520R SV build has commenced

[JohnCh]

Fair point and something to consider.  I've asked Caterham to confirm the wire is 1.5 mm² (assuming it's metric) which is a touch larger than 16 AWG.  First priority now though is getting the engine running well enough that I can journey to the dyno and have the tune sorted.

[sf4018]

I didn’t upsize the Caterham wiring for my external pump, works fine. Didn’t notice any heat build up in the cable either so doesn’t appear to be close either. Only reason I can think of to do is if the voltage drop for 16AWG at 10A is too high (can’t imagine it would be, the wires are not that long).

[JohnCh]

Thanks @sf4018, do you happen to know the amperage draw or can you share the make and model so I can look it up?  At this stage I have three options: (1) leave as-is, (2) replace the wire, or (3) replace the pump with the version rated at 165 LPH.  That's still more than enough and draws half the amps (<5) but likely won't lift the fuel as strongly.  I'd love to go with option 1.

 

 

[sf4018]

18 minutes ago, JohnCh said:

Thanks @sf4018, do you happen to know the amperage draw or can you share the make and model so I can look it up?  At this stage I have three options: (1) leave as-is, (2) replace the wire, or (3) replace the pump with the version rated at 165 LPH.  That's still more than enough and draws half the amps (<5) but likely won't lift the fuel as strongly.  I'd love to go with option 1.

 

 

https://caterhamparts.co.uk/pumps-senders/726-fuel-pump-for-bag-tank.html?search_query=5249970000&results=1

its a Bosch pump, 100lph. The fuel system is pressured to 60psi, the return is local to the tank, not at the rail. My guess is it’ll draw 2/3 the amps of yours. Don’t have the p/n.

[JohnCh]

Given I'm in a holding pattern waiting on parts to arrive before I can attempt driving the car again, I decided to finish up two projects: mirror mounts and glove box.  

 

The mirrors mount fairly high on the Caterham.  Coupled with the lowered floors, the mirrors are in my line of sight when going around corners; a pet peeve of mine (PM me for the full list).  Using the lower threaded mounts rather than the ones in the middle, and printing spacers that work around the lower sidescreen hinges, results in mirrors about 3.75" lower than stock.  Still a little higher than the Westfield, but close and it's enough to move the mirror out of the way. 

 

 

Next was the 3D printed glove box.  Thundersport, who makes the tunnel covers for Caterham, provided a matching piece of leather with my half hood order.  That covers the outside of the glove box and doubles as the hinge.  Magnets glued into the lid and into an adjustable fixture that bolts to the top of the glove box keep it closed.  I also added a mount for the 12v+ outlet and inserted a USB adapter which is easy to reach from the driver's seat yet not visible from the cabin.  Given the parasitic drain from the USB adapters, a simple on/off switch was added to avoid battery issues. 

 

 

 

 

It turned out ok, but mounting was compromised.  My original plan was to use countersunk screws through the bottom of the glove box, with nuts on the underside of the tunnel top.  Unfortunately, the chassis has a steel plate in that location and clearance between it and the tunnel cover is really tight.  The fix was switching to dome head screws that go up through the bottom of the tunnel cover with the nuts inside the glovebox.  Not optimal, but good enough.  A few printed caps to cover the nuts and screw ends should protect objects from scratches.  V1 of the cap design is shown below.  I might tweak it before printing the final set of four.

 

-John

[MV8]

Any plans for a cluster visor? I've used textured abs sheets off ebay, molded in the oven over wood form then allowed to cool to room temp to set but you could print something (add a thicker rolled edge for appearance). Attach with small, velcro, adhesive backed buttons for easy removal.

Test with cereal box and blue painters masking tape (leaves no residue) for best angle to your line of sight and hand clearance when turning.

 

Should work well for midday overhead with the top down and from the side near sunset/sunrise.

[JohnCh]

23 minutes ago, MV8 said:

Any plans for a cluster visor?

 

It's a last resort situation since it would require bringing the steering wheel closer to my body to gain sufficient finger clearance.  I'm hoping the mount that angles the display towards the driver's face will eliminate glare issues in most situations.  The Westfield has a Motogadget tach/speedo unit with an LCD display for speed and various other functions.  The angle is similar to the AiM, but It doesn't have anywhere near the brightness or contrast, yet glare is rarely an issue.  That gives me hope, but the only one way to find out is to drive it in various conditions.

 

 

-John

[JohnCh]

I finally received the various parts required to assemble the redesigned fuel pump module.  The big change is a new, smaller Hydramat that attaches directly to the bottom of the pump, eliminating the restrictions from the earlier design and freeing up a lot of space.  That in turn required a new pump bracket and convoluted PTFE tube on the pump outlet to accommodate the big S bend.  Unfortunately, a quick 200' drive on my property resulted in the same behavior as seen before: full lean, stalls, easily restarts, then repeats about every 200'.  Argh.... 

