How to measure a Bosch LSU 4.9 wideband lambda sensor

Everyone loves a happy ending. Share your diagnostic success stories here.
KimAndersen
TwoWaves
TwoWaves
Posts: 246
Joined: Fri Nov 30, 2012 2:53 pm
Location: Denmark

Re: How to measure a Bosch LSU 4.9 wideband lambda sensor

Post by KimAndersen »

Hi FF

Under what conditions are these waveform capture taken - I mean what are the engine rpms during your test.

Your readings are more moderate compared with mine, but my readings are out of range, i think so.

It could have been nice to make a comparison between the TC versus the RTD sensor, but that not so easy, just to plug in a other sensor with those compact engine compartment there are in those new cars.

I will soon come with a update on my RTD sensor project as i came a little off track with regards to information i found about the supply voltage to the RTD sensor, it got me a little confused.

I must admit, there a lot more to account for when you must make those calculations on this sensor. :wink:

Regards
Kim

User avatar
Fat Freddy
TwoWaves
TwoWaves
Posts: 335
Joined: Fri May 04, 2012 4:52 am

Re: How to measure a Bosch LSU 4.9 wideband lambda sensor

Post by Fat Freddy »

Hi Kim

I'm posting up an image explaining the initial couple of minutes. But I'll see what I can do in regards of giving more info on what the engine is actually doing.

The whole run was about 15km's busy highway and a steady climb of about 1000ft
It could have been nice to make a comparison between the TC versus the RTD sensor,
That was something I was thinking about. The electronics may require some working out. At least it gives some comparisons.

I would like to see your results. I'm interested in seeing the differences in EGT between a modern diesel and my old tractor.
Explanation.png
Cheers
FF

KimAndersen
TwoWaves
TwoWaves
Posts: 246
Joined: Fri Nov 30, 2012 2:53 pm
Location: Denmark

Re: How to measure a Bosch LSU 4.9 wideband lambda sensor

Post by KimAndersen »

Hi FF

Here are my update regarding measuring of the exhaust gas temperature sensor (G235) which equals the Bosch PT200E sensor. What i did to verify that the G235 sensor was showing the correct values was firstly to find out whether or not the resistance/voltage values that i have used in my lookup table was the right ones.

If you are the happy user of the software VCDS is that a easy task. You simply look at measuring block in group 094 and at the same time, with a potentiometer connected to G235 sensor plug - the one that goes to ecu, can you now observe the exhaust gas temperature change when you turn the potentiometer up for more resistance.

Don´t start the engine.

From this resistance-test could I establish, that it was the correct values that i had been using in my lookup table. So what could cause the high exhaust gas temperature readings i was getting !!!.

The only thing that i might suspect is the lookup table and why so that, because the datasheet over the Bosch PT200 sensor has a resistance table with a interval of 50 degrees celsius - going from 100 to 900 degrees celsius.

What I did, was to create a lookup table with a interval of 5 degrees celsius from 100 to 950 celsius and that involves a lot calculations, which i did.

I must say, when i look at result of the this waveform capture, which is a acceleration test going from 2 gear 6 gear with a peak rpm at 3900 and a exhaust gas temperature at 730 degrees celsius, is it more likely to be correct.

The temperature at idle speed is between 140 to 150 degrees celsius and when i compare it to your old "Tractor" as you like to call it, would i say, that it is very close when we only look at idle speed temperature - that is interesting to think about. :wink:
ACC_TEST_2_6_GEAR_3900_RPM.jpg
ACC_TEST_2_6_GEAR_3900RPM.psdata
(210.15 KiB) Downloaded 684 times
Regards
Kim

KimAndersen
TwoWaves
TwoWaves
Posts: 246
Joined: Fri Nov 30, 2012 2:53 pm
Location: Denmark

Re: How to measure a Bosch LSU 4.9 wideband lambda sensor

Post by KimAndersen »

Hi there

I'm back with a update on how to measure a Bosch Wideband 4.9 LSU sensor on the road, while the original measurement was done with a standstill engine.

I wanted to see how the lambda readings were during a dynamic road test.

What should one expect on such a measurement when I have not been able to find any information that cover this subject significantly.The dynamic road test I did was conducted with the a car acceleration from around 20 km/t to 150 km/t, starting with the engine in 2 gear and ends in 5 gear.

The rpms reaches 4000 during this test.

