Hi spencer, I cross fingers for the next weekend! :)

Steady progress through the week in preparation for this weekends activities.

Great evenings work tonight; most important thing was to get the oil circuit primed and make sure we have no leaks. And that we get oil pressure, of course. Removal of the lower plugs in order to give the starter motor an easier time, and then a video call into Chris, who very kindly witnessed the first cranking of the motor, live from Blighty. It's a good job he warned me that it will take a good 20-25secs of cranking (2x 15sec pulls, no longer), to get the oil pressure needle off the bottom of the gauge; I know I've done this before, but paranoia is at play, and knowing that the time it's taking to get oil pressure up was very reassuring.

Great news, we have just under 2bar of pressure at (faster than usual) cranking speed. A thorough check under the engine and in the engine bay, and I can't see any leaks. Looking good, but not counting my chickens just yet.

Next job, to fill the water cooling system. Good news there as well, no leaks! Again, I'm not home and dry yet; coolant has only filled as far as the electric pump, including the cooling matrix, but there's quite a few joints that are now "wet", so pretty happy that it's not weeping from anywhere.

The plan for the weekend has changed a little; Simon (wiring) is now coming along tomorrow to perform a systems check, make sure we can see all the sensors, etc. Sam is still coming along on Sunday to build a base fire-up map, start the engine and do some "snagging", make sure there aren't any silly problems before we book dates on the dyno.

It's moving along...

Thanks mate! Fingers, toes, your families fingers and toes....heck, even your neighbours please, I'll take double doses of lucky tablets as well ;)

SmileWavy

The message passed them Spencer, everything goes well take place! :RÉ

:) ^^^ Not quite, as it turned out. Although your new found use of GIFs did make me chuckle Mr C!

Right, some slight alterations in the plan for last weekend meant Simon and I were doing harness and sensor checks on Saturday morning, entering the various parameters for (most of) the sensors dotted around the car. We identified a few strange things, like the ECU not powering down as you'd hope, but nothing that should've stopped the car from firing up on Sunday. Checks done, and I dashed off to the races for a mates 40th birthday. Very pleasant it was too, English countryside on a perfect summers day, gloriously Pimms O'Clock.

Sundays agenda was then for Sam to come and play with the system, build some initial maps and see if we can get her to fire up. I have to confess, if ever there was a good use case for for the tag #mindblown, then this was it. Another incredibly clever guy, it was like having an 8hr masterclass in calibration theory. And quantum Physics (joke). Although I absolutely "got" the concepts, there's not a chance that much of it has stuck in the grey matter!

It was during this time that it became very apparent just how good, and capable, the Syvecs unit is. I honestly hadn't quite appreciated this, not knowing quite what the background to Life Racing is (ex Pectel, sold to Cosworth, then went to develop engines - AER - then needed an electronics arm to control them, hence Life Racing). Then understanding where Syvecs have come from with their OEM partnership with Life Racing...all very interesting. To me, at least.

Very reassuring to witness some of the intricate rules Sam has written that are now part of my calibration file, making absolutely sure that if there's anything untoward with the parameters that the ECU is seeing, it will cut the engine....way quicker than anyone could monitor from the cabin. However, it's not just a case of "if Oil-T = < xx ºC then kill engine", we're able to also build additional conditions (arguments) that specify a time element, that for example would cater for hot engine being shut down (ie fuel stop) and then needing to restart some short period later, perhaps catering for heat soak. I'm sure that makes no sense whatsoever, and it's only meant to be illustrative (it's also my poor interpretation and exhausted brain)...needless to say, it has it's roots very much in motorsport levels of robustness. Aerospace grade components used throughout with wider heat tolerances than off-the-shelf components, code written by the same guys who code the Life Racing range. Sitting adjacent to Sam in the office, in fact.

The successful calibration between the throttle position pot, and the eGas throttle body, was completed; Sam being particularly happy with how accurate that has turned out to be. Again, he started showing me a myriad of oscilloscope trace windows (in the main Syvecs SCal interface) that were showing how much trim compensation and suchlike was needed in order to get the behaviour of the throttle body as he wanted it. Bear with me, I was struggling to even grasp the concepts of this; apparently very much in the PLD arena in terms of systems control. (#mindblown)

But....did we manage to get the engine fired up? Nope. Unfortunately not, and I have to say, I was bitterly disappointed. But, a few days later, and a little perspective thrown against the situation, and it's not all bad. In fact, a long way from that. It was a very productive days work with building the calibration, discussing the project (it's the first time that Sam has seen the car) and importantly, sorting out a plan of action.

We were hindered by an own-goal on my behalf. The bloody crank position sensor....again!

If you cast your memory back 3 years (and about 30 pages) ago to when the engine was on the dyno for Build v1.0, we had issues with picking up a signal. In our rush to just get the damn thing fired up and running-in, we changed both the sensor for an inductive type, driven by the fact that the only spare MBE 9A9 ECU which was available, was configured for inductive sensors, not hall effect (in MBE systems, its not changeable, set at a HW level; Syvecs you can use either). We never did get to the bottom of why it didn't work, and suspected perhaps a faulty ECU.

