Jump to content

Zel's Motoring Adventures...Peugeot, Renault, Rover, Trabant, Invacar & A Sinclair C5 - 25/03 - Trabant back in action...


Zelandeth

Recommended Posts

Had planned to attack a few tasks today, but instead got utterly sidetracked by getting the most stubborn laptop in the world working. If I had any sense I'd have just tossed it back in the stack of broken stuff...but I'm stubborn. As such I've decided that it isn't going to beat me.

 

Looks like on OS install number six we're actually getting somewhere.

 

Tomorrow the cleaning supplies will be coming out I think. The tree has blown most of its pollen now, so hopefully I can turn the currently green cars back to red.

 

For now though, off to the cinema to see the re-release of 2001.

Link to comment
Share on other sites

Speaking of PWM and multiplexing...

 

Why on earth can't they run the systems in modern cars at a high enough frequency that the blasted flicker isn't visible? It drives me spare in the vast majority of modern cars. I really can't see any reason for it aside from lazy design...

 

Isn't it efficiency? As I understand it, the switching transistors are dissipating barely any power in the on or off states, but do lose some during the actual switching. So the more switching operations in a given time, the less efficient the operation. No idea if this is significant between visible and non-visible PWM though!

Link to comment
Share on other sites

It was just about bearable for me when it was just one or two things - but *EVERYTHING* in the current VAG cars we see through our door does it. ...and it's astonishing the number of things they've found excuses to stick lights in. Do you *really* need lights under the front seats, in the door handles (fine - have those light up when you turn the ignition off), or silly little lights shining from the roof down at the gear lever? ...and they ALL flicker.

 

Granted, that's usually item somewhere around 379 on the list of things I hate about them anyway, and I find any excuse not to travel in them.

 

Last November I seriously considered driving the Lada up to Livingston to avoid a six hour trip as a passenger in whatever we had at the time...

Link to comment
Share on other sites

In preparation for the mad activity hopefully to be getting underway over the next couple of weeks, I got the Invacar back off the axle stands.

 

IMG_20180604_165228.jpg

 

I then set about trying to get it started.

 

You may recall that not long ago it had managed to suck some disgusting brown sludge out of the bottom of the temporary fuel tank I've been using. Like a complete idiot, despite cleaning the carb out following this, I totally forgot about the fuel pump and the line between it and the carb. So as soon as I tried to start it today, it went and gummed itself up again.

 

After a somewhat protracted battle - including a *proper* backfire that probably scared our neighbours half to death - I eventually managed to get it going. It'll idle sweet as you like, but any attempt at touching the throttle makes it die.

 

So will be cleaning the carb *again* tomorrow.

 

I'd really prefer to be able to back it out of the garage and into a suitable corner to get the work done rather than having to faff about pushing it...so more on that tomorrow.

 

In other news, a new oxygen sensor is on the way for the Lada. Fingers crossed that will resolve most of the running issues... we'll see.

Link to comment
Share on other sites

AARGH!

 

New oxygen sensor fitted to the Lada.

 

Check engine light immediately pops up the moment it drops into closed loop. Errors logged...13 and 34 still. 13 being the sod as that's "no oxygen sensor signal".

 

The sensor is brand bloody new. The ECU is known to come from a working car. What the heck?!?

Link to comment
Share on other sites

There's a wire between the sensor and the ECU.

There are four of them in fact, and they all test perfectly.

 

Looking at the voltage on the sensor, it's hovering around 0.8-0.9V - this should be telling the ECU "WAAAAY rich, reduce fuelling..." but it isn't being interpreted as such.

 

Reverting to the age old mantra of electronic stuff..."Thou shalt check voltages."

 

Usefully, the documentation has a nice chart telling you what to look for pin-by-pin on the ECU connectors. Everything checks out - with the exception of the pins for the oxygen sensor - in particular the oxygen sensor ground, which the documentation states should be directly passed through to the ground point on the engine block via the ECU ground wire - this is floating at 1.51V with relation to chassis ground. There is also conflict between diagrams as to whether this should be a brown or brown/white wire.

 

This is actually the very symptom which had me condemning the original ECU (that and the non-functional diagnostic system)...

 

I need to do some research...

Link to comment
Share on other sites

Sadly not. Nivas - even late ones with injection - are very thin on the ground these days.

 

The injection loom is self-contained though so it shouldn't be impossible to figure out, just going to take a lot more pain than I'd hoped.

