Zel's Motoring Adventures...Volvo, Renault, Rover, Trabant, Invacar & A Sinclair C5 - Updated 13/11.

[Zelandeth]

Glad to report that as of this morning the Activa has once again got fully functioning air conditioning.

I've also mostly got the new software configuration set up on my web server. Hopefully will get that back online this weekend at some point.

Evening edit:

Given that it's just coming up on six months I've had the Activa it's time for her routine service. Fresh oil and filter being the main part of that (it's not the least accessible filter I've had to get at, but it's not a million miles off)...

...While the oil was draining I topped off the windscreen washers, checked the coolant, tyre pressures and did a bit of a walk round looking for anything amiss. Total findings: one duff number plate bulb which I had a spare for on hand.

Noticed when idling the engine to circulate the new oil before the final level check that the throttle cable had quite a bit of slack so adjusted that. LHM level was checked, then pronounced the job done.

Here's to another uneventful six months hopefully!

[Zelandeth]

Given that progress has (temporarily) slowed down a bit with the Invacar, and the Saab will be off to pastures new shortly hopefully, I decided it was about time that I got my lazy tail in gear and set about getting the Lada ready for an MOT.  I'm briefly ignoring the windscreen issue in the hope that a solution will present itself if I stop looking.

The main thing (aside from the aforementioned stuffed windscreen) is that the injection system was in need of some tidying up.  It had initially been set up with the sole intention of proving that it would work - as I wasn't entirely convinced that it was a viable project when I started out.  As such there were some bits that any sane MOT tester would have laughed at.

Since I'd started the project though I'd obtained a proper injection spec Riva fuel tank and in-tank fuel pump (the Riva was sold in Canada and some of the Scandinavian counties in 1.7i form like the Niva apparently).  This would allow me to get rid of a good deal of plumbing and the Saab in-tank fuel pump that was cable-tied to the underside of the car.

First up, I needed to drain the tank down.  I'd deliberately let the fuel level get quite low before the winter layup for this reason.  The fuel light flickering at any movement of the car whatsoever.

Easiest way I decided was to pull the fuel supply line off the throttle body and direct it into a can, then stick a jumper in between the contact terminals of the fuel pump relay socket.

The fuel gauge lies like a rug.  I wound up very nearly filling a 20 litre Jerry can before it ran dry.  Having said that, there's no doubting the ability of the pump to provide volume - it took about a minute to fill each 5 litre can!

Time to dig out the shiny new fuel pump...

...and tank...

...and combine these two things - also taking time to attach the hose tails, and to LABEL the flow and return lines, as the odds of me remembering which is flow and return while crawling around under the car were somewhere between slim and zero.

The side trim in the boot of the car is only held in by a couple of self tapping screws, and once removed gives access to the tank filler neck and the world's most over-complicated tank breather system.

The blue hose is the main tank breather and goes to a port on the top of the tank.  I'm pretty sure that this should have a hose clip on, but there's no evidence of there ever having been one there...not that this necessarily means anything on a Lada!  I pulled it off the filler neck here so that it could just be removed along with the tank.

The access hole in the boot floor allows access to the fuel flow and return connectors and the fuel gauge sender, these were all disconnected.

The fuel filler connection is hidden behind a guard panel which like the engine bay splash guard is made of thick enough metal that you could probably jack the car up with it.  A trio of bolts hold it in place, one visible here, and two in the nearside rear wheel arch.

Once that's out the way access is easy.

Definite evidence of someone being in here before - mainly that one of the rubber sections there has a VW/Audi logo on it - albeit date stamped 1993, and covered in the same underbody paint as the rest of the car...so possibly a bodge dating back to when the car was nearly new?

The filler to tank connection was altered a bit later in the model run, instead of the three-piece design shown above, there's a longer "tail" on the tank, and a single-piece rubber boot that connects it to the filler neck, so that whole lot had to come out.