What there is remarkable to me during this test is the lambda reading at acceleration, where it has a steady reading at 0 mA which equals lambda 1 or expressed as 14.5:1 air/fuel ratio in a diesel engine.Look at this picture where the circle to the left and right represent a gear change and then look at the readings between the circles which is the actual acceleration and this is where my expectations doesn't match the real world.
Dynamic_Roadtest_Wideband_Sensor_LSU49.jpg
I would have expected it to go in a rich mode.

The last two frames of this measurement are where the car coasting down from 150 km/t to standstill and in this period the accelerator pedal position equals 0 degrees.The lambda readings during this period are also 0 mA and therefore lambda 1 which does not correspond with my expectations of engine is drawing pure air through the engine without fuel being added.

I have expected higher lambda readings than 0 mA or lambda 1 at this stage. Pure air readings are around 2.54 mA.The only time the wideband sensor are showing maximum air readings are when the ignition are switch off.

I reckon that the lambda readings I´m getting from dynamic road are a part of diesel parcticle filter strategy (DPF) regeneration control whereby oxygen content in the exhaust gas are held at a minminum, especially during deacceleration of the engine.
Dynamic_Roadtest_Wideband_Sensor_LSU49.psdata
(985.79 KiB) Downloaded 568 times
Regards
Kim

Steve Smith
Pico Staff Member
Pico Staff Member
Posts: 1581
Joined: Sun Aug 25, 2013 7:22 am

Re: How to measure a Bosch LSU 4.9 wideband lambda sensor

Post by Steve Smith »

Hello Kim and to all, thank you for this continued feedback and all the posts.

These make for thought provoking discussions and challenging scope techniques for sure.

I hope the information posted here promotes the use of the scope when looking at broadband sensors as we can now see true response time and fuelling activity.(Not so when using a scan tool)

Not sure I have an explanation for the periods of inactivity of our broadband sensor when installed to these diesel engine vehicles but this is not uncommon.

I too was amazed when monitoring activity across a BMW 318d broadband sensor during climbing and deceleration, at how little response there was from the broadband sensor, so doubting the whole technique!

My thoughts surrounding the clear activity during gear changes would be as expected and indicated in your image above, where we have a sudden change in fuelling during momentary deceleration, (Lean) followed immediately by a gear change and then acceleration accompanied with the applied load (momentarily rich). Similar to how we shake a response out of a broadband sensor using repeated WOT techniques.

From my experience it seems easier to see the system lean thanks to over-run fuel cut, but difficult to send the signal rich.

Could this be testament to modern engine managements systems and the finite control of fuelling under load conditions (Diesel)? Zero mA output during period of acceleration between your gear changes

When we look at fuel consumption for diesel engines in comparisons to petrol, diesels tend to produce similar fuel consumption figures whether you “drove it like you stole it” or potter from A to B, however, the same cannot be said for petrol and reading a long note on Wikipedia :

“For any given partial load the fuel efficiency (mass burned per energy produced) of a diesel engine remains nearly constant, as opposed to petrol and turbine engines which use proportionally more fuel with partial power outputs”


http://en.wikipedia.org/wiki/Diesel_engine


Another theory regarding 0 mA output during buffers 5 and 6 could be characteristic operation of broadband sensor control (PCM) during “foot off gas” over-run fuel cut conditions

Would there be any value in pumping oxygen out of the broadband sensor measurement cell under these conditions when no fuel will ever be applied under these driving conditions?

Massive food for thought all round

Looking at buffer 6 there would appear to be activity form the broadband sensor, possibly due to “foot back on gas” or fuel re-established after the fuel cut, deceleration period?
FUEL ACTIVITY AFTER FUEL CUT.jpg
Thanks again take care…….Steve

KimAndersen
TwoWaves
TwoWaves
Posts: 246
Joined: Fri Nov 30, 2012 2:53 pm
Location: Denmark

Re: How to measure a Bosch LSU 4.9 wideband lambda sensor

Post by KimAndersen »

Hi Steve

You are absolutely right about the last bit off buffer 6. It was me putting the foot on the accelerator pedal before the car comes to a standstill.

With regards to the lambda readings I was getting have I been doing some research on the subject. By studying a document from SAE on how to control of a diesel engine with DPF (diesel particulate filter).

If I should summarize some of the claims in this document with regards to lambda values and the control of the DPF. It all comes down to controlling the inlet temperature to the DPF during the regeneration phase which is vital for correct operation.

It's no surprise to anybody that the manufacturers wants to reduce the emission from those car they are manufacturing as the emission demands are getting harder to achieve.

In order to reduce the emission from the car are the manufacturers using different algorithms to do this. One of the algorithms they are using are a way of predicting the amount accumulation (PM) