This is a short story of electronic detection: of using clues and tools to
work out something which - at first glance - looks impossible. The outcome is
specific to one make and model, but the approach is universal.
The Problem
Many grey market Japanese imports come with navigation systems. That’s great,
but not so good is the Japanese software and Japanese language directions.
Initially it seems a good idea to convert the navigation system to English
and local mapping software, but this is very hard to do. Much easier is to
simply pull the navigation computer out of the car, so saving a bit of weight
and at the same time, freeing-up some space for the installation of an amplifier
or other electronic bits.
But that move can have some unexpected outcomes. Firstly, in some cars, the
nav system operates the car’s clock. So what happens to the clock when the nav
computer is pulled? In the case of the guinea pig car – a ’99 Toyota Prius – the
result was good. Whereas previously there had been no way of setting the clock
(and it always read wrongly by an hour), with the nav computer unplugged, up
came clock setting buttons on the colour dashboard LCD screen. Excellent.
But not so good was that at the same time the driver’s side door speaker
stopped working! There wouldn’t seem to be any connection between a navigation
computer and a car radio, but there is. What happens is that the nav computer
sends its oral instructions through the driver’s side door speaker. Because it
doesn’t want to be yelling its instructions over the top of the radio, the speaker is disconnected from the radio when the
instructions come through. That’s
fine - but when the nav computer is unplugged, the speaker goes silent all the
time.
And how the hell do you fix that?
Thoughts
As always, the first thing to do is to consult a wiring manual for the car.
But unfortunately, while I have a manual, it’s for the wrong model. The manual I
have is for an NHW11 model, not my NHW10. And the wiring changed substantially
between these models.
But thinking about the way the system could work, there were at least two
possible approaches.
The front-right speaker signal from the radio could be fed to the nav
computer. The nav computer could either pass this signal to the speaker or, when
it wanted to voice instructions, disconnect the radio and switch the speaker
output to the navigation system.
Alternatively, the nav computer could signal the radio when it wanted the
radio to cut off the speaker. The nav computer would always be wired to the
speaker but it would voice instructions only after the radio had been instructed
to turn off its feed to that speaker.
And here’s where the wiring diagram was useful – as shown at the bottom of
this diagram, the radio is always connected to the four speakers (approach 2),
rather than having one of the speakers fed through the nav system (approach 1).
So the nav computer must signal the radio to shut down the feed to the
front-right speaker. With the computer unplugged, this signal was present all
the time, so switching off this speaker.
But how on earth could the shape and form of this signal be found, and then
recreated and fed into the right wires? It seems impossible....
Testing
The nav computer is connected to the wiring loom with three plugs – large,
medium and small. The small plug carries only one shielded wiring connection –
almost certainly to the GPS aerial. The medium sized plug contains eight
connections, and the large plug nine connections.
The wires that signal the radio to shut down the speaker must be in one of
these plugs - but which? The first step was to unplug each in turn, isolating
which plug contained the relevant wires. Pulling the medium plug caused the
clock to come good, and pulling the large plug caused the speaker to switch off.
So the relevant wires were in the large plug - that simple move had halved the
number of possibles from 18 to 9.
But the nine wires that were in the large plug could be for anything: power,
earth, video feeds to the central colour LCD, speed or braking inputs. (But, on
reflection, probably not the video feed to the LCD. Because the LCD screen clock
came good when the small plug was pulled, this plug probably contains the LCD
feeds.) But how do you sort out which of the wires in the remaining large plug
are which?
Thinking about it further, most of these wires are going to inputs to the nav
computer. The only outputs are likely to be to the radio, to tell it to shut
down the speaker. So the inputs will be there whether the plug is in the nav
computer or is pulled, but the outputs will be present only when the plug is in
the nav computer. If that’s the case, some measurements with a multimeter
would reveal all.
I earthed one lead of a multimeter and with the ignition switched on,
measured the voltages on each of the plug wires with the plug pushed into the
nav computer. From the back of the plug, the results looked like this.
I then pulled the nav system plug and again measured the voltages. As can be
seen, two of the pins (highlighted) which previously had 6.2V on them had
dropped to zero.
It therefore seems likely that the radio looks for 6.2V on these two pins
before it activates the front-right speaker. I grabbed two short lengths of
hook-up wire and connected the 6.2V supplies to their adjoining pins.
The speaker came to life.....
It was then a simple case to cut and solder together the correct wires,
covering the joins with insulating tape.
Conclusion
I was a little lucky – if the car had a CAN communications bus and looked for
signals communicated on the bus before it activated the speaker, I’d have been
stuck. But by the same token, finding the right wires and then feeding-in the
correct voltages was no easy task.
In this type of situation:
- Think about the different ways in which the system might operate
- Isolate the relevant plug by pulling the plugs in turn
- Remember that input signals will be present irrespective of whether the plug
is pulled or not, but output signals will only be present when the plug is
inserted.*
- Recreate the required output signal and feed it to the appropriate
pins.
*an exception to this is where the sensor needs a power supply, or pull-up or
pull-down resistor, to function.
See, even what initially appears impossible can be done without blowing
anything up!
|