, but I've included the extra components in the physical layout for Mattaus's benefit when he puts the components together. You might want to think about doing something similar to your design.

Yes, filtering the power supply will come at the cost of making the S/S kill device more sluggish. Not sure if a 100nF capacitor will do much to stabilise the supply rail on your board, but I don't think that it will do any harm if you leave it in-circuit. There may need to be some tweaking to my design in respect of power supply conditioning, but I'll wait for feedback from Mattaus.

With my circuit and looking at your circuit too, I've no doubt that they will operate as intended, The real measure of success though will be the long-term stability of the devices with events like the air conditioner switching on/off (causing the 12V rail to fluctuate) and human practices like switching the ignition on/off quickly. I guess a few hiccups can be tolerated, but systemic misfiring is the real issue with both our designs (IMO). But let's wait until the devices are actually installed and trialled

Thanks for the 555 link - good resource!
Cheers
Don

[mattaus]

OK, so taking the centre console apart a second time was a piece of cake. Literally took me about 30 seconds. I checked; nothing is broken lol.

I fashioned up some super professional needle probes:



First problem I discovered is that the wires VW use are NOT easily pierced. Like, it's impossible without tearing them apart or stabbing yourself in the finger. Thankfully I managed to get a bit more slack out of the wiring loom and just shoved the needles into the back of the plug itself. It's actually a better way to do it because there is zero damage being done

Here's a video of my efforts:

https://youtu.be/X17jhW6QAtU

Yes, I had the leads connected backwards. Stupid, but the DMM was actually upside down when I hooked them up and I wasn't thinking straight. Also, I kept calling the bloody button "traction control". You know it isn't, I know it isn't. Let's just put it down to a Monday spent in 6 hours worth of meetings that some bright spark (not me!) decided should start at 8:30am. Oh, and when I was uming and ahing about what to call the blue and black wire, it was because I thought there was a more technical name for it. I guess the best I can come up with is S/S switch wire

Anyway, as you can see the voltage I was getting across the switch itself was only 2V. Dropped to zero as expected. The voltage from T15 to GND is 12V when ancillary power is on, and 14.1V when the engine is running. Also as expected. There was a delay in the T15 voltage jumping from 12V to 14V though; probably around 5 seconds. It doesn't budge when you press the S/S button. No video of that sorry.

So all-in-all, nothing we didn't expect?

[DV52]

OK, so taking the centre console apart a second time was a piece of cake. Literally took me about 30 seconds. I checked; nothing is broken lol.

I fashioned up some super professional needle probes:

First problem I discovered is that the wires VW use are NOT easily pierced. Like, it's impossible without tearing them apart or stabbing yourself in the finger. Thankfully I managed to get a bit more slack out of the wiring loom and just shoved the needles into the back of the plug itself. It's actually a better way to do it because there is zero damage being done

Here's a video of my efforts:

https://youtu.be/X17jhW6QAtU

Yes, I had the leads connected backwards. Stupid, but the DMM was actually upside down when I hooked them up and I wasn't thinking straight. Also, I kept calling the bloody button "traction control". You know it isn't, I know it isn't. Let's just put it down to a Monday spent in 6 hours worth of meetings that some bright spark (not me!) decided should start at 8:30am. Oh, and when I was uming and ahing about what to call the blue and black wire, it was because I thought there was a more technical name for it. I guess the best I can come up with is S/S switch wire

Anyway, as you can see the voltage I was getting across the switch itself was only 2V. Dropped to zero as expected. The voltage from T15 to GND is 12V when ancillary power is on, and 14.1V when the engine is running. Also as expected. There was a delay in the T15 voltage jumping from 12V to 14V though; probably around 5 seconds. It doesn't budge when you press the S/S button. No video of that sorry.

So all-in-all, nothing we didn't expect?

Matt: Well done- I like no surprises!! Also thanks for the video!

The 2 volt DC reading is interesting and somewhat surprising. I assumed that this would be a DC voltage, but perhaps it's actually an AC trace - can't imagine why though given that it's simply shorted out by the S/S switch. Could be that PIN#60 on the BCM needs to see an actual signal form of some sort (rather than a DC voltage) when the native VW S/S switch is not closed (only guessing). I'm glad that my design uses a relay contact rather than a solid state switch - because the electrical isolation that results from the raw contacts doesn't impact in any way on the conditions of the signal on PIN#60.(hint to AJW - don't use the FET relay)
Don
PS: Try not to pierce the wires too often, If you can, leave the needles intact - less damage to the copper core this way

[Snail Style]

Ewww don't use a multimeter, your gunna find some problems if you do that.

[mattaus]

I assembled the circuit and after a quick phone call to Don in order to verify that I am in fact an idiot, the device was verified as working and ready for a test run in the car:

VW Automatic Stop/Start Deactivation Device Test



My observations:

• I didn't fiddle with the timings at all. For what it's worth I actually like how quickly it activated.
• Any delay you build into these devices that starts from the moment the circuit senses power, will literally start the nanosecond you press start/turn the key. So you definitely want enough delay to ensure the S/S system is deactivated AFTER the engine is at idle. Not sure what the consequences of it deactivating too early would be, and I don't want to find out.
• There was some concern that the jump from the ancillary 12V to cranking 14V would cause the device to 'reset'. This didn't happen.
• The 2 second activation time is completely unnoticeable (by virtue of how the switch actually functions). It could be cut down a lot, but it doesn't hurt either way. Probably safer to leave it as is.
• Plenty of room to install the device in and around the gear shifter. Your mileage may vary with a manual shifter though.
• We need to work out a robust and easy way to actually interface with the existing wiring loom. I'm going to do some research and see if there is an adapter we can buy that would suit.
• More in-depth testing is definitely needed, but this cannot be done until I have a way to interface with the car better (see the point above).

All in all, a pretty damn good result for a first try. Testament to Don's brilliance if anything

- Matt

[DV52]

^^^^ Matt: I don't want this to sound like a mutual admiration thing, but my very deepest appreciation for your efforts. I'm making no comment about the way that the components were soldered together (in the video), but the ultimate test is -does the damn thing work -and apparently it does. Wooo-weee halleluiah big G!

So, it seems that the preliminary results are encouraging. But before we both get too excited, I think that we need to understand if there are any longer term issues to address. (i.e. with the car operating under lots of normal conditions).

As discussed, I reckon that you should settle on initial values for T1 and T2. Then dab a dob of glue onto the two potentiometers (something not too strong). Then package-up the components into a sleeve of heat shrink tubing and semi-attach the device to the three wires (so that it can be removed if further modes are needed). Then just drive the car as normal and observe any idiosyncrasies that may arise (with the device- that is).

I'm anticipating receipt of the second monostable oscillator by post later in the week. I'll solder together my unit and post a pic of the arrangement.

Again many thanks for your sterling efforts and for your input to the early stages of this fledgling device!
Don

PS: There's a simple solution if the voltage rise from 12V to 14V (because of the alternator kicking-in) becomes problematic. We could just run the device at a lower voltage (5 Volts, say) by placing a 3 pin regulator in front of the mono. The regulator would provide electrical separation for the mono from the voltage rise (all that would happen is that the volt-drop across the 3 pin regulator would increase when the alternator started - end effect would be an increased power loss in the voltage regulator. But since the device doesn't use much power, the increased heating would be marginal) I would need to change the specs of the relay if this was necessary, but this is not a problem (the TRR1A12D comes in a 5 volt version).