 

 

With an action camera livestreaming video of the gauge to my phone and extension cables allowing me to hold the multimeter next to it, I confirmed fuel pressure drops to zero before power is cut.  No doubt. this is not an ECU issue.  

 

 

Thinking through the possible causes given this didn't start happening until after I had covered about 12 miles, and then suddenly started doing it repeatedly with shrinking intervals, I began suspecting fuel level.  It was one thing I was certain had changed over that span.  Initially though, this seemed a red herring.  Based on measurements and subsequent calculations performed while fabricating the new pump bracket, I had about 0.9  gallons of fuel in the tank, and the Hydramat was fully submerged.  But what else could it be?  The pump and fuel pump module were all new and the fuel filter and fuel lines showed no blockages.  

 

Opting to add just 1/ 2 gallon as a test, I drove it again on the same route: 200' up the driveway, left U-turn onto the lawn, then back down the driveway before turning left to face the garage.  Things were better.  The fuel pressure went to zero in the same spots – as I completed the U-turn and as I prepared to stop in front of the garage – but in 3 out of 4 instances, the fuel pressure recovered before the engine could die.  Hmm....  Left turns.  Granted these were made at idle speed in first gear, so not exactly a g-force inducing change of direction, but it was consistent.  I then drove a full circle to the left at idle speed.  Zero fuel pressure..  Repeated two more times with the same result.  Then 3 circles to the right.  Fuel pressure stayed steady at 44 psi.  After adding another gallon to the tank, the issue went away when doing circles to the left.  Woo hoo! 

 

My theory on what is happening and why is best explained with a cross section drawing of the tank:

Note, the above is not to scale and is an approximation based on my memory.  The factory fuel pump pick up is much higher than mine and misses about 1-1.5 gallons of available fuel in the tank.  My setup, however, is at the bottom and in theory provides access to all fuel in the tank.  The baffle separating the left and right sides of the tank that's intended to minimize fuel sloshing has its opening on the bottom front.  This is below the factory pick up but at the same level as mine.  This means that when turning to the left, the baffle keeps the fuel available for the factory inlet as it approaches its empty point.  However, it allows the fuel to vacate to the right side of the tank under similar circumstances for my set up.  

 

Hopefully this is the issue.  I'll find out tomorrow when I add more fuel and attempt road driving.   

 

Couple of other updates.  I heard back from Caterham about the wires feeding the fuel pump.  They are 2mm² which is 14 AWG equivalent.  That should be sufficient for the current pump, although I may still downsize it to avoid stressing the alternator during those rare times when all the big load items are on (headlights, fan, heated seats, and pump.) 

 

Lastly, after 4 long months, the no-logo ITG filter has arrived.

 

 

-John

[MV8]

Hooray! Back to basics.

The slosh issue can be fixed with an external tank mod and no additional space required but you might lose a half gallon of total fuel capacity. The tank would be cut along the bend line to the right of the strainer to the end of the tank. Cut 2-3 inches into the rear wall and bottom at the ends of the long cut to form flaps. Fold the flaps in to overlap then trim the overlap so they butt together, then tig. The resulting right end tank corner overhang is cut off and moved to the left next to the strainer to fill the gap. Gravity would keep the sump/cup full and the cup capacity would be enough to keep from going lean at high rpm for the duration of a high G left turn, also considering the return flow would also flow into the cup.

[JohnCh]

Thanks MV8, I'll keep that in mind if for a winter project.  The interesting thing for me, assuming my hypothesis is correct, is that my mods didn't create a problem, but rather exposed a shortcoming of the stock tank.  If I treat empty as the same fuel level where the stock pickup can no longer access fuel, then the issue will remain masked.  I should be able to program the AiM to reflect this virtual empty point.  

 

Looking back at when the engine first died, I was driving in a straight line, and the prior turn was a gentle sweep to the right.  This tells me that yesterday's experimentation was either fraught with false positives, or there were actually two issues at play: in-tank fuel plumbing issues (now addressed and possibly causing the random lean spikes) and fuel surge as fuel was consumed.  I'll report back later with today's results.

 

Thanks,

John

[JohnCh]

I put 22 miles on the car today.  No issues, and looking at the logs, the random lean spike seems to have disappeared.  However, it still goes lean (about 1.10 Lambda, 16 AFR) when adding throttle.  Looking at the map, it appears SBD didn't enable either of the acceleration fueling options. Does anyone know if that's expected with the MBE?  If it's not, does anyone have settings from another Caterham I could try as a baseline?

 

 

Thanks,

John

[MV8]

255ms is a very short period but there must be more to it than that number. I would enable just the accel with 130ms, increase/decrease by half to find the point where it is just enough to cover the lean spot (no more), look at decel to pull at least as much fuel on closing throttle. Pull more than added by accel to aid engine braking.

 

Need more accel enrichment in: first third of throttle, first third of rpm range, cooler engine temps. Switching to being based on just manifold pressure would create conflicts.