All that disconnected, just remained to remove the 13mm bolt nut and a couple of washers from each corner of the tank to release it.

...At this point I also noticed that I've lost an exhaust hanger - which may well explain why it's been rattling so much.  The temporary fuel pump can also just be seen peeking into this shot.  The exhaust in general is pretty shot...time for a stainless one from the cat back methinks.

With the nuts all removed, the tank was carefully lowered out without and drama, and set to one side.

The kink in the breather hose I suspect may be responsible for why filling the tank was such an annoyingly slow affair, so I'll put some measures in place to try to rectify that.

Doesn't look half bad inside given that this is from 1993.

Now this is where my trying to be pro-active caused me a bit of a headache.  I initially decided to replace that breather pipe with something rather more flexible (I had some fuel hose in a suitable diameter handy - it's the same size as the PCV line from the cam cover to the inlet manifold on a Saab 900...), which it turned out was a mistake.  The new hose has a very slightly wider external diameter, which meant that it stubbornly refused to nearly slip through the hole in the chassis rail as the tank was situated.  It's awkward enough trying to get the filler neck and fixing holes lined up on their own, never mind without trapping that hose.  I'll revert back to the original one with an anti-kink spring added when I next have a shot at sorting this.

Sadly that's where I had to abandon things yesterday to get cleaned up and cook dinner - and it's not stopped raining today.  With a bit of luck I'll get it all sorted next time around, as once the tank's in it should be a pretty simple matter to get the pipework connected up.  All that will remain then will be to trim and optimise the routing of a few hoses in the engine bay as it's a bit messy just now, and fit the new fuel filter while I've got stuff apart anyway.  After that it will be a thorough check over everything to make sure I'm not missing any hose clips (including the breather, vapour separator and purge system pipework), and check it still works.

Bit annoyed I didn't manage to get the tank properly in place yesterday, but that's just how it works out sometimes isn't it?

[Zelandeth]

First achievement of today was *finally* getting my web server back online.

I'd taken it down with the intention of actually getting a proper clean OS installed on it, rather than a very nearly 10 year old bodged and hacked about Ubuntu installation which I think dated from three (maybe four) machines ago.  Despite that fact, the thing was absolutely stable and I don't think I ever had a single crash to report.  However, there were a bunch of stuff that didn't work properly, and most recently local network access added itself to that list, meaning I had to resort to using a bloody USB stick to get pages onto the thing.

Wiped the disk, stuck CentOS7 on it, got Apache etc installed and set about configuring it.  Initially had some file permission issues which took me a moment to figure out (probably because last time I set it up was nearly a decade ago), but didn't cause any real headaches.

...What did take me three days worth of on and off poking the blasted thing was getting it to actually serve a page rather than a directory listing.  Eventually I tracked this down to being because the configuration file was asking for pages named index.html rather than index.htm which was actually on the drive...Rookie mistake.  Anyhow, it's working again now so I'm not going to complain.

Back to the cars...

So I left the Lada with the new fuel tank half hanging out the back end of it propped up by an empty fuel can.  Less than ideal.

I spent about half an hour today fighting with various different configurations until eventually giving in and sticking the original kinked breather hose back in place (figure I'll cut it in the future and stick a metal elbow in), as at least I knew that was the right shape!  It only took another hour or so of swearing at it until I eventually managed to get both the breather and the filler neck to behave and let me slot the tank properly into place.

Filler neck hose clips put in place...

...I definitely reckon that the older style three-piece joint here is *far* easier to work with.  I really hope the rubber piece there is made of better stuff than the similar part I bought for my VW T25 a few years ago which turned to powder after about six months.

I did however stick a hose clip on the breather hose inside the car.  Hoping that this resolves the occasional stink of petrol in the cabin on hot days...as I honestly don't know where else it can be coming from.

