SVA Preparation
With SVA looming I got stuck into the usual small list of jobs that takes twice as long as you think.
Rent-a-cat
When I dropped off the bodywork up at The Kitcar Workshop I rented what has become known as a "rent-a-cat" system from them in order to get my carbed R1 engine through the Stringent post-95 emissions test.
|
|
|
|
|
|
|
|
The rent-a-cat consists of a catalytic converter, heated (4-wire) lambda probe, electronic control unit (ECU), two air bleed valves, water temperature sender and various bit of pipe, tee pieces and mounts. The system works by monitoring the exhaust gases with the lambda probe and then bleeding air into the inlet manifold to keep the mixture stoichemetric (lambda = 1.0). The catalytic converter then reduces the amount of C02 and HCO in the exhaust gases at the tailpipe.
Installation was very straightforward. It was necessary to get an M18x1.5 thread pitch boss (half nut) welded into the final collector, just in front of the silencer, in a position where it was 'seeing' the gases from all four cylinders. The lambda probe is screwed into this and the cable run to the battery tray and plugged into the ECU.
The catalytic converter goes behind the lambda probe, in my case I welded this inside the front of the silencer, onto the back of the reducer I had made to fit the standard R1 silencer onto a 2" pipe. it was necessary to dismantle the front of the standard R1 silencer to get the cat in it. There was a surprising amount of room inside there! I also got a tight radius 2" 90 degree elbow welded on to the outlet to stop the rear wing getting burned.
The two air valves each feed a pair of cylinders. I used the screw in adapters that came with my Morgan Carbtune to attach hoses from a pair of inlets to a tee piece and then on to an air valve. With my exhaust manifold pairing cylinders 1&4, 2&3 I decided to pair inlets 1&3, 2&4 into an air valve for no better reason that it seemed like the right thing to do to average out the mixture across the collectors. I mounted the air valves on a bracket I made off the stat mounting and ran the cables back to plug into the ECU.
The water temperature sender was simply cable-tied to the top hose and the cable run back to the ECU along with those from the air valves.
I positioned the ECU on the battery tray, secured by self-adhesive velcro and connected it up the +12V and ground.
With everything in place it was time to set it all up. The instructions that came with the rent-a-cat had obviously been translated to English by a non native English speaker, which made them a little cryptic, particularly in the area of setting up the pilot screws to achieve the correct operation of the system as indicated by the three LEDs on the ECU. This is what I worked out, which may be of additional use.
The green LED indicates that lambda =1.0. This is on continuously while the water is too cold for the system to be working and will blink on and off as the lambda drifts around 1.0 when the system is working.
The yellow LED indicates that the water is up to operational temperature and that the control unit is operational. In my case this happened at 60C.
The red LED goes out when an air valve is opened. Thus is will blink on and off as the air valves are modulated to control the flow of air.
When you first start the engine the green and red LEDs should be be illuminated indicating that everything is connected up OK and the air valves are closed. When I originally connected the lamdba sensor I plugged it into the second port on the ECU (lambda 2) rather than lambda 1 and the green LED did not illuminate.
With the water temperature below 60C the yellow LED remains off indicating that the system is not yet working. Let the engine idle until the water temperature rises and the yellow LED comes on. If the lambda is >1.0 at this point then you'll hear the air valves start to chatter at the same time.
If the lambda is <1.0 at this point then all LEDs will be on, but the air valves will not start to operate as the mixture is too lean.
The objective of setting the mixture screws seems to be to get the mixture just rich enough so that the air valves can compensate by modulating additional air into the inlets to keep the lambda equal to 1.0.
The instructions suggest adjusting the idle mixture (pilot) screws at 2000rpm. It is a good idea to screw all four pilot screws right in to their seats and then back out by the number of turns indicated by the service manual. This ensures that they are all set the same. I set mine initially to 3 turns out and increased the idle speed to 2000rpm.
The mixture should be leaned off, by screwing in all pilot screws by the same amount, until the green LED remains permanently lit. In my case this corresponded exactly to 3 turns out on the mixture screws as I had initially set them.
The pilot screws should then be screwed out a little at a time to richen the mixture until the green LEDs just begins to blink and the air valves can be heard beginning to operate. The red LED will flash in time with these. I found that this point was with the pilot screws out 3 1/2 turns and that in this position the air valves were operating cyclically as varying amounts of air (including none at all for a short period) was bled into the inlets to keep the lambda equal to 1.0.
I reduced the idle speed back down to 1,000 rpm and the air valves continued to operate cyclically.
One side effect of the rent-a-cat being fitted is that with the system fitted, but not yet working due to the water temperature being too cold, the idle speed will be lower than when the air valves are open and the mixture will be a little rich. This will lead to a slightly lumpy idle until the water warms up and the system begins functioning.
Other misc jobs
 |
 |
One thing that I had been putting off for ages was doing something with the HT leads. The SVA requires the HT leads to be suppressed, which bike leads are, but they are not marked as such. As I didn't particularly want to trash a perfectly good set of HT leads I decided to leave the original leads alone, tidy up how they were mounted and hope common sense prevails at the SVA.
The bias bar setup on the brakes was another cause for concern as the SVA testers are supposed to set them to provide the most rear bias for the brake bias test if they are not permanently locked in position. I decided upon several strategies for overcoming this including hiding the bias assembly behind a permanently fixed cover and going to the SVA armed with lock nuts and if these were deemed not permanent enough some roll pins so we could establish where the bias should be set then drill the pivot and just into the bar itself before securing with a roll pin.
In order to soften some of the many sharp edges on the Striker I bought a load of U shaped rubber trim, which went round the cycle wings, boot opening and along the aluminium angle the was standing proud of the nosecone. I put nut covers on the front suspension bolts and track rod ends and some kitchen worktop corner protectors on the bottom front edges of the chassis. I could have gone on forever covering bits up, but decided to do just the obvious bits and let the SVA tester point out where else he wanted some putting.
I also swapped out the complete upper steering column and replaced it with one to which a standard sierra wheel was attached with plenty of padding and a nice big diameter to exempt all my switches from radius testing.
I fitted a handbrake warning switch to the end of the wires that had been hanging there waiting for ages to be connected. As usual this took longer to do than it should have as there was no obvious means of attaching a switch to the lever provided in the kit and the first switch I bought needed to be screwed to the hardened bit of the mechanism, which was impossible to drill. Also the Digidash needed a +12V signal to illuminate the warning light rather than the more usual short to ground. In the end I found the perfect switch in the form of a waterproof microswitch with lever actuator from RS.
One area I was concerned over was the rear edge of the battery tray as this was below the bottom of the dash and thus would be subject to radius testing in case a knee should come into contact with it. This would be difficult to radius as it had rivet butts sticking out of it. Talking to Martin Bell whilst picking up the bodywork it seems to correct thing to do is to cover this along with the underside of the dash. Therefore I made a cover out of some 20G aluminium to conceal this and also conceal all the electrics under the dash.
I had also not bothered to connect up the headlights and front indicators, which needed a little debugging in the home-made loom to get the hazards working without the ignition on and the indicators with the ignition on, but nothing a bit of a quick circuit diagram couldn't put right and after all I did wire it together in the first place.
When the bodywork was collected there was the small matter of fitting the rear lights and connecting them up to the loom, which was done a little late in the day, but otherwise I was ready for the test.