 

[JohnCh]

Thanks MV8.  As per the screenshot below from the software, there are a number of individual pages associated with this function and it appears they all have data.  However, it's unclear if that data is for whatever he normally uses as a base map -- which could be old and for a very different engine -- or if these are the values used in his maps for his similarly tuned Duratecs.  I'm hoping I can get a hold of an MBE map for a Duratec so I can compare.   If I don't get a hold of one in a bit, I may just click the enable button for Synchronous Transient fueling and see what happens.

 

 

[MV8]

In regard to tanks with no baffles, here is a '90s ranger poly fuel tank (installs longitudinally in the chassis) replacing my '83 carbed v8 ranger steel tank as part of an upgrade. These newer tanks use a single in-tank pump where the earlier tanks have nothing to prevent slosh and used an in-tank lift pump that goes to a quart reservoir that feeds the external high pressure pump. The reservoir has four lines; one from the tank lift pump, one to the high pressure pump, one from the engine fuel rail, and one to the tank as the return. The return and lift pump keep the reservoir full. The return to the tank is the upper most line to carry off excess without increasing the pressure in the reservoir. They make similar reservoirs now with integral pumps, similar to mercedes oem.

[ashyers]

John,

It's not uncommon for the accel functions to be switched off in light vehicles with ITB's. I monkeyed with it a bunch back when I worked on Atlantics and with the 7. I found that a good basemap was hard to beat. The accel functions often just muddied the water.

 

Andy

[JohnCh]

Thanks Andy, Simon was kind enough to share his map from SBD for comparison.  It also has the acceleration fueling disabled, so this seems intentional.  In contrast, Emerald does enable it on the map I'm using in the Westfield and it didn't show those lean spikes when logging after the engine was completed.  Perhaps part of the issue is that I'm seeing the lean spikes as they occur because the Lambda is shown in my AiM display?  I see it, so I'm aware.  I'll see what happens if I intentionally ignore the display under normal driving.  If I feel the issue, then I'll know it's not simply my reaction to a reading.

 

 

[sf4018]

The Lambda for my duratec is all over the place when accelerating/decelerating, between 0.85 to 1.30, but the engine feels good.

https://youtu.be/uqRcBIpMkSY

 

[JohnCh]

Thanks Simon.  My lean spikes are at tip in at low load sites.  I haven't yet tried heavier application like you would have on track.  That said, when I drove the car yesterday and tried to ignore the Lambda readings when tipping in the throttle to avoid a bias, it seemed to run better.  It might be good enough for the 35 mile trip to the dyno.  Well, there is an asterisk to that comment which reads "if the engine isn't bad."  <cue foreboding music>

 

I'd covered over 60 miles and it was time for its first oil change.  My recollection is that I initially put in 8 quarts. However, after sucking out the oil from the tank, removing the finger and oil filters, I had a shade over 5-1/4 quarts in the Pella.  The plugs were sooty but dry, yet when looking into the bores, a coat of oil on the top of every piston stared back at me.  After refilling with fresh break-in oil, I confirmed there is no visible oil smoke from the exhaust.  There was a brief puff of very black smoke but it coincided with a stumble when blipping the throttle and was consistent with earlier Lambda readings.  

 

A compression test showed 240psi across the board.  That seems high, but it could be my gauge.  It has historically shown 200-210 psi with the original iteration of the Westfield's Duratec which had 10.75:1 pistons vs. 12:1 in the 2.4L.  Leak down showed 2% in each cylinder and a borescope revealed nothing out of the ordinary beyond the thin film of oil.  Esslinger responded to my email Sunday evening (nice customer service).  His initial reaction was bad and led to much swearing and invoking the names of various deities on my part.  However, after seeing the leak down numbers and my confirmation I've never seen visible smoke from the exhaust, he thinks it might have been oil from initial startup before the rings had begun to seat.  That would also mean my estimation of how much oil is missing is, well…wrong.  

 

At this point I'm not sure what to think.  Other than the issues related to the fuel assembly and the map not being ideal -- yet according to the logs, nothing unsafe -- there are no other signs of issues.  No odd noises, starts and runs fine, no smoke from the exhaust that I've seen or smelled, and it doesn’t run hot.

 

If the rain holds out, I'll try to put another 10 miles on it today to see if there is measurable oil decrease in the tank and also keep an eye open for smoke.  There is still the possibility that I missed it or that running it for several minutes wasn't sufficient.  When I started this thread many, many months ago, I recall writing this was going to take a while.  This, however, is not quite what I had in mind.

 

-John

[KnifeySpoony]

I typically get a little over 6qts out during oil changes. That does leave a bit unaccounted for, but I assume you had a lot of blow by before the rings were seated. More importantly, how much did you have to add back in to get the level you were at before draining?

 

 

[JohnCh]

Thanks, I added nearly 6-3/4 quarts to get it just below the top of the baffle, but I can't guarantee I filled it to exactly that same spot before.  The fact that you add a little over 6 quarts makes me feel much better.  Now who is going to come along and make me feel much worse?