A few months ago, I decided to tidy things up with regards to the wiring; the inductive sensor we hurriedly used had about 75cm of cable attached to it, when we only needed 10cm. So I asked Simon to revert back to the HE sensor I'm using all over the car, for wheel speed and cam position; a Honeywell "1GT". I could then carry 1x spare sensor in my spares package, and be comfortable that I could change any one of 6 sensors if I needed to, and avoid being stranded.

This then caused Sam and I some massive head scratching. VERY weird traces on the ECU diagnostics. Swapped a spare sensor in its place - same result.

Chris very kindly then called us (what a legend), and guessed exactly what the problem was....the bloody 1GT sensor. In short, not suitable for 60-2 tooth flywheel resolutions. Old 964 Turbo dual mass flywheel? Possibly, less teeth, lower resolution.

So, the problem was completely of my own making, me succumbing to OCD.

At the moment, I'm waiting for a new sensor (inductive) to arrive.....I gave away the one that I *was* using to the guy who bought my old MBE ECU!! And it's no longer made by Bosch....With fortunate hindsight, Simon had run 4 feeds to the Deutch connector for the crank position; 3x electrical feeds and a screen, so no major surgery required to accommodate an inductive sensor, which obviously doesn't require a +5v feed.

So, unfortunate but nevermind....it was a productive weekend, albeit hidden inside a lovely black anodised black box, rather than vapours being pumped out of exhaust tips.

Spencer--

This a really instructive, informative post, that shows not only how far you have come with this project, but insight into the "reality" of sensor inputs. We like to think these are stable, constant inputs with predictable resistance ranges-- in fact, they are all over the place, influenced by EMFs, ground loops, weird transients, noise. In the analog world we just muddle through-- what do you really care if your oil pressure gauge oscillates below the level of the needle's dampening-- but in the ECU world it's GIGO-- garbage in, garbage out.

The fact that you are plugged in to a community that fundamentally understands the forces at work, and how to compensate for them, is a HUGE differentiator for the success of your project.

And for us. Carry on.

Evening folks, time for another update. A significant one this time!

The last few weeks have been spent being a) incredibly busy at work, and b) relentlessly pushing forward with diving activities, and getting Lucy qualified with her Open Water certs, ready to "hit the water diving" in Malta in 3 weeks time (and me continuing my Advanced OW, practicing a few things in readiness). That's now done, Lucy's now OW certified, and we're good to go. Time to crack on with the car project.

Simon made another trip across last weekend to make some changes to the harness, and to make a jumper harness for the new crank position sensor. The previous HE sensor used a Deutsch 4w connector, with +V, -V and Gnd, the new inductive (VR) sensor uses Sig +, Sig - and a Screen; thankfully Simon had the foresight to run a screen through the harness, so a simple adapter from Deutsch to the successor to the 3w junior mini-timer connector that Bosch are now using. Some health-checking and it was job-done for the day.

The new Bosch crank sensor is an even easier fit into the original sensor mounting bracket that mounts onto the back of the crankcase half, poking through the bellhousing aperture. No adapters needed to cater for a longer C-C on the mounting bolt, nor any spacing required to cater for the sensor being longer; it just drops straight in.

http://img.photobucket.com/albums/v2...d/IMG_3471.jpg

Firing up the Syvecs S-Cal s/w to hack my way around and find a way to monitor the crank pulse waveform, and all looks good (although the shot below was taken before I zoomed into the top trace, which revealed the 60 pulses, then the indexing gap)

http://img.photobucket.com/albums/v2...d/IMG_3472.jpg

Every evening for the last week I've then been playing around and getting the car into a state where we can drive it this weekend, as opposed to just getting it to a point where we can fire the engine up. Lots of odd-and-sods jobs to be done, not least of which to sort out a few leaks on my water cooling system. Nothing major, just teething problems.

Sam arrived again yesterday morning for session #2; the obvious aim being to ultimately get the car out onto the public highway for a shakedown and to ensure all parameters are reading correctly before leaving me to it, to run in the engine on the road over the next month or so.

So, first check was to make sure we had a good crank pulse - and we did, no problems at all. Second cranking on the starter, I thought was going to be to just check phasing of cam position v crank. What I wasn't expecting (or I would've rigged up the GoPro in advance) was that engine would burst into life!

<iframe width="560" height="315" src="https://www.youtube.com/embed/V-P95pKpILk?rel=0" frameborder="0" allow="autoplay; encrypted-media" allowfullscreen></iframe>

What I then discovered moments after shutdown, was that the top set of coils weren't even connected! We (Simon and I) deliberately disconnected them a few weeks ago, in case the ignition was inadvertently switched on without the correct setting for the coils being set up - didn't want to fry them, of course. I'd got as far as reconnecting the easy-to-get-to lower plugs, but not the tops - I wasn't expecting a start sequence! I think you'll agree, she sound pretty damn sweet for the very first firing!!! (especially on only the lower 6 plugs)

This is the second fire-up. The safety strategy that Sam has programmed in prevents the engine firing until a decently high oil pressure has been achieved, hence the long "pull" on the starter motor.