Connecting the lines up was downright child's play compared to trying to convince the breather hose to go where it belonged...Let's hope I'm still saying that when I've got the pump running!  Haven't bothered with the cable ties yet (to keep them clear of the brake bias control actuator rod and the nearside shock absorber), want to make sure we don't have any leaks first.

Then discovered one slight hiccup when I came to put the petrol back in it...in that I couldn't track down the spout to the fuel can.

Not one to be discouraged I improvised.

Worked like a charm...even managed to avoid getting a mouth full of petrol this time!

No leaks from the filler (thank goodness) in evidence.  I'll need to figure out the connections to the fuel pump and sender tomorrow, tidy stuff up in the engine bay a bit, then hopefully wind up with a car that still works at the end of the day...

Actual technical difficulty of that job?  At worse a 2/10.  Sheer bloody awkwardness: 8/10.

What a bloody faff.

[Zelandeth]

Well...had a number of things I'd planned to get done today...However this ended up summing up the actual day.

Bloody weather.

[Zelandeth]

Figured that I really ought to leave these videos here for you guys from when I was doing the drive system tests on the Invacar...

Firstly a general view looking forwards - not quite capturing the earth-shattering noise levels in the garage though!

Secondly with the cover removed and looking at the belts as the movable cones do their CVT dance...

Sorry I totally forgot to upload these at the time!

[dollywobbler]

Gosh your engine (probably exhaust related) sounds much smoother and quieter!

[Zelandeth]

Gosh your engine (probably exhaust related) sounds much smoother and quieter!

I have noted that this appears to be a different exhaust to the one you have (there's no turn at the tip), not sure whether this belies any internal difference in construction that might have an effect on noise levels.

 

It does sound astonishingly happy mechanically speaking, and I have managed to lose a good few bits of time just listening to the thing burbling away at idle.

 

Fair to say it sounds better than it looks!

[dollywobbler]

I can do that now! After clearing out the carb and turning the idle up a bit, it will actually smoothly idle for as long as I want it to. This is some novelty.

[Zelandeth]

I have to admit to having been astonished how much better my one ran once I turned the idle speed up a little bit as well.

As mentioned, the exhaust on mine looks very slightly different to the one on DW's Invacar.  Not sure whether this is purely external, or if the internal structure may be slightly different too.

She's got a bit buried this week with me concentrating on sorting the Lada out!

Hopefully normal service will be resumed soon there.

Today however I mostly got the Lada finished off.  Crawled around underneath and removed the wiring to my old fuel pump and rerouted it into the back of the cabin.  Then spent a while playing detective while working out which terminal was what on the fuel pump/gauge plug.  I'd failed to notice that this being a later part has a different socket to the Niva one I was used to.

At this point my fancy digital multimeter decided to for reasons unknown, turn itself into a doorstop.  This was unhelpful, especially as I couldn't find any of the dozen or so "cheap 'n' Nasty TM" yellow ones that I know are scattered around the house for moments such as this.  So wound up hauling the old AVO 8 down to the driveway.

I seem to keep ending up using this thing working on the cars...

Everything connected up, and it seems to work.  Oh...my...word...it is so quiet compared to the old one, which used to whine away in a really annoying way.  No evidence of any fuel leaks, and the car seem to be running smoother.  Hard to say if that's because the fuel pressure delivered is now what it's expecting, or if it's because I've poked and prodded a few things in the engine bay and sorted a couple of potential vacuum leaks (which are the bane of this injection setup, even the smallest of leaks can cause it to throw a major tantrum).  Not really going to know for certain until we get it on the exhaust gas analyser, but fingers crossed it's running properly in closed loop mode and such.  It seems to smell right at least.

Just need to get a couple of last hoses changed in the engine bay (more in the interests of tidiness than anything), and try to sort out a better bracket arrangement for the throttle linkage to cable joint (cable ties are currently a critical component), then I'll call that side of things done for now I think.

Hopefully get it booked in for an MOT in the next week or so...

