Another step forwards in solving the problem...now, Lambda sensors info please?

[Fanum]

I spent the day with Neil yesterday trying to get to the root of the odd and erratic problem with my 1150GSA Twin spark.

THIS is the last thread, in which it was concluded that compression formed a major part of the problem.

So yesterday, after a long and very unpleasant ride to Roehampton from Harwich with only one cylinder firing, no idle, coughing and farting (the bike as well ) and random stalling, we took the LH rocker cover off and had a poke around.

The valve clearances were WELL out...although Neil had set them correctly when he stuck the second hand head/valves/piston and block in a few hundred miles ago, the clearances were just way too large.

Good news, because a) it would account for the compression on that side of around 6.5 bar, and b) it means that the valves weren't opening much (if at all) which hopefully means no damage to the piston crown etc.

So out come the gauges and Neil set the clearances up again......out comes the compression tester and BINGO!......12.5 bar...awesome!!

a quick check on the valves on the RHS and although they weren't as bad as on the left, they were too loose as well, so those got tweaked up pronto....the only other thing that may be relevant that we found at this stage was that the main plug on the left hand side looked perfect, but the one on the RHS was fouled and wet (the opposite of what we were expecting )

Anyway......all refitted, oil topped up, bike fired up and instantly sounded a LOT smoother, revving nicely and but idling still hunting about and farting a bit, with an occasional pop/backfire

Out come the harmoniser and laptop to set TPS, and it wouldn't balance up, even under the guru's touch.

Another screen on the GS911 showed that the Lambda sensor is flatlining.....no response whatsoever on the graph, and 180 or so faults on it.

(Bike currently fitted with original headers, original cat and stubby, with original lambda probe fitted)

Next step.....lambda plug disconnected , so fuel profile should revert to a general setting, 'Steptoe link' removed from plug in fusebox.

Bike then runs smoother on idle, less hunting, less popping.

It's late by now, and without a replacement lambda to hand and the bike running a little better than it was, I take off round the block....it is smoother up to about 30, but then fluffs out, no power, runs on one only and pops on overrun

After a quick discussion, there's nothing much else we can do so on go the panniers and kit and I head back to Harwich......top speed is marginally up to 65, it idles better in the 8 miles of stationary traffic I have to filter through on the M25 to get off onto the A12, but it's still a most unpleasant ride

Conclusion of the day....Lambda dead (probably poisoned by having been run for a couple hundred miles on one cylinder and excess fuel shagging up the sensor)

A search here last night finds this:

Could be the lambda - I'd rode my 1150GSA for 20k on a trip abroad, all fine, road it back from the freighters, all fine. Left it in the garage for 3 months then started riding it regularly again and it was an absolute BITCH. Stalling, stuttering, really flat and horrible especially about 3k, revs would just drop to nothing on clutch in. It was a really inconsistent though and an absolute pain. I changed the lambda (for the third time!) and it was a different bike.

and a couple of other posts that are remarkably similar to that, and my symptoms.

Ok, so I now have good (if not VERY good) compression, but the TPS is almost impossible to balance, and I have a flatlining lambda....

A new lambda sensor (with BMW plug) is over £100.....a TKN lambda with BMW plug is around £70, or a 'Universal' 4 wire lambda (18mm) is about £20 off Ebay.....

Splicing the 4 BMW wires on to the 4 wires on the 'universal' lambda is no problem...
according to the Ebay ad:

[FONT=Arial, Helvetica, sans-serif]The best way to determine if it will fit your car is to check the number and colours of the wires on your existing sensor. Basically, if your existing sensor has the same number and colour of wires, 1 Black 1 Grey & 2 white in this case, then you can be 95% certain that this sensor will do the job.[/FONT]

I just called Neil (who sounds like he's at death's door with his manflu ), but he looked at his stash of dead sensors and the BMW Wiring is the same....one grey, one black, two white

There are still some odd things happening with this problem.....when we swapped an injector for another for example, the bike ran perfectly for 80 miles on the way home, them progressively fluffing and losing power again until it was back to square one when I got back to Harwich....that doesn't make much sense.