<iframe width="560" height="315" src="https://www.youtube.com/embed/KcSEKhGtQEA?rel=0" frameborder="0" allow="autoplay; encrypted-media" allowfullscreen></iframe>

So, all's going well, but we had a few hiccups shortly after. The Lambda sensor simply wasn't reading for some reason. After some head scratching, we realised that the Syvecs S6+ motherboard is "wired" for the Bosch 4.2 type sensor, not the 4.9 sensor which I'm using. However, a very easy thing to fix; a dozen or so hex bolts later and we're doing open-heart surgery on the main board, un-soldering 2x bridging pads and re-soldering them to different positions. If you can imagine 3x pads in a line, A, B and C, where B is common. B bridged to C = Bosch 4.2, or B bridged to A = Bosch 4.9. This needed to be done on 2 sets of pads to effect the change (and is entirely above-board, described in their manual, so not a hack by any stretch, it's just needed in order to support a wide range of sensors, incl NTK).

All OK after that? Hmm, not really, we could now see readings, but the sensor simply wasn't heating up as it should. Although it was a new part for Engine Build #1, it's entirely likely that it became contaminated with what happened last time, blow-by gases, or perhaps a temperature shock, etc. So, an hours round-trip drive to buy the one remaining unit that was in stock in the local area, and it's all great again....perfect readings. £100 lighter but a big relief. (back-to-back re-testing of the old sensor later confirmed it was shot). Apparently it is known that the Bosch units are a little fragile, certainly not as robust as the NTK units AER/Life Racing use for their motorsport applications....but they are 1/4 > 1/3rd the price. And available on a Saturday afternoon.

Many hours then spent with Sam tweaking maps, whilst I continue my work to get the car drivable. Items deliberately left unfitted, like the intercooler, inlet snorkel, breather catch tank, etc, all needed to be fitted now that we had inspected for leaks. I tell you, the workout I got for that whole days activities was immense. As some of you know, we're going through week 4 of a heatwave here in the UK (nothing compared to Arizona or some parts of SA, for sure, but for us Brits, it's very nice and hot). This coupled with a stinking hot car that's heating up the garage nicely, and I'm sweating like a good'un.

Wheels back on, off the ramps and time to drive it. Gulp. First impressions? Wow....not at all bad for the first stab at making sure it has road manners. Some initial over sensitivity was easily dealt with by Sam, perched on the back set (no front passenger seat in the car yet!) with laptop on, er, lap.

This thing wants to make boost REALLY quickly; My insistence on Sam applying "Nanny State" limits to revs, boost and throttle opening meant that he had to knock the eGas opening down to just 20% (from 40% initially) so that it's not quite so quick to crash into the 0.7bar boost max that he originally set.

A 10-15mile run out, no dramas at all. No stalling at junctions, pulling away from standstill is beautifully smooth and she idles superbly. Oil temps all well under control.

There are still some tasks to be completed though. Firstly, the tacho drive either needs to be enabled from the ECU, or some wiring needs to be looked at. It was previously wired into the harness for the MBE ECU, so I'm hoping nothings changed in that regard, and it's just s/w enablement.

The oil level gauge on the dashboard is all over the place, reading intermittently. Needs pulling and testing.

The oil pressure gauge is over-reading compared to the ECU. At 3,000 rpm the Syvecs is showing 3.9bar, but the dash is declaring >5 bar (i.e. off the scale).

There's a fair amount of electrical noise when monitoring #6 cylinder knock. It's not knock itself, more of an unclean signal versus the other cylinders (not seen it on the S-Cal sw, so can't comment any further than this)

There's an annoying buzz from the front fuel pump; I think there's metal-to-metal contact somewhere, perhaps an inlet/outlet fitting.

Underbody panels need refitting, as does the rear bumper, then time to pile on the miles with frequent inspections.

So, a fantastic weekend, major milestone reached. But...I'm trying not to get too excited. Been here once before, where everything sounded great but suspicious symptoms proved otherwise. I'm sure with all the rules and safety settings that Sam has plumbed in, nothing untoward can happen. In his words, everything is set to "paranoid" levels of safety, where even the briefest loss in oil pressure will kill the engine.

I am very satisfied for you Spencer! :)

Great news, another major milestone behind you.

Thanks mate!! It won't be long until I'll be cruising through the French countryside in it!

Haha, I'm so cautiously optimistic at the moment mate, really am NOT counting any chickens until it's been on the rollers, done some decent pulls, and leakdown tests have been done.

Okay, mini-update....

We were originally scheduled to be up with Northampton Motorsport on the 11th August for the dyno runs; unfortunately, due to work commitments, Sam has had to postpone....now looking to be early September. Not the end of the world, a) it's very close to being road legal and usable, and b) 11 Aug all of a sudden feels quite close, especially with a diving holiday in between. We all know how tight things get if there are any snags that need addressing, so I'm not too upset about it.

