Clonking rear end - resolved

In all the engine excitement I had forgotten that I have actually resolved one of the many issues since the car has been on the road.

The clonking rear end when coming on/off the throttle, and clunking when reversing. I thought I had narrowed it down to CV joints but while I was under the car doing something else I thought after almost 1000 miles its probably a good idea to do a general spanner check. The suspension has been adjusted a few times so I went round everything I could see to make sure it hadn't come loose.

It turns out a few things were surprisingly loose. Bellhousing bolts for one, on the passenger side. One of them was only hand tight!

Then the big one - Diff carrier mounts. Took a FULL turn of a socket. They weren't loose, they just weren't effin' tight, which I believe is the precise torque rating required. Since nipping these up there have been NO clonks whatsoever when sharply  transitioning on/off the throttle.

A recurring theme is that every fixing which has needed nipping up has had a plain nut and spring washers. I am now of the opinion that spring washers are useless for anti-shake. One job for the future is to go round and replace any plain nut/spring washer combos for nylocs (unless its somewhere toasty hot of course). Some places like the bellhousing are a bolt into a threaded hole so there I'll leave the washers in (for clearance as otherwise I'll need shorter bolts) but will apply some loctite to the threads.

Engine scare - seems to be ok

The gunge in the header tank caused much worry and headache but appears (not 100% convinced yet) to be resolved. We drained the coolant and flushed it through with the garden hose pipe. We thoroughly flushed the block, heater matrix circuit and the radiator. I started asking questions and reading as much as I could about what brown gunge could be a symptom of - I already knew of course, oil in coolant from a blown head gasket. This tallied up with the car getting quite hot on the rolling road so I feared the worst.

Thanks to some excellent guidance from the cobra forum I tried to remain calm and worked methodically.

Checked oil on the dipstick - appeared normal.
Drained oil and inspected it - normal, no emulsified oil (mayonnaise)
Rocker covers are baffled so no mayonnaise there.
Removed all spark plugs. The rear two on the drivers side are a pig to get out as the engine is actually offset slightly so there's barely enough clearance to get the plugs out. For interest, this is what the plugs looked like:

They look a bit white which would suggest its running lean but from what I gather with modern fuels you can't really read spark plugs any more. Either way what it does confirm is that if I'm to make more progress with the carb tuning I desperately need to get my AF meter installed! Anyway, on with the diagnosis:

No coolant poured out of any cylinders (good!)
Plugs were dry, if a little oily.
Compression test - bought a compression tester but stupidly bought one with a very long body so couldn't get it in the aforementioned rear two cylinders so I couldn't check all the cylinders. The cylinders I did check all read 170psi which is absolutely normal.
The important test - coolant sniffer test. Borrowed a kit from a local garage. I had never used one before, you take the cap off the header tank and shove this chemical tester in. If combustion gases are present (i.e. leaking head gasket) the fluid turns yellow. If not it stays blue.
Test fluid stayed blue.

Feeling brave, we put the engine back together, fresh fluids and went for a short 20 mile drive. The hot running problems in traffic still persist but generally everything was fine. Checked coolant again afterwards, no brown sludge.

The next day, feeling very very brave we drove 90 miles each way to Warwick. Gave it plenty of stick, various conditions. Long periods at very light throttle cruise. Some sharp bursts of hard acceleration, some more prolonged stints at higher revs. Got there and back with no issues apart from the expected - hot in traffic.

Following advice, I've now got the header tank filled far less as it was consistently pissing out the overflow but I just kept on topping it up without thinking. the level is now about 1/3 in the header tank. I had been obsessed with getting the header tank as high as possible so that halfway up the tank equates to the highest point in the coolant circuit (thermostat + top hose bend) but of course that's where my air bleed to the header is, and the system is pressurised during operation as the coolant expands, so it doesn't matter. As an engineer I'm ashamed I didn't realise this and needed it pointing out to me.

Anyway, it seems to be quite happy but I can't help be a little suspicious. All of this meant we had to cancel the paint job but we are now back on for painting next weekend.

