Bosch LSU 4.9 sensors dropping like flies

dr_jones
Posts: 162
Joined: Wed Jun 08, 2016 3:24 pm

Bosch LSU 4.9 sensors dropping like flies

Post by dr_jones »

So I've been using the Bosch LSU 4.9 wideband sensor with the S6Plus since the end of last year - the guys at Syvecs set up the calibration for the sensor for me (thanks!!) - as its been completely trouble free until last week where I had my first LSU 4.9 fail on me (output went high / lean AFR - it was something like 12 months old)

Then weather got much colder (UK) and I saw the first real issue with my fuelling cal - ie way too much fuel fuel correction for low ECT - and way too much preliminary fuel injection - managed to flood the engine and in the process potentially destroyed the second LSU4.9 - well initially it seemed OK but it died maybe 10 minutes of light load running later after I'd resolved the flooding condition / startup fuelling calibration.

Number 3 LSU 4.9 initially looked good - got 40 minutes out of that this morning - checked it was functioning (just a free air check) midday and then it failed maybe one minute into my journey home (engine not even up to running temp.

Now I'm in something of a fix - having used my entire stock of new LSU 4.9s in a couple of days I'm hesitant to just throw in another and have it fail immediately - these things aren't the cheapest!

Initial ideas I have:
- Sensor isn't reaching operating temp fast enough in the colder weather (problem with the sensor cal / problem with the S6Plus LSU4.9 control itself)
- Wiring issue - using a JTinnovations supplied cable - so decent quality - but the connection to the LSU 4.9 is under the car up by the side of the gearbox - no obvious crappage on it but I guess water ingress *is* possible.
- Position of sensor - its at around 11 oclock but would be protruding very slightly into the gas flow - also it *is* further than might be optimal from the turbine housing (approx 90cm).
- Batch of duff sensors - unlikely

I'd be interested if anyone else has experienced similar or might have any further ideas what on earth could be causing this....
pat
Syvecs Staff - Cleaner
Posts: 356
Joined: Fri May 23, 2008 10:23 am
Location: Out there... somewhere
Contact:

Re: Bosch LSU 4.9 sensors dropping like flies

Post by pat »

It would have been good to see datalogs of these failures, in particular it would be good to see what the LSU temperature reading was and also the heater drive. If you do have that sort of log then I'de be interested to have a look.

If you are concerned about destroying another sensor, you could consider using an NTK UEGO instead. These have been in use for longer and have a good record of reliability.

Regards,

Pat.
dr_jones
Posts: 162
Joined: Wed Jun 08, 2016 3:24 pm

Re: Bosch LSU 4.9 sensors dropping like flies

Post by dr_jones »

Thanks Pat,

I'm afraid I don't have a complete picture (annoyingly i'd just moved to not capturing data < 3000 rpm to try limit the amount of captured data) and I wasn't capturing LSUFeedback_U03V or LSUFeedback_U03 or Lam1R (which I *think* would be the appropriate data (?).

What I have is here:

https://dl.dropboxusercontent.com/u/202 ... -01-05.zip

I'm not sure the data is enough to draw a solid conclusion - but would certainly appreciate your giving it a once over.

There's certainly some odd behavior in terms of AFR in the penultimate log at closed throttle (could well be linked to my overrun cutoff not working - separate post).

-Ed.
dr_jones
Posts: 162
Joined: Wed Jun 08, 2016 3:24 pm

Re: Bosch LSU 4.9 sensors dropping like flies

Post by dr_jones »

Have in addition added the missing LSU related channels and captured intial sensor warm-up to engine start:
2017-01-12 09-49-09 plus 0h00m00s S6#1628.SD
(271.46 KiB) Downloaded 958 times
2017-01-12 09-49-09 plus 0h00m00s S6#1628.SD
(271.46 KiB) Downloaded 958 times
2017-01-12 09-49-09 plus 0h00m00s S6#1628.SD
(271.46 KiB) Downloaded 958 times
Again - not certain of the usefulness right now - this is after-the-fact - on a now seemingly dead LSU4.9 (the most recent one)
Attachments
2017-01-12 09-57-43 plus 0h00m00s S6#1628.SD
(280.76 KiB) Downloaded 855 times
2017-01-12 09-57-30 plus 0h00m00s S6#1628.SD
(259.71 KiB) Downloaded 844 times
dr_jones
Posts: 162
Joined: Wed Jun 08, 2016 3:24 pm

Re: Bosch LSU 4.9 sensors dropping like flies

Post by dr_jones »

...I also managed to snap the moment the sensor died (prior to me logging) with my phone...
Attachments
Lambda failing.JPG
Lambda failing.JPG (25.54 KiB) Viewed 30502 times
dr_jones
Posts: 162
Joined: Wed Jun 08, 2016 3:24 pm

Re: Bosch LSU 4.9 sensors dropping like flies

Post by dr_jones »

Still struggling with this - I'm loathed to swap over to NTK sensors without getting to a sensible conclusion.