As I type this, the car was due to be at my pals for its annual MoT test, however, something that we didn't do before Sam left on Saturday, was to leave the car with a calibration that'll get it through the emissions test. Should be easy enough to correct, some remote dial-in work onto my laptop and then supercede the calibration file?....well in theory yes, except that despite me having a valid connection from my laptop last week (Macbook Pro, running Win 10 under Parallels), it just will NOT connect again now. Beyond annoying. So, time to try 3 other Win10 devices I have, all running native Windows rather than virtualised - same result with them, constant "Offline" being displayed.

The Syvecs unit doesn't use the full TCP/IP protocol stack, thus doesn't use/issue IP addresses (therefore DHCP is no use). Difficult to diagnose, no way to ping it, only to monitor I/O traffic via the Windows networking setup and diagnostics screens. Indeed, there was handshaking going on, but once done, no continuous comms. At one point I thought I had a faulty Thunderbolt>RJ45 adapter (and still not convinced it's not broken), so another quick dash out to the local PC shop to buy a USB>RJ45 adapter, but to no avail.

However....this lunchtime, I decided to have a go at configuring the TPLink mini router I bought specifically for this job. I hadn't used it thus far, working on the KISS principle, surely a plugged in network cable has got to be simpler than trying to configure a wi-fi access point? Well, after switching some of the TPLink settings (for it to actually behave as an Access Point) it all connected beautifully!! Both my Mac and my Acer tablet. Excellent. Team Viewer also installed, so as soon as Sam has a spare window, he can remote in and do-the-necessary.

"Look Mum, no wires!"
http://img.photobucket.com/albums/v2...d/IMG_3562.jpg

http://img.photobucket.com/albums/v2...d/IMG_3568.jpg

Chaos in the cabin, yet to be mounted but it's all live....
http://img.photobucket.com/albums/v2...d/IMG_3569.jpg


In the meantime, the rear bumper has now gone back on, but not fully bolted in place under the rear quarter panels. The lights and reflector are back in, and bloody hell, she looks gorgeous again...Ok, a little dirty, but I've got my 964 back!!

http://img.photobucket.com/albums/v2...d/IMG_3567.jpg

Oh, and I finally got around to adjusting the engine cover latch! For ages, it's been a case of stretching between the lid release puller and simultaneously using your right hand to lift up the wing. I won't pretend that adjusting it was the work of 5 minutes - it very much wasn't - but it now releases with my dignity intact.

Its alive! ALIVE!

Classically, there are a bunch of small items. Of course you have the right approach, make a list and check everything off.

Nitrox?

Nitrox? :confused:

Being blown away by the level of detail and commitment in your work, I have silently been following, hoping to see you bring this back where it belongs. I am very happy to catch up on the latest achievements, it appears you are closer than ever to where you deserve to be.

A huge congrats from Denmark.

Cheers

;) Cheers Jesper, appreciate the generous comments. Yeah, hoping that the end is in sight, there are so many other projects I want to switch my attention to (not car projects! I'll never do this again!!)

Okay, time for an update.

The bumper's back off, along with inner arch liners. :mad::rolleyes:

Remember the new Bosch Lambda sensor I had to buy a few weeks ago? Well, we had another remote session for Sam to adjust the calibration in order to pass the emissions test for the UK MoT, but all was not well again. No heating control despite a clear heat-up ramping profile. Cue the face/palm despair. And swearing.

I should explain at this point that the Syvecs S6+ unit has a daughterboard, essentially the bot that makes an S6 into an S6+. It adds inputs/outputs, but also has the Bosch CJ125 controller chip mounted on it; the chip that controls the LSU4.2 and LSU4.9 sensors. The mainboard is made by Life Racing, as well as the code for this being written by them, but the daughterboard is made for Syvecs by a 3rd party, and I presume that the associated code is written by either the 3rd party or by Syvecs themselves.

Why is this relevant? Well, it turns out there was some issue that was discovered with that CJ125 chip concerning the code. The only announcement that I'm aware of was the final post in a thread on the Syvecs forum! Hmm. So, we've potentially f***ed a brand new Bosch sensor because of a code issue. Terrific.

A code reflash to the daughterboard later (easy enough), and time to give it another go. Any better? No. Different behaviour but ultimately still not working well enough. Looking at the control parameters of what the new code is now doing, it's not subtly different, it's radically altered (apparently). The Lambda now looks to be heating up fine, but then "dropping off a cliff" once up to temperature. A complete head-scratcher, which nobody seems to know what to do about. Emails off to Syvecs support, and no reply back. Not that impressed, I have to say.

I've read and learnt more about Lambda sensors than I ever wanted to know, and it seems that Syvecs (Life Racing) very much seems to be designed around the NTK sensors rather than Bosch (a conclusion I've come to after reading about more frequent updates to linearisation curves, and far more chat traffic on the forum from the design guys at Syvecs).

The NTK sensor is controlled directly by the mainboard, not the daughterboard, so is using Life Racing code. There's essentially 3 different types of NTK probes;

There's some mention of an L1H1, but this was superceded by the L2H2 some years ago, with a part number of LZA-09-E1. It's an off-the-shelf unit as fitted to Honda Civics (amongst others). Readily available and is anywhere from £115 - £180.