Hmm

Hmm well this doesn't look good. The water pump was leaking yet again so no I'm well versed in removal and refitting I drained the cooling system and just happened to glance into the header tank and its all full of brown sludge:

Here's the dregs of what I drained out

This is the offending sludge scraped out the inside of the header tank:

Return to the Rolling Road

We went back for another rolling road session yesterday morning to verify there's no detonation now the timing is correct and also to check the idle circuit is set up properly and see what its like on light cruise/part throttle and do a power run too.

We managed to achieve all this, which was nice. For some reason we couldn't get a reliable lambda value, it kept waving around between 0.8 and 1.0 throughout the rev/load range and at idle it was as high as 2.0 but again varied a lot so we didn't trust it, at idle this variation was probably because of the cam. CORRECTION on this - the cam plays a part but its also because the exhausts are such large diameter using a gas analyser probe shoved up the end of the exhaust the air in the room can get in at low rpm's and muck with the oxygen reading. I have a plan afoot to fit wideband O2 sensors in the collectors, connected to a suitable gauage/datalogger. I'm in the early stages of research at the moment but the Innovate LC1 and AEM 30-4900 systems look pretty good.

We did get stable CO for the 70mph load (to test throttle range) and full throttle power run and it was hovering around 2.5-3.0% which is tad on the rich side at around 13:1 A/F ratio which in my limited understanding is acceptable for these engines.

For the power run we had difficulty stopping the wheels spinning on the rollers, we had to spray WD40 on the tyres to soften them up and strapped the car down to get some load on the tyres. Here is Naomi being used as ballast:

Eventually we got a good power run in and this was the satisfying result:

376bhp and 392ft.lbs torque aren't bad, considering the engine spec I bought was 375/375! That would also explain why my neck hurts after a few good 2nd and 3rd gear blasts! Very satisfying to see the engine is making what it should but there is still the little niggle at light throttle it doesnt run as well as other people with similar engines/hot cams say their engines run (unless they're telling porky pies of course). Another visit might be on the cards at some point but now the focus moves towards prep for painting next week.

The cobra made a new friend at the garage where the rolling road is, a very nice looking GT40 (kit)

Non leaking water pump. Definitely, this time.

A while ago I replaced the gasket on the rear cover plate on the water pump. I used a good quality felpro gasket and some hylomar blue but still managed to make a balls of it and it was weeping slightly from the bottom (the bit you can't get to with the pump still on the engine).

I decided to replace all the water pump gaskets including the pump-to-block gaskets with black RTV. It took a while to clean all the old felpro gasket off where it had been practically welded on by the heat but 50 miles or so later and its still dry. I'm cautious to say its a success so I'll keep an eye on it.

This was the mess that the old gasket had made - lower two screws on the back plate were weeping and making a mess of the harmonic balancer.

While I had the cooling system in bits I moved the fan temperature switch to before the radiator instead of after so the fan kicks in based on the engine temp (or at least closer to it) rather than the cooled water that has just been through the radiator which seems more sensible. I also managed to shuffle the header tank up a little higher.

Correct Ride Height

Ever since the first drive for IVA, the suspension has been jacked up to comedy height so the indicators passed the minimum height test. With all the engine problems, sorting the ride height out has been at the bottom of the list but on Saturday we found time to address it.

According to the Dax manual, the correct ride height for a De Dion is 5.25" front, 6.25" rear. Ours was wound up to 6.75" front and 7.5" rear. The height wasn't really noticeable in normal driving but on heavy braking the nose pitches forwards alarmingly and it was vague and wallowy in the corners.

We set about winding the spring seats down and re-measured. With the fronts as low as they would go we managed 5.5". Perhaps with a little more time it will settle even lower so I'll keep an eye on it. The rears we set to 6.5" to maintain the 1" rake from rear to front.

Changing the ride height of course completely knackers the camber and toe at the front so we made a string box round the car and adjusted it back to spec. 0.25° negative camber and 0.25° toe in. Rear toe was unchanged, rear camber is not adjustable.