In terms of values that I *should see* can anyone help me out - right now all Lambda related data is flatlined at the following:

LSUFeedback_U03 = 100
LSUFeedback_U03v = 4.993

Lam1 = 15.3
Lam1R = 657
Lam1V = 1.537

Which suggests that the heater is now permanently being driven at max - (previously this was not the case) and that the sensor is giving a reading albeit a meaningless one.

Connections to the sensor all ring through so I'm struggling to know where to go next in terms of troubleshooting.
dr_jones
Posts: 162
Joined: Wed Jun 08, 2016 3:24 pm

Re: Bosch LSU 4.9 sensors dropping like flies

Post by dr_jones »

I've spent more time looking at this and found the datasheet relating to the Bosch Lambdatronic LT4 controller to be of help.

I got to suspecting it was no coincidence that my sensors began failing as the weather turned cold.

My theory is that with the colder weather comes additional liquid phase water in the exhaust and for longer so an increased risk of water getting on the cold ceramic surfaces of the sensor and causing premature failure....and to add that increased thermal shock as the sensor is quickly heated from a lower starting temperature.

Looking at the LSU 4.9 control I have on the S6Plus it does not look like there is any sort of dewpoint delay at all. So as the ign is switched on an the engine
fires the sensor's heater receives full heating voltage and starts rapidly heating immediately. I looks like a typical dewpoint delay for the LSU sensors is around 30s (and typically this delay is triggered below ECT of 50 degs) - during this persio the sensor is heated but at a lower voltage (rate) - what I don't know right now is exactly what that reduced voltage should be.

I can't readily see that I can modify the heater control output duty table to runTime (or indeed ad an ECT trigger) but I see that I can swap out vbat for ECT and so create a calibration that heats the heater at a lower rate as ECT is low and does not allow full heating power until the engine is warmed and so hopefully reduce the risk of water ending up on the hot sensor.

So the heater output duty table becomes something more like this:
Attachments
heater_control_ect.JPG
heater_control_ect.JPG (81.09 KiB) Viewed 30439 times
TimH
JT Innovations
Posts: 718
Joined: Wed Sep 01, 2010 3:51 pm

Re: Bosch LSU 4.9 sensors dropping like flies

Post by TimH »

Not sure it's a good idea to omit the VBat correction, otherwise the heater volts will be wrong.

LSU4.9 heater should be ramped up at 0.4V/s and shouldn't really be started until the engine's running; and limited to an effective 2V during the condensation phase...so the critical question is how long that is. I've used 8s in my expander, but maybe 30s is more realistic and not a big deal for a performance car where start-up emissions are less critical.

Probably best to speak with Pat as he may be able (and maybe even willing!) to rework the GP+ firmware to sort this.
dr_jones
Posts: 162
Joined: Wed Jun 08, 2016 3:24 pm

Re: Bosch LSU 4.9 sensors dropping like flies

Post by dr_jones »

Thanks Tim - thats useful info on the heater initiation - I'll have a word with Pat and see what f/w related possibilities there are

I'm not sure I actually ever had actual working vbat correction - I should have really posted a before and after - but below is the **before** showing the vbat against LSUFeedback_U03 - I had assumed vbat was only in there are *something* had to be!
Attachments
heater_control_vbat.JPG
heater_control_vbat.JPG (299.32 KiB) Viewed 30425 times
dr_jones
Posts: 162
Joined: Wed Jun 08, 2016 3:24 pm

Re: Bosch LSU 4.9 sensors dropping like flies

Post by dr_jones »

Just saw f/w version adds "'runTimeLR' which mirrors runTime but freezes at 300.00 seconds and is available for CM axis choices" - which is great(!) and should be a better choice that ect for heater initial warmup - though I guess I'll still need to dump vbat correction.
Post Reply