The reason the valve clearances went out so quickly is also puzzling.

Fitting a new lambda is the next step....the compression has been cured, fitting the lambda will certainly help but may not cure the problem, but hopefully will allow us to set the TPS properly.

Any thoughts or comments on other aspects would be welcome though, and as soon as I get the new lambda from Ebay and get it fitted, I'll report back again

 

[Rich C]

Nothing wrong in principle with universal lambda sensors. These bosch 4 wire items on the GS are standard zirconia sensors that were used on virtually everything with Bosch engine management up untill a few years ago. The universal replacement approach works because the only difference is the connector and length of wire depending on the vehicle.

Just cut,solder and heat shrink. The whites are not polarity sensitive so either way round works fine. These two wires operate a heater within the sensor to get it to temperature quickly before the exhaust gases have had a chance to warm it up.

Motor factors will also stock universal sensors if that is easier than ebay plus you may get a better brand. There's some chinese shite out there as always as well as some good chinese product.

 

[Fanum]

Nothing wrong in principle with universal lambda sensors. These bosch 4 wire items on the GS are standard zirconia sensors that were used on virtually everything with Bosch engine management up untill a few years ago. The universal replacement approach works because the only difference is the connector and length of wire depending on the vehicle.

Just cut,solder and heat shrink. The whites are not polarity sensitive so either way round works fine. These two wires operate a heater within the sensor to get it to temperature quickly before the exhaust gases have had a chance to warm it up.

Motor factors will also stock universal sensors if that is easier than ebay plus you may get a better brand. There's some chinese shite out there as always as well as some good chinese product.

Thanks for that....I was wondering about the two white wires, but had assumed it would be something like that

Just ordered from HERE on ebay.....universal, corect size thread, correct 4 wire fitting, £17.75 delivered


I have some combined crimp/heatshrink connectors that I'll use, and the immense price difference between this and even a motorworks one is making me a happy geezer

(I looked up a BMW OEM one and some places seem to be charging over £400 FFS!!)

Once fitted, I have the option of returning to Y piece and Remus as well..Rosie likes that idea 'cos it sounds a lot rortier than the cat and stubby, but I'll see how it runs with the current setup and new lambda first


I'm not quite sure why disconnecting the dead lambda didn't return it to running on a set programme, but it's got to the stage where I'm removing known problems one by one and hopefully, eventually will have removed any possible causes and it will suddenly all come together

 

[Aidan1150]

What a palaver Bill. Stick with it though, it's bound to come good in the end.

 

[Steptoe]

Out come the harmoniser and laptop to set TPS, and it wouldn't balance up, even under the guru's touch.

, but hopefully will allow us to set the TPS properly.

The TPS is set correctly.

It's the throttle bodies that wouldn't balance up.

I
I just called Neil (who sounds like he's at death's door with his manflu ), but he looked at his stash of dead sensors

See, that's why you should never throw anything away

 

[Timolgra]

Conclusion of the day....Lambda dead (probably poisoned by having been run for a couple hundred miles on one cylinder and excess fuel shagging up the sensor)

Wasn't that the conclusion 6 weeks ago when Steptoe wrote ....

"Remember your lambda was flat lining on my lap top when it should have been peaking and troughing . I'd be looking in that area.

 

[Fanum]

Wasn't that the conclusion 6 weeks ago when Steptoe wrote ....

"Remember your lambda was flat lining on my lap top when it should have been peaking and troughing . I'd be looking in that area.

Yes, but the compression had to be sorted as well.....it's knocking things off the list one by one

 

[Smug]

The valve clearances were WELL out...although Neil had set them correctly when he stuck the second hand head/valves/piston and block in a few hundred miles ago, the clearances were just way too large.

Good news, because a) it would account for the compression on that side of around 6.5 bar, and b) it means that the valves weren't opening much (if at all) which hopefully means no damage to the piston crown etc.

So out come the gauges and Neil set the clearances up again......out comes the compression tester and BINGO!......12.5 bar...awesome!!

Resetting the valve clearances would not improve the cylinder compression by that much.