We made the changes and went for a short drive and it was immediately apparent something was wrong. The steering wheel was way off centre and the car darted around as if it had somehow shed 500kg. Good for racing perhaps, not good for a bumpy road. We quickly returned home and rechecked everything - turns out somehow we had managed to put 1° toe-in on the front! No idea how we managed that. Corrected to 0.25° and its now nice and stable. We decided to check the damper settings while we were at it and the rear was too soft, should be 10 clicks from softest and I had it at 2 clicks.

The car isn't completely transformed but is definitely far more stable over bumps, the steering rack height was set for the "true" ride height, not the IVA ride height so the bump steer has been reduced. The increase in damper rate on the rear has made it more stable under hard acceleration on a bumpy surface but the front seems to have lost a bit of directness. The initial turn-in is a bit sluggish but once your in a corner it feels very good and stable even if you're on/off the throttle (so long as your not jabbing at the pedal like a maniac of course). I think perhaps the front could take a click or two harder on the dampers, or maybe I need to check tyre pressures. but generally it feels very good and its just small tweaks now to find the optimum.

This is it now at the correct(ish) ride height. More pictures to follow.

Rolling Road - Follow - Up

After the not-too-successful rolling road session, I had a friend round and we tried to work out what was going on. One of the things suggested was to make sure the timing mark was at TDC so we pulled the plugs out and turned the engine over by hand. As I rotated the harmonic balancer we noticed (with associated groan of realisation) that there are two white marks, we were using the wrong one! I don't know what the second mark equated to, but it explains why the timing was all messed up. We started from scratch and verified TDC and scratched off the second mark on the damper.

We set that straight (after having to go out and buy a new timing light as my ebay one was broken after only a couple of uses). Went out for a drive and surprise surprise, no detonation. The car sounds better, pulls harder and runs smoother at low revs although its still far from perfect below 2000rpm. There's also still a bit of a bog when going for full throttle from low revs, possibly the jet change we made on the RR wasn't necessary now we've got the timing right.

Now that's sorted I'm booked in for another RR session in 2 weeks to eliminate the bog and try and recheck  the idle mixture.

We drove from Cambridge to Letchworth to visit a friend and it was extremely civilised. We used a quarter of a tank which is about 18 litres and did just over 70 miles which equals 17.6mpg, which is frankly astounding!

Rolling Road

Spent the afternoon on the rolling road at Revivals Garage in Thriplow near Duxford yesterday afternoon.

Was very interesting, turns out we were having some detonation problems on higher loads at low-mid revs. We put the timing back from 34° all in to about 25° all in which reduced it but didn't cure it entirely. Either the air/fuel mixture is incredibly lean or the TDC mark on the harmonic balancer is in the wrong place. We looked at the mixture first and changed the standard rods (0.113 main jet and 071x047 rod) to stage 1 rich +4% - calibration point 10 for edelbrock #1407 carb (0.116 main jet and 073x047 rod).

This helped a lot but still didn't entirely eliminate the problem but sadly we ran out of time. We learned an awful lot though. The road was set up to hold 80mph which is about 3800 in 4th and we then observed what was happening throughout the full throttle range. In the end we had it running nicely up to about 2/3 throttle as the secondaries start to come on then it detonates. Not perfect but a lot better than it was. I thought I had a rattle in the exhaust where I put some screws in to hold the collector on but it turns out it was pinking! Whoops!

Next jobs are to check the timing mark is actually at TDC as it was a bit odd we had to back the timing off so much. Then after that we'll change the standard orange rod spring for the stiffer pink spring from the carb calibration kit to keep the rod up and get more fuel in (assuming its a lean problem).

The Hunt for the Clonk

I've noticed (and mentioned on here previously, I'm sure) a clonk from somewhere in the drivetrain when coming sharply off the throttle and got under the car today to have a look. Initially I thought it was diff backlash, the diff is reconditioned and supplied by Dax. Following advice from the forum I jacked the car up, put it in gear to lock the prop and moved the wheels, carefully inspecting each joint in the drivetrain to see where the problem was.

Turns out there is practically no backlash in the diff at all. All the play is in the CV joints. These are off a BMW M5 and in the rolling chassis kit supplied by Dax they are not new parts. If  I had realised at the time then I would have just put new ones in, seems daft and is annoying as it's a big job to disassemble the entire back end to replace them, but at least I've identified the clonk.