13th February 2009

Bodywork - more painting

  The main tub and the bonnet needed some GRP work before spraying them. Some of the GRP work was due to things I had done - like trimming the rear wheel arch returns too much so that there wasn't enough material left to get an SVA-friendly return with. Some of the work was needed due to the quality of the moulds, and the fact that they've been poorly repaired with masking tape, with the result that the bodywork needs quite a bit of remedial work to get it looking presentable.

As for the painting, it was the same procedure here as with the sidepods and the scuttle. First, a coat or two of etch primer...

...and then two coats of matt black cellulose.

Now I've fitted the bodywork it turns out a missed a few spots on the bonnet, but I can always touch up those areas. Since it's only cellulose paint it's not too toxic and doesn't have to be used only in the spraybooth, unlike the primer and the clearcoat I've got, which are epoxy and 2K respectively - both not the kind of stuff you want to be spraying anywhere else.

More wiring

  Having established that the engine would run with a relatively unmolested engine loom attached to it, I've then started adding in the wiring properly. So, high current cables running to and from the FIA switch, a new fusebox, and bolting the regulator down into place. The next step on the wiring plan was to join together the engine loom and the dashboard wiring (without adding the rest of the wiring, such as the lighting circuits and the sensor wiring) to the point where I could start and stop the engine from the controls on the dashboard.

After working out that I'd muddled up the two connectors on the DASH2, and correcting that, the DASH2 worked nicely as you can see. The dash warning lights are for neutral (green) and oil pressure (red). I prefer to have a rather prominent neutral light since in both races I've started I've managed to shift seamlessly from first into neutral at the start of the race, which does not help one make a blinding start off the line.

You can also see the small sub-dash to the right hand side of the picture. This has 6 buttons, which control the DL1 and DASH2. It will eventually also have the brake bias adjuster mounted in the large hole, and the pull-handle for the plumbed-in fire extinguisher mounted in the smaller hole on the left.

Having wired everything up, it was time to see if the engine would start from the dash controls. FIA switch on, and the DL1 lights up as it should do (I've followed Tim's advice and wired the DL1 in on a permanent +ve line, so it's only switched by the FIA switch and not the ignition switch. Ignition key switch on, and the warning lights and DASH2 come on (as you can see above), but the fuel pump doesn't prime (which it should). And the engine wouldn't start. Bugger.

I started off trying to work out why the fuel pump wasn't priming properly, which I assumed (incorrectly) was due to a fault in the fuel pump wiring circuit. When I'd formerly got the engine running the fuel pump wiring was a total lash-up, whereas I'd now wired it properly through the dash switch (the left-side toggle switch with the red emergency cover to the right of the FIA switch cuts power to the fuel pump). It turns out, of course, that it was nothing to do with the wiring for the fuel pump. However, in investigating the fault, including rather destructively taking apart the relay/diode pack which contains the fuel pump relay and starter circuit interlock relay, I did learn a few things.

1. Haynes manuals obviously just copy the factory service manual wire diagrams. The service manual shows a resistor in the relay/diode pack, albeit one which according to the diagram isn't actually connected to anything. There is no such resistor - just lots of diodes, 2 relays and a zener diode. Which Mr. Haynes would know if he'd taken apart the diode pack rather than just copying Mr. Yamaha's diagram...

2. There's an error in the Yamaha service manual for checking the fuel pump relay works. There are two blue/yellow wires coming out of the diode/relay pack. One is the earth sink for the fuel pump relay, the other is the earth sink for the interlock relay. The diagram in the Yamaha service manual shows the wrong blue/yellow wire as being the sink for the fuel pump relay - it should be the other one (and they're not internally connected - I checked).

3. The relays for the fuel pump and starter circuit interlocks are tiny. Roughly 8mm x 4mm x 5mm. And yet can handle 40A. Jolly clever these Japanese chappies.

In the end the problem turned out to be the tip-over sensor, or rather its absence. I'd rather fancied keeping the tip-over sensor - after all, shutting the engine down if the car turns upside down seems like a jolly sensible idea to me. However, the problem with these is that they're generally based on pendulums (pendula?) which when mounted in a car get tripped every time you go round a corner fast because (once again) bikes lean over in corners whereas cars don't. I was convinced that the Yamaha tip-over sensor couldn't be large enough for a pendulum-based device, but I was wrong again. It contains a tiny pendulum with a magnet on it and some circuitry (Hall effect sensor) mounted in this small box.

In order to work out how to bypass the tip-over sensor I back-probed the sensor. It showed impedances of 1.8 million ohms for one side of the circuit and 155k ohms for the other side. Based upon a supply voltage of roughly 12 volts this gave an output for a potential divider with R2 as 155k ohms and R1 as 1M8 ohms of about 1 volt, which is apparently what the ECU expects to see when the bike is upright. So to fool the ECU into thinking the tip-over sensor was present and working properly, I added 1M8 and 155k resistors between the signal wire and the +ve supply and earth wire respectively.

There were 2 problems with this otherwise incredibly cunning plan:

1. The +ve line isn't 12V - it's 5V. The ECU generates a 5V line which is distributed throughout all the blue cables in the engine loom. Presumably the sensors need a regulated voltage to work consistently, and 5V is easier to generate than 12V.

2. I suspect the tip-over sensor contains active circuitry rather than just working as a passive potential divider. Also, I suspect that by using resistors with impedance as high as 1M8, the internal impedance of the ECU was having an effect on the voltage being created by the potential divider. At any rate, if I left the signal wire unconnected to the ECU it gave a voltage of about 0.4V. Once I connected the signal wire to the ECU this went up to over 4V.

This was, of course, why the engine wasn't working and the fuel pump wasn't firing - the ECU thought the bike was lying on its side. I reduced the value of the R2 resistor in the potential divider all the way down to 1k2 ohms before I got an output voltage in the correct region (just over 1V). As soon as the voltage from the tip-over sensor bypass circuitry got into the right range, the fuel pump primed and the engine started with no difficulties.

It took 8 hours to get this sorted, and sadly none of it was very photogenic...

  So, after all this labour, I had a working ignition circuit and an engine that could be started from the dashboard. However, I also had a whole heap of rather untidy wiring. Unlike Tim, the problem with the R1 loom wasn't that the various wires weren't long enough - quite the opposite, the problem was that there was too much of them, in particular the 16 wires going from the ECU to and from the injectors and various sensors, servos and motors in the throttle bodies. I did think about simply bundling them all up and hiding the excess beneath copious layers of loom tape, but in the end I decided I couldn't bring myself to do that, so the next job was to tidy up the engine loom wiring...