[Zelandeth]

Just a quick update for today as I suffered from a severely inefficient afternoon where I couldn't find anything.

The Lada is running far better.

Blipping the throttle before resulted in the engine bogging down, then revving up slowly like an old N/A diesel.  It's much more responsive now, and no longer produces any smoke when the pedal is prodded.  Hopefully this will result in a nice screen of green rather than red numbers on the emission tester next week.

...Also, I don't think I'll be returning the middle silencer to the system.  It's currently got a straight bit of pipe where that was after it disintegrated one day (I kid you not, it pretty much exploded), as the soundtrack seems much more in fitting with the general design era of the car.

She really does need a new system from the cat back though...it's very much showing that the car spent most of its life doing short journeys, and is just knackered.

[Zelandeth]

Sad day...after six years on my fleet, the Saab has now moved on to pastures new.

Sounds like she's got a good shot at being restored to her former glory though, so that feels better than it might.

First time I've not owned a Classic 900 since 2004...

[Zelandeth]

First show/event of the year done!

Okay, first successful event of the year...attempted the Stoney Stratford one on New Year's Day, but never even got close to it as it was so overcrowded.

Nice little get together at the British Motor Museum, the a convoy run over to Bicester Heritage. Had *no* idea how much stuff was buried in there!

Was really surprised to find this very forlorn looking little Visa hiding in a corner by a hanger though.

Don't worry though, a couple of other concerned Citroen enthusiasts have already made sure to ask the relevant questions, so hopefully it will have a bright future - it's parked on the right estate for it, that's for sure.

[Zelandeth]

MOT time for the Lada tomorrow.

Had a look around today, and save for a missing exhaust hanger don't see any definite issues. Don't have any of the right sort laying around so will deal with that tomorrow.

Have cable tied all the loose fuel lines into behaving (have tethered them to the vent line which is still firmly held in by the original fasteners), and made sure all the lights are working, there's water in the washers etc.

Also did some very period accurate bodgery to deal with the ragged edges of the rusty wing tops.

... it's the early 1990s all over again!

New wings will be going on soon. Had hoped to get a body shop to do it, but don't have the odd £1500 to throw at it just now...so will get myself a spot weld drill and just get stuck in I guess...

Wish me luck tomorrow.

[Zelandeth]

FAIL!

Not epic fail though.  Emissions are high as she's obviously running rich.  Stable though rather than wandering all over the place, so first suspicion is coolant temperature sensor having drifted. 

First stop:

First time on the road since early October though and you wouldn't know it to drive the car.

[Zelandeth]

So this afternoon's task was to figure out what the heck was going on with the emissions on the Lada.

This isn't actually as much of a trial as you might expect, not least because the throttle body injection system is really well documented, and the documentation is available from the Lada UK website over here.

Obviously some tasks assume that you've got a LadaScan/Tech I diagnostic tool (it's an ALDL system, not OBD), but it takes you through a lot of diagnostic tests and gives information which can help you come to logical conclusions even without anything that specialised.  Obviously though, one of those is very much on my wish list, and I periodically poke eBay in the hopes of finding one.

I already know that my ECU has some form of issue because the output (well it's actually a switched ground...but I'm calling it an output here for clarity's sake!) for the check engine light doesn't work.  This is particularly annoying as that's involved in reading out stored fault codes and also allows some degree of engine-running diagnostics to be done...Without a working light though, you can't do any of that.

The first thing I did given that the car was obviously running rich was to check the coolant and air temperature sensors, they're easy to get at, measure and there's a table in the documents.  While the values were a little out, they weren't unreasonable and even the manual states in bold letters that all values are approximate...so probably not that.

Next step was to have a look at the lambda sensor signal to give me some idea whether the car was running in closed loop mode at all.

Conveniently the ECU connector pins are accessible from the back of the connector, and are even labelled. 