 

First thing I'm going to check tomorrow is if there's any potential grounding issues - while it *seems* fine at the ECU there are several points where things come together in terms of grounds, and I'm not discounting something back feeding.

 

If an electrical fault makes no sense, it's probably a dodgy earth after all!

 

I'll update tomorrow, probably will include a few diagrams and measurements to see if you folks think I'm speaking rubbish.

 

I'm bloody sure that once this odd voltage offset is sorted that we'll have it sorted.

 

Currently we have around 0.8-0.9V *across the oxygen sensor* - that indicates a very rich condition - correct.

 

However because the reference that's being measured against is effectively offset by +1.5V, the ECU isn't seeing that... it's seeing garbage, hence flagging up the code 13 (no oxygen sensor signal) as it's totally out of range.

 

Then there's the code 35...but one headache at a time! Though if there's some bloody backfeeding earth problem I wouldn't discount them being connected.

Link to comment
Share on other sites

I should have taken a before photo for that one above. It was sitting at about 7-8% CO at idle, and would go full scale rich at fast idle (it was measured at just under 5% at the MOT). The meter gets slightly confused by the extra water content in the exhaust of a cat equipped car so reads slightly high, but it's still a useful guide.

 

So it's easy to see the difference today!

 

The real key to me cracking this - he says not having the MOT in his grubby paws yet - was a bit of background from a Lada expert. The key bit of data was that Nivas apparently were often an absolute swine to get to run right with *any* non-Lada lambda sensors, and usually due to grounding issues. This set the little grey cells going...

 

Now, the original cat equipped Rivas (and the early cat Nivas before the TBi system came in) had a three wire lambda sensor. Two pin plug for the heater, one flying lead for the sensor signal, signal ground through the case. When the TBi system came in, this changed to a four pin one. Now, my theory here is that Lada, being...economically minded...decided to continue using the same tooling, just adapted to suit the four wire design, with the sensor ground still *also* tied to the case.

 

This wouldn't be a problem generally. It's a needless "double ground" connection which is bad practice, but it should work. However here's where it gets interesting...as far as I can tell, the sensor ground pin on the ECU is NOT properly ground terminated...it will float at 1.51V with relation to chassis ground. Now keep in mind that the signal we're looking for is only between 0.2 and 0.9V...so adding a 1.5V offset to that is just going to result in it being interpreted as garbage by the ECU. This will happen with almost any non-Lada oxygen sensor because the sensor ground (pin C) there is fully isolated from the case...so there is no ground path!

 

What I've done today was to connect pins D6 (oxygen sensor ground) and A11 (CTS & TPS signal ground) at the ECU together. This correctly ties the oxygen sensor ground level to the signal reference ground...and suddenly the ECU can actually see a meaningful signal!

 

Previously the voltage across the *sensor* was floating at around 0.8-0.9V, indicating that the mixture is well rich...but the ECU couldn't make sense of it...as it was effectively seeing -0.6V. Totally out of range, to the extent that the ECU was just giving up and declaring that there was no signal at all.

 

The moment of truth for me was watching the sensor voltage as the car warmed up, and suddenly when it dropped into closed loop mode and the voltage plummeted down to 0.2V and then wandered around as the system regulated itself...and the exhaust gas reading speaks for itself. Fast idle now will see it peg itself at the lean end of the gauge rather than full scale rich.

 

We'll see what the MOT gas analyser says tomorrow...

 

No idea whether this logic is sound...but it seems to make sense.

 

I do have to wonder how many Nivas have been written off due to emission issues because of this issue - especially as the OEM Lada oxygen sensors have been unobtainable for some time now.

Link to comment
Share on other sites

We have an MOT!

 

As for the emissions - here's a before and after of the lambda sensor shenanigans.

 

Before:

IMG_20180608_114748.jpg

 

After:

IMG_20180608_114445.jpg

 

Think those results speak for themselves...and finally justify why I embarked on this project!

 

... especially with the pass threshold having been reduced to 0.2%CO in May...trying to get it under 0.3% with the carb setup was always hard enough. Last one was 0.297%!

 

Now to sort proper insurance, get it taxed, then I can actually drive the thing!

Link to comment
Share on other sites

Had an FTP today!

 

Stopped by the local hand car wash after being out running errands. Planning to hit it with the polished soon, but wanted to blast the worst of the gunk off as it was getting hard to see out of the thing.

 

Even managed to stop them using the horrible spray cleaner as it attacks the paint.

 

IMG_20180611_183010.jpg

 

Then while they were vacuuming it out, it cut out and wouldn't restart.