Then there's the LZA-03-E1, known as the "motorsport" sensor; higher quality, but twice the price, and not at all easily available, with supply contracts in place between (I think) Life Racing and NTK.

Above that, there's a Lab grade NTK sensor, used for calibration and workshop monitoring systems, i.e., the kit used on dyno's. These are reassuringly, breathtakingly expensive, and not what's used on the car itself.

Another point of note is that the Bosch sensor has calibration resistors built into the plug, and so you can't shorten and re-terminate an LSU4.9 (or 4.2, I think). The NTK doesn't have these resistors built in, and can be re-terminated with a non-NTK plug if desired; commonly used is the MIL spec plug, more heatproof than the stock plastic unit, or Deutsch for something cheaper and smaller. Calibration – if required – is then done in the linearization curves, which are in the SCal software.

Long story a little shorter, I don’t want to continue cocking-about trying to get a Bosch sensor working, and then worrying that it may just decide to give-up one day – it’s a pain to change, the bumper has to come back off to access it. So, I’ve gone with an off-the-shelf NTK L2H2 unit.

http://img.photobucket.com/albums/v2...d/IMG_3674.jpg

http://img.photobucket.com/albums/v2...d/IMG_3675.jpg

Sam’s very happy with these sensors, having used them many times before in motorsport applications (non-works, “privateer”) and in his own cars. Northampton Motorsport (who are going to be dyno’ing the car in September) also speak highly of the NTK unit, having had poor experiences with the Bosch sensor previously. They also said anecdotally that the linearisation curves, when compared to their very expensive workshop unit, are spot-on identical; all pretty reassuring. I’m also keeping the stock termination; if I ever do need to replace one “in the field” (and they ought to be considered a service item really), I certainly don’t want to be thwarted by then having to re-terminate with a MIL/Deutsch connector.

So, decision made – the next job was to rework the wiring harness, which has a Bosch connector. Not the easiest thing to get to, however, removal of the intake snorkel and intercooler and I can *just about* get enough access. Chop off the Bosch plug and refit a 6w Deutsch connector.

http://img.photobucket.com/albums/v2...d/IMG_3681.jpg

http://img.photobucket.com/albums/v2...d/IMG_3680.jpg

http://img.photobucket.com/albums/v2...d/IMG_3682.jpg
(heatshrink boot deliberately not fitted until we test it)

Simon has made me a 40cm interfacing harness, Deutsch at one end, NTK plug the other all wrapped in Raychem DR25.

http://img.photobucket.com/albums/v2...d/IMG_3684.jpg

Lambda now screwed into the exhaust, and wired up – we’re ready for another fire-up attempt tomorrow evening. If OK, then I can book the MoT for when I get back from vacation.

http://img.photobucket.com/albums/v2...d/IMG_3683.jpg

In parallel, I’ve also been trying to get the hard-wired Cat6 cable from the ECU to talk to the laptop without going through the wifi router….not that I want to use it this way, but it *should* work like this…and it isn’t…and it’s pissing me right off. Tried everything I can think of, checked the Tx/Rx connections into the ECU, which are all fine, I’ve even made up another Cat6 cable and substituted it into the harness. Slightly different result; SCal now intermittently (when it does detect the ECU) declares that the calibration files are locked, implying that there’s a concurrent connection into the ECU from another device that’s also running SCal. And there most certainly isn’t.

http://img.photobucket.com/albums/v2...d/IMG_3667.jpg

Life’s too short, and ultimately I only want to use Wifi to connect anyway. But with only a single wifi aerial in the Mac, I needed to get an additional wifi adapter, or else continue using a 25m Cat5e cable that runs from the house into the garage (far from ideal).

£10 for one of these via Amazon…..

http://img.photobucket.com/albums/v2...d/IMG_3685.jpg

….and at last, I’m able to setup one aerial for connecting to the house router for Team Viewer remote access, and the other connection to the Syvecs wifi router for SCal access, all fine running on Win10 virtualised under the Parallels hypervisor and OSX High Sierra (as tested by using Team Viewer in the OSX environment, to then remotely connect to the Win10 environment and in to SCal).

More updates over the next day or so. Hopefully.

Freddie Uncle Charlie Kate. I should have pressured you into Motec when I had the chance.

Your perseverance has taken on entirely new dimensions. The stiff upper is now forged titanium.

You are a better man than I: anytime I run Parallels to emulate windows, my uncrashable Mac. . . crashes. I think it can SMELL MS code.

I close with this:

<iframe width="560" height="315" src="https://www.youtube.com/embed/pdRH5wzCQQw" frameborder="0" allow="autoplay; encrypted-media" allowfullscreen></iframe>

I think they all have their "issues" to be honest, I've heard great things about MoTec, but I've also heard many more not-so-good things. And some of their units are now looking distinctly old-hat.

Believe me, I'm FAST losing the will to even finish this, genuinely wish I'd never started; just thoroughly bored by it now, finding any reason I can think of to NOT work on the car...just sick to death of it. Never, ever, again.