Pins D6 and D7 are the ones I'm interested in, D6 being the sensor ground and D7 the signal.  Wait...What is going on here...This is the reading from D6...which should be near as dammit 0V.  It's not even a "notional" ground, it should just go via the ECU then straight to the earthing point down by the coil pack, which is bolted straight to the block where the distributor used to be.

I checked with an analogue meter as well just to make sure this wasn't some meter glitch...Nope...It's resolutely sitting at 1.5V...Essentially meaning that there's a 1.5V offset on the lambda reading.  Given that this signal is only meant to swing between 0.1V and 1V, it goes without saying that it's unsurprising that the system is struggling.

A bit of further investigation revealed that this voltage was appearing courtesy of the ECU itself rather than a crossed wire further down the line. 

Nothing really to lose at this point as the essential summary is basically "the ECU is stuffed" I figured I could externally ground that terminal and see what happens.  The answer is that the sense line latches up at around 0.7V and the car runs noticeably worse.  I'm not seeing any sign of a proper lambda signal at all, disconnecting/reconnecting the sensor has no effect on running whatsoever and the voltage on the sense line is a solid 0.7V.

Balls.  I really need a known good ECU before I can chase this much further I think...

The thought occurs to me that I may well have just found the fault which originally took the vehicle which I got the injection kit from off the road!

[PhilA]

Does the ECU have any form of data output at all? (Or is that what the scanning tool looks at?).

 

If it does, is the data stream documented?

 

Phil

[Zelandeth]

Does the ECU have any form of data output at all? (Or is that what the scanning tool looks at?).

 

If it does, is the data stream documented?

 

Phil

Yes there is, there's a serial data port (ALDL protocol, essentially the predecessor to OBD), and I'm pretty sure it is reasonably well documented. Building a PC interface has been on my to do list for a while, as the EFILive package can then read real-time data out.  The basic ECU output if you don't want a detailed analysis, is done via the check engine light.  You can read out error codes via that with the engine off, and put the unit into a diagnostic mode and read some basic running information about how the emission control system is working via it as well.  However, the check engine light output is dead on this ECU.

 

At present I don't have any way of reading said data, but I can check it's there at least...not too hopeful of it doing me any good anyhow given there are pretty clearly issues with the analogue side of the ECU itself.

 

 

A theory springs to mind.

 

I know that Lada bought the injection system in from GM as a complete unit.  Now GM aren't going to have built an ECU to order for Lada given that in their grand scheme of things it would have been a very low volume order.  No, they will have used an off-the-shelf unit, simply configured to suit the car.  In the case of this ECU that's done via a plug-in board referred to as the "Vehicle Identity Module" in the documentation.

 

Having done a little bit of rummaging, the ECU used in some variants of the Vauxhall Corsa B looks physically identical.  Same connectors, same cover at the top of the unit (where the aforementioned identity module slots in) etc.  Haven't actually managed to find a diagram showing me what pin does what on the connector as is so helpfully listed in the Lada documentation, but I've a sneaking suspicion that these are physically identical units...

 

What are the odds of me being able to get hold of a Corsa B ECU and swap the VIM over and it working do folks think?

[PhilA]

Highly likely.

[Zelandeth]

Okay, given that they were available for £20 delivered, I've grabbed one of the identical looking Corsa ones.  Should hopefully be pretty obvious if they're radically different once the cover's off.  They're both from a similar era, both use the ALDL comms protocol and are both from GM vehicles, so hopefully luck will be on my side.

 

Difference is that Corsa B ECUs are easy enough to come by.  Lada Niva ones aren't...hence the only one I could find on eBay today being listed by a guy in Germany who's asking £170 for it.

 

Fingers crossed this will work...If nothing else, it will be useful to know whether it's a viable mod for others with a duff ECU.

[PhilA]

Reason I say that is my Renault uses all GM sensors, they're a late 80's staple.

All the thermistor curves are the same, the voltage.. only thing different is the protocol on the serial output.