 

Was shoved out the way so they could get back to their work while I sorted it out. Was quite obvious it wasn't getting fuel.

 

Was quickly tracked to the fact that they had nicked the wire that runs to the fuel pump, which I had totally forgotten was still trailing under the passenger seat. This had yanked the wire out of the crimp connector. This had then shorted out on the seat bracket, blowing the fuse.

 

I twisted the wires together, stuffed a spare fuse (which I had in the boot of course!) in to get me home.

 

I'll get that properly rerouted tomorrow...had honestly forgotten it was still under there. Moderately embarrassing, but easily sorted at least. Plus this is what local shakedown runs are for.

Link to comment
Share on other sites

New speed sensor for the Lada has arrived.

 

IMG_20180612_163554.jpg

 

Has been fitted (monumentally fiddly due to minimal space between the side of the gearbox and bodywork), but not wired up yet. Sadly the loom leading to this had been mangled on the donor vehicle, so I need to figure out which pins are what. I've run a cable from the sensor up to the engine bay though so I'll be able to get it connected once I work out the pin connections.

 

Annoyingly they're not numbered on the connector unlike most Lada sensors.

Link to comment
Share on other sites

Today a small trailer has joined the fleet.

 

IMG_20180613_172125.jpg

 

This will prove hugely useful in allowing me to catch up with our garden work backlog as I'll finally have a way to get the trimmings to the tip other than one bin every week.

 

It will also get a bit of a tidy up. Some fresh paint and varnish will make quite the difference. Electrics will benefit from an overhaul too as the grounding seems less than stellar just now, though all the lights do actually work. That puts it well ahead of 90% of the ones I see on the road!

 

It's actually a lot better mannered than the last couple trailers of this size I towed and didn't seem too interested in bouncing around much even when empty.

Link to comment
Share on other sites

They were indeed, as were the Xantias in the 90s - though I reckon that's probably more down to their self levelling ability than damping.

 

It wouldn't matter if you stuck a load of weight on the tow ball of a Xantia or BX, it would still ride at the correct height - unlike a conventionally sprung car which would be dragging the rear mudflaps along the ground. That concept being taken to extremes with the CX Loadrunner of course...

 

It still provides me with amusement watching the suspension on the Activa react to the weight change when passengers get in or out of the car.

Link to comment
Share on other sites

So, having wired up the speed sensor in the Lada, went out for a test run. Two roundabouts in...indicators stopped working again. Argh!

 

Instrument panel out, and sure enough they started working again...squeezing the flasher unit would make them start and stop. Pulled it out and had the cover off, sure enough there were several dry solder joints on the PCB.

 

IMG_20180615_130531.jpg

 

Resoldered all of them. It now appears to work reliably.

 

Also have stuck the stove pipe between the exhaust manifold and air intake back on, and cleaned the windscreen washer jets again as they clogged up yesterday.

 

Will go for a longer test run later and see if we've got rid of the sporadic code 24 popping up whenever you coast anywhere.

Link to comment
Share on other sites

Reed switch with coil and parasitic ferrite feedback antenna?

 

Phil

These old Russian flasher units seem ridiculously over complicated... I've never actually traced the circuit out...I really should some time. Precisely what the reed switch is doing I've no idea off the top of my head.

 

At a total guess they're using the coil around it as a current shunt, causing the contacts to close to provide feedback to sense that the correct amount of current is flowing...so blown bulb detection. Surely there's an easier way!

 

I don't think I've ever seen two of these that are identical either.

Link to comment
Share on other sites

Without using semiconductors? There's only 2 in there that I can see. Clever design using simpler components.

 

Phil

True enough... trying to design that in without adding any more transistors does get tricky. I get the impression it's been set up to replace a thermal flasher somewhere in the past too. The connections to it are a permanent live, ground, load out and dash lamp out.

 

This does also have the side effect of meaning that the hazard switch is hilariously complicated and has something like 12 terminals. That you do *not* attempt to take to bits if you ever want it to work again. I did that when I was about twelve, I found the last couple of bits that flew out in all directions when we moved out about twenty years later.

 

Just glad I've got a proper one in there now, the modern replacement (aside from going nuts when the headlights were on) was way, way too quiet...you can actually hear the tick now. Well...clunk more like. It's bolted to the dash support, so you can hear the *bonk, thunk, bonk, thunk...* reverberating through the bulkhead from outside too...

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
×
×
  • Create New...