To be fair, most of the time Parallels is fine on my setup. Latest versions I've had (v11 then skipped to v13) have caused the fans on the Mac to run full-tilt; there's a known KB about it, which I've tried but makes little difference. I don't use it as my work tool (any more) so it's not something that's a 12hr/day annoyance. The only thing which is hard work compared to native Windoze is the adapter connections and bridging networks.

So, another remote dial-in session this evning......surprise-sur-****ing-prise, it's still not playing ball. Time to check the wiring again (although I'm 99.9% sure it's fine), everything in SW is set for the new Lambda, and yet we're not getting it to heat correctly.

I'm off on vacation this weekend....hoping the car gets stolen in the meantime.

:-(

Release(Leave) nothing Spencer, you are going to arrive there!!! If you give up I would be very disappointed and disgusted for all the efforts which you made, you have to go at the end of your project for you and for the forum, it is an obligation! ;)

Hang in there! We’re all pulling for you.

Fingers crossed there’s an a-ha moment and it all comes together.

Best wishes!


Sent from my iPad using Tapatalk

Hi Spenny just seeing if you made any progress?or if you found any lost treasure during your under water adventures?

Sent from my SM-N950U using Tapatalk

Hi Guys,

As-of yesterday evening, yes, some progress thanks. Sam went on vacation just as I came back from mine, so in the meantime I've been speaking with Syvecs in order to ensure that the changes needed on the mainboard of the ECU are what's needed to switch it from working with a Bosch sensor to an NTK. After some slightly contradictory advice, we've now got that nailed....

http://img.photobucket.com/albums/v2...d/IMG_4071.jpg

http://img.photobucket.com/albums/v2...d/IMG_4072.jpg

After getting back from another days diving on Saturday (PADI Rescue Diver course, great fun), I popped a brave pill and took the car out for a 30mile drive mid-evening with a mate and his "trade plates". Hopefully nice and quiet should there be any difficulties.

All's well, no oil leaks. A few rattles to sort out, not least of which is a buzzing from the front fuel pump. I checked this a few weeks ago, nothing obvious I could find, but I'll have another look; perhaps it's the cover thats vibrating on a fitting. I'm also going to switch back over to the stock 964T engine mounts from the 964RS mounts currently in there. I'm hearing too many noises and feeling vibrations that I don't remember from when the car was stock - it could be nothing to worry about and simply a function of harder mounts. There also appears to be a rumbling from the alternator or cooling fan, dammit....could do without that, now that we're only a fortnight away from dyno-day, with 500miles to get driven before then.

Yesterday was the day the car had been re-booked in for it's annual MoT test, and I have to say, I wasn't at all confident that it would pass the emissions test after I had observed the readings in the Syvecs SCal "dashboard"...but then I have very limited knowledge of such things and how to interpret it, so....

Well, as predicted, it didn't pass all of the emissions test; it DID pass the hydrocarbons element, but not CO2. Oh, and I also have a sidelight bulb blown.

http://img.photobucket.com/albums/v2...d/IMG_4101.jpg

http://img.photobucket.com/albums/v2...d/IMG_4100.jpg

So, coming back home, "kicking the cat" (metaphorically) and cueing up a few Plan B's (involving returning the ECU for a thorough diagnostics check and if required, a session up at Northampton Motorsport this weekend to really get to the bottom of it), I then touched-base with Sam in the evening. A quick video of what I was seeing on the Lam1 trace in SCal and he confirmed that it was absolutely working, just not very aggressively. Purely down to how he's set it up. At the time of typing, we're scheduled for another remote session this evening for him to rework the control calibration, and I'll go back and see if it then passes the test.

I do know that it's going to be very tricky, as the catalytic converter is integrated into the silencer casing, thus very small....but we'll see (and yes, I do have to retain a converter and comply with emissions, as it's a 1993 registered car which is the first year of cars needing to abide by emissions tests in the UK).

We also need to get that water cooling circuit operational, as even a light drive over a short distance is causing it to boil over and spew bright red coolant.

Keep you posted....

Okay, just got the A/C pump off to make way for the RH engine mount swap....and the pulley on that definitely has some play.

I'm presuming there shouldn't be any play at all, despite there being a clutch mechanism in there?

So close, Spencer!

Stand firm Spencer! Your experience(experiment) to cross(spend) my engine efi has me cools little, that seems complicated enough even by making things completely as you made him(it). ;)

Well finally some positive news and distinct progress!!

For various reasons, the remote-in session a couple of nights ago didn't happen. We did get a quick session in last night, and hallelujah, the Lambda sensor is at last working beautifully.

[Interesting side note, and funny how "religious" folk are with preferences; Sam was pretty cutting with thoughts on Bosch sensors versus NTK, in his mind they're like chalk and cheese with a preference of NTK every single time. Interesting, I'd never really given it that much consideration before this project, a lambda sensor being just a lambda sensor, surely?....Nope, in my research I now understand more about these things than I ever wanted to know, and then consider the level of engines that Sam works with in his day job (LMP1, WRC, WTCR, Mazda Daytona Prototypes, etc), and what they use (NTK), perhaps that's useful info for some out there.....for reference, my sensor is the NTK "L2H2" which supercedes their previous L1H1, which translates as p/n LZA09-E1. Not quite the same as they use in the professional applications, but perfectly good for road cars and anything other than top-flight works applications. The R5 rally cars that run Life Racing, for example, use exactly this sensor.]