Wear in the valve train is the only thing that would cause the valve clearances to increase by a huge amount. If something in the valve train has lost its surface hardness you will find that the valve clearance will keep increasing until you run out of adjustment.

 

[Stewart H]

I have fitted a couple of the pattern Lambda sensors to various motors with an 100% success rate

 

[Timolgra]

Yes, but the compression had to be sorted as well.....it's knocking things off the list one by one

Has it been 'sorted' though mate? The valve clearance should have only altered by a few thou, if at all.

 

[Fanum]

Resetting the valve clearances would not improve the cylinder compression by that much.

Has it been 'sorted' though mate? The valve clearance should have only altered by a few thou, if at all.

I can't remember what the numbers were tbh, but the collets were visibly rattling loosely......the numbers on Neil's gauges are so worn he has to count in from the end of the pack to find the right one, but he may remember which one finally filled the gaping chasm.

As for not making much difference.......compression in LH pot before valves tweaked 6.5 bar...after tweaking, 12.5 bar....nothing else done whatsoever.....I'd call that pretty significant

And Tim, yes, they SHOULD have only changed by a few thou, but they hadn't.....the gaps would have intimidated Eddie Kid.

Why they had loosened so much is another question.....Having seen the way Neil works many times, and the fact that the bike was running fine initially after he'd put that second hand left pot/piston/head on it, I have no doubt at all that he'd set it all up correctly....he's too damn methodical to have not done it.

Whether a new oxy sensor cures it or just brings us another step towards a final fix is yet to be seen, but I have high hopes

 

[Smug]

I can't remember what the numbers were tbh, but the collets were visibly rattling loosely......the numbers on Neil's gauges are so worn he has to count in from the end of the pack to find the right one, but he may remember which one finally filled the gaping chasm.

I'd have thought that the collets could only be loose if the valve springs have gone soft or broken.

Important to identify why the valve clearances had changed so much. Cams? Followers? Push rods? Rockers?

 

[ChrisH Kerry]

Can't understand how tightening the valve clearances raises the compression - doesn't make sense Could it be a sticking valve/s? I hope you get it sorted - it's the sort of thing that could drive a man to drink

 

[Fanum]

Can't understand how tightening the valve clearances raises the compression - doesn't make sense Could it be a sticking valve/s? I hope you get it sorted - it's the sort of thing that could drive a man to drink

They were so loose they weren't being opened, so little air was getting in, so there was nothing to compress.

That's the theory anyway

 

[Timolgra]

They were so loose they weren't being opened,

Hmmmmm

 

[ChrisH Kerry]

They were so loose they weren't being opened, so little air was getting in, so there was nothing to compress.

That's the theory anyway

That's a new one to me

 

[Timolgra]

That's a new one to me

This is the latest vacuum engine.

 

[roddy]

....I'm not quite sure why disconnecting the dead lambda didn't return it to running on a set programme, but it's got to the stage where I'm removing known problems one by one and hopefully, eventually will have removed any possible causes and it will suddenly all come together

Would removing Fuse 4 or 5 (can't remember the exact one) help clearing out the memory of the ECU incase it's caught in some sort of loop ?

They were so loose they weren't being opened, so little air was getting in, so there was nothing to compress.

That's the theory anyway

Doesn't make a lot of logical sense that ?

plugs out, air has an easy path into cyclinder, fit compression guage, turn engine over see what the guage goes to ?

I'd have thought even if the intake valve only opened a fraction it would be enough to suck in the correct amount of air or you'd see the compression guage going negative/decreasing as a vacuum was created before the compression went positive ??

 

[ChrisH Kerry]

This is the latest vacuum engine.

Ah yes, the BMW R575GS

The clearance would have to be MASSIVE for the valves not to open at all, I rebuilt an SR125 some years back that had clearance so big the rocker arms had battered the tops of the valves - clearance was in mm rather than tenths of mm.... it had still run - sort of

 

[Smug]

I'm not pulling your leg but does the camshaft still have a complete set of full lobes?

I have seen engines where a lobe has worn down causing strange problems.