I made a CNPS harness the previous weekend (with no capacitors), installed new Iridium BKR7EIX-11 sparkplugs and tried firing up the car with the GM D514A CNPS but it wouldn't turn over, just coughing and farting - very similar to the Fab9 COPS I tried previously. I figured it was probably because capacitors are really needed from 12V to GND and ordered some new bits to make another harness.
This weekend's plan was to finish up the GM CNPS installation since I finally had everything I needed.
My CNPS brackets were slightly too short since there wasn't any clearance between the rocker cover and the harness socket's locking tab - the socket's locking tab angles downward and touches the rocker cover.
I found that two small washers underneath each of the 4 bolt-points were enough to provide the required clearance but after dropping a few into the engine bay, knew that was NOT a long-term solution. Luckily I didn't have the splash tray fitted and being small and round, they bounced around the engine bay and rolled out underneath.
Two 3mm aluminium shims solved this issue:
They've been shaped so that they're not visible.
Next issue was the FM Big Spark Kit Magnecor wires:
They've been made specifically for the FM bracket so are too long for me.
I ordered some 180* boots and connectors from Summit racing so that I could shorten the FM wires. Unfortunately the Summit boots are way too tight - you can force them onto the coils but getting them off again is a wrestling match. I then bought a set of Commodore (Aussie GM vehicle that uses the LS engine) Procomp 10.5mm wires from eBay, figuring I could re-use their boots and connectors.
Here's what the Procomp leads look like with the ends removed (left) vs the Summit kit (on right):
I'm using 3 180* connectors and will use the shortest FM 90* lead for cylinder 4.
The Procomp boots fit nice and snug and their connectors are much longer than the Summit connector - they're also sheathed.
Here's the Magnecor 8.5mm lead (top) vs the Procomp 10.5mm lead with the connectors removed:
- much tidier!
The previous wiring harness I made used a common ground wire for all CNPS and the wiring was as short as possible but it made the harness very chunky and inflexible so this time I made the leads longer and made them as two separate harnesses, one for cylinder 1&4 and one for cylinder 2&3, each with their own 4700uF capacitor from 12V to engine GND.
Here's a great DIY soldering tool you should make if you don't have it already (shown holding some scrap wire for demonstration purposes):
- much better than "helping hands" and really made my job of in-engine-bay soldering the ID1000 plugs into the factory harness "a doddle".
The GM factory wiring uses the crimps shown on the left below:
These crimps are great since they make it easier to join 4 wires together - trying to solder 4 wires together when their other ends are already connected to plugs is an exercise in chronic frustration. Unfortunately I couldn't find any so I hacked up some of my blade connectors (shown on the right), cutting the two end sections off for use as wire crimps. These DIY crimps plus soldering "for good measure" produced nice compact joins which passed the "tug test".
Here's the new harness (the other one looks the same):
The capacitor is sheathed in some clear heatshrink and zip-tied to the harness.
Everything all in place, battery fully charged, dwell time set to a conservative 3200us, turn the key and...
CRANK, NO START!
I played around with the dwell time, going up in increments of 200us, all the way up to 5000us; the best I got was some coughing and farting again. All this tweaking really took a toll on the battery - at the end I noted that it was dropping to 9V during cranking and was at about 11.7V when I stopped.
I then spent a couple of days reading everything I could about this and suspect it has something to do with my Ignition MAP which looks like it hasn't really been touched.
I put the original NGK HKS M35i sparkplugs and OEM wiring back in, dropped the dwell time back to stock 2100uS, and everything fired up as normal.
More investigation required!
The D514A coils are also known as LS7 coils and I found this on the internetz:
Here are the dwell settings for the LS7 coils, again from Motec Aus via IJ:
Batt V Dwell
Those dwell times below 11V are a bit concerning. Should I be concerned about increasing the dwell time beyond 5ms?
The D585 coils are known to prematurely fire at 5ms but I haven't heard this is an issue with the D514A/LS7 coils.