So, onto last nights tweaking session - I could detect the joy in Sams voice on the end of the phone when he could see (and hear) the closed loop control actually working for the first time on this engine; I think it's the most excited I've heard him, haha. A few moments later and unbelievably quick flitting of screens in SCal via Team Viewer (it was truly a blur), and I've got a refined calibration for annual testing purposes, and a further optimised calibration for my running-in duties.

A call over to my pal this afternoon, and he had a spare 30mins to get a re-test done on the emissions. Because of the rush to get over there, in my haste I forgot to switch over to the "MoT calibration" once the car was in the workshop (what a dumbass), but...it still passed! So, I'm all legal, and able to now start driving it once we rig up the water cooling circuit.

We've identified that it could be marginal driving the fluid pump directly from a PWM output on the ECU; steady state current draw is fine, but the inrush on startup may pop the drive output. Same goes for the Spal cooling fan.....so we're going to stick on the safe side and eventually drive it via a relay. No speed control by doing so, but not really needed in this instance, just on/off @100% controlled by the ECT sensor that's in the circuit.

In order to just get up and running, and over to Simon to make the modification properly, I'm going to temporarily rig up a +VE feed from something like the rear screen heater, ignition switched but with control via the demister switch.

Once done, I have a fortnight to get 300-400 miles driven and an oil change done, before heading up to Northampton Motorsport for a days dyno'ing.

Fantastic news, Spencer! So relieved for you, I know it's been a major source of stress.

The change from PWM control to switched should be a simple one, and I cannot imagine you'd do any real harm driving it without that functioning as we drive these in 40C+ weather with no cooling circuit and don't experience bearing failures.

Looking forward to your driving reports!

Cheers Chris, I'm more worried about the boiling-over of the fluid that is steadily "kettling" itself without any significant airflow over the radiator, as it's 100% shrouded by the cooling fan mounted to the outside.

Just got back from the first real drive with empty dual carriageways, where I could start leaning on the boost. It's mapped very "softly" in terms of throttle opening but still with a boost limit of 0.7bar, so it gets there but not in a rapid manner. The primary reason for testing was to monitor the coolant temps using my Windows tablet and the Syvecs SMon software. Surprisingly, the airflow in that part of the car does look to be pretty significant; cruising anywhere between 50mph and 70(ish) mph, and the temps stay between 62-66ºC, and that was for about 18miles of driving. As soon as I slowed down coming back into my street and parking in the garage, they went up to 75-76ºC straight away.....so, in order to pile on the miles, I'm going to drive it over to Simons place in the morning, so we can rig-up a temporary solution to switch the fan and coolant pump on. Reckon it should be safe.

So, first thoughts on the drive itself....hmm, needs work. Predictably of course. Light throttle opening, i.e., when trundling through town, and it kangaroos all over the place. The only way to overcome it is to be more brutal with the throttle, and lift more frequently. This (I think) is due to a change Sam made the other evening once we had a working Lambda sensor, as it didn't do it on my first drive out. It's also stalling a lot easier, but again, a relatively easy change to make.

More worrying though, is that I've still got a very coarse vibration, I'm sure from the engine. Feels just like it's completely out of balance; like a washing-machine that's on a spin cycle, but with an uneven load in the drum. Sounds horrible, feels worse. Better under acceleration, but can't live with it like it is when cruising.

Engine mounts were changed the other day back to OE stock items from the genuine RS ones, and it's made no difference. Bit of a head-scratcher, as it was there on Engine v1.0 as well, exact same sensation. Wondering if it's perhaps a function of a single-mass flywheel? I still have the dual-mass item stored (and think I saw a receipt in the file, indicating it was only a thousand or two miles old when I bought the car).

"Answers on a postcard"....

Brilliant, I am great satisfied for you Spencer! :)

"More worrying though, is that I've still got a very coarse vibration, I'm sure from the engine. Feels just like it's completely out of balance; like a washing-machine that's on a spin cycle, but with an uneven load in the drum. Sounds horrible, feels worse. Better under acceleration, but can't live with it like it is when cruising."
Spencer, it is maybe a room in badly centered or badly balanced rotation.

Hi Pat,

I'm certain it's not a balancing issue; if you remember from the backend of last year, I took the flywheel, rods, pistons and crank pulley (and even the cam position sprocket) up to one of the best balancing experts in the UK - it was all balanced to within a fraction of its life! (but this is what it feels like).

Chatting this through with Chris over the last few evenings, an excellent suggestion which I'll try, is to get the timing lights onto the crank pulley, and see whether there is timing variance on any of the firing cycles. I have to confess, when Chris mentioned it, and how easy it is for this to happen, I did think "that's EXACTLY what it feels like, mis-timed ignition"

My other thought was whether I have something as "simple" as a dead cylinder, but to have two failed plugs, or coils, or sections of wiring harness would be pretty unlikely....although not impossible, as it's exactly what we had on the engine dyno some years ago.