 

 

Phil

[DodgeRover]

Is there an immobiliser built into the Corsa ECU?

[Zelandeth]

Is there an immobiliser built into the Corsa ECU?

Unknown... entirely possible given that it has control over ignition, fuel pump and injector pulses. I'd expect the "enabled or disabled" state of that feature to be in the configuration module though.

 

I guess we'll find out soon enough!

[Zelandeth]

Balls.

 

The ECU I've just had delivered is totally different. Same connectors, same case...but totally different board within, most critically, lacking the place for the plug in config module.

 

Back to square one.

[Zelandeth]

Well it looks like in a massive stroke of luck that an ECU may have been found by a friend.  Not going to get my hopes up too far until I actually have the thing in my grubby paws, but fingers crossed...

I'll be doing a thorough wiring check before it goes in as well just in case it turns out that the old ECU was killed rather than dying of natural causes.  Last thing I want to do is fry the new one, given that they're apparently made of Unobtanium these days.

I will have a shot at repairing the original one though.  I'm pretty certain that the fault will lay among the transistors etc which lay between the digital innards of the ECU and the analogue outside world.  However without meaningful part numbers or a known good example to compare to it's near impossible to really fault find.  With a working example to compare to and some patience however I should hopefully be able to work some of it out.  If I could just find a damn schematic it would be easy...

So, current Lada project status: Cautiously optimistic...

Here's a question for the Vauxhall/GM experts:

Anyone recognise this ECU as the one that was used in any particular models?

This case style:

This this sort of module under the top cover:

This is what it looks like with the whole case removed:

Anyone have the foggiest idea what this was used in aside from Ladas?  NOT the Corsa B apparently...The one I got for that is externally identical, but has a totally different setup internally, lacking in the space for the configuration module.

[dollywobbler]

Were these ECUs only used on UK Ladas?

[PhilA]

There's not much on that board that should age badly. Tried running it and hitting it up with freeze spray?

 

It could be something as simple as the ROM for the table has gone bad. What's behind the blue plastic?

 

Phil

[Zelandeth]

Used here in the Niva 1.7 and I believe in some Scandinavian regions where the same power unit was also used in the Riva for the last couple of years. Keeping my eyes out for anyone who's broken one as I need the vehicle speed sensor pickup from one ideally.

 

So one model and for 1995-1998 I believe. After that they switched to the multipoint setup for the domestic market.

 

The fault with the ECU I have seems to be solid, not affected by hot, cold, shock, pressure on the board etc. I don't *think* it's a ROM data issue as the errant voltage sitting on the lambda sensor input pins is there the moment power is applied irrespective of whether the ROM module (which is the blue plug in cartridge - an EPROM and resistor network) is plugged in or not.

 

One of the first things I'll be doing once I have a known good one here however is taking a backup of the ROM data though...It is a UV erasable type I believe, so bit rot is a potential issue as they get older...

[Zelandeth]

Nothing specifically car related done this weekend, but a nice little workbench upgrade has been installed.

This replaces the old Telequipment D61a that I picked up nearly 20 years ago when I first got into electronics. The new Hameg is still pretty ancient, but has some basic storage abilities. The lack of that was probably the biggest drawback with the old one where dealing with pretty much anything digital. It's not a modern DSO by a long shot, but is a big step forward. Especially for less than fifty quid.

Will look to get the old one moved on to someone else who's just getting into electronics as I was.

[somewhatfoolish]

Snap! I've got a HM205 in a cupboard somewhere; I was trained patchily in the use of silly scopes 20+ years ago using HM 205 and bought one off ebay in the days when you posted people cheques for £20. Not used it in ages but for some things nothing else works. I iz wel jel ov yore dedicated printer tho.

[GrumpiusMaximus]

Does anybody preserve ECU firmware?  I'm just interested.  I'd imagine an adjustable ROM reader and some software could mean that the firmware is saved for the future and potentially adapters made for generic chips?