The diagnostics work will commence after the weekend....

Good evening Spencer, I look forward to knowing the bottom line of this history as regards this vibration, where from she results. You soon arrive at the end of your project, I am satisfied for you. :)

So another update is due...

Today was the longest journey I've done in the car on Engine v2.0, about a 65 mile round-trip. Sounds trivial when talking about a modern car, but not so trivial when you're listening out for every little noise or sign of something not-quite-being-right. And of course, these old things are wonderfully old-skool; heavy steering, offset pedals, hugely mechanical gearshift, none of the DCT flappy paddle tech-fest of my daily driver!

With my Windows10 tablet booted up and connected to the wifi router for the whole journey, simply to monitor the engine coolant temp parameter, the journey was spectacularly uneventful. Sure, all of the quirks I mentioned in last nights post are still there, but I was able to drive around the kangarooing throttle, and so long as I kept moving, the temps remained in the mid 70ºC zone.

As mentioned yesterday, I'd intended a relatively quick visit to Simon, just to get a temporary harness rigged up to enable the water pump and fan....however, Simon very kindly said we may as well just crack-on and get it done properly with new relays installed. Passenger seat out, and we were soon at work making it so. All tested, everything working perfectly. The fan has been programmed to come on with ignition (for the time being), but it seems the pump hasn't yet been programmed to come on at all, as I had similar temps on my return trip. Of course, the fan is cooling the water in the radiator, but when relying on any natural syphoning of the fluid through the bearing housing, well, it's not too quick to cool the whole water circuit down. (A quick solution, of course is to short out the relay to permanently enable it until it's re-programmed).

I'm going to relocate my MAP sensor very close to the plenum of the intake, rather than having 45-50cm of silicon hose connecting it to where it's currently mounted on top of the 1-2-3 bank. Sam's keen to minimise any possibility of hose deformation under open throttle, as well as any delay with that amount of "dead air" in the pipework. Getting it closer will sharpen throttle response. So, another extension to the loom required, with two joining sections; the first section is to re-join it back onto the existing sensor I'm using. The second section will replace that, and connect to a different sensor I have, which will mount either into the plenum itself or very close to it and joined with a short section of -4 teflon hose. Another job done.

Next up was diagnosing why I don't have a working rev counter. We're still not completely on top of this one yet, however, strong progress today with some more modifications to the Syvecs mainboard, and a swap around between two of the PWM outputs. Essentially, by using "PWM 1" or "PWM 2" outputs, you then have the option to solder one of the bridges on the board, to incorporate a 470 Ohm pull-up resistor to 12v....which is what 99% of tacho's need, a PWM signal pulling to ground for each pulse. Measuring this output in several places on the 964 (easiest one being pin 14 in the diagnostic socket) meant we could monitor a pulse....but still nothing at the tacho. I'll trace it back, now that I'm back home and can remove the steering wheel and column shroud (can't remove the top half of the shroud as there's a hidden 3rd screw behind the steering boss...can't remove the tacho without removing the top shroud!). Hopefully it's just a plug not properly seated, but I guess it could be a blown tacho that needs repair.

Finally, a job I wasn't expecting to get done (or rather, another job I wasn't expecting to get done) was re-commissioning the A/C system. If you stretch your memory back a few years (about 3), I needed to re-work the A/C lines completely, as stock 964 Turbo lines kind of run diagonally in a curved/looped fashion to the A/C pump. No good for me, too much new stuff in the way, so I made new lines that ran tight to the bulkhead and behind the power steering pump to the front of the engine bay. Remarkably simple, once I’d also completely fabricated new mounting blocks onto the pump itself, so they clear the intercooler and other bits-n-bobs. Up until today, the welds had only been tested with a penetrating dye but never subjected to A/C gas or even a vacuum test. Now, as luck would have it, Simon also has a full A/C recharging rig – this is the other side of his business.

http://img.photobucket.com/albums/v2...d/IMG_4156.jpg

As the hoses had been open to the air for some while, with the engine builds happening, I had intended replacing the (brand new) drier cannister. But it wasn’t high on the priority list. Simons advice, however, was that if we vacuum the hell out of the system for an hour or two, and use Nitrogen to flush it, then we should be fine…..fingers crossed that the fabricated connections and brand new rubber lines and fittings, are all good. And they were!! Vacuum held perfectly, and an hour or so after letting it suck everything out, we had a fully charged system, chilling down to 11ºC (with significant heat soak into the car and radiator, as we’d been firing it up throughout the morning to test things. No forward motion of course, either. Very happy with that, the A/C pump still works, the clutch engages, perfecto!

So, I’m sitting here typing this at (almost) midnight, still grinning – a very positive day, lots of progress. Still a few jobs to get done of course, primarily checking the timing, but it’s moving in the right direction finally!!

Which just goes to show that you should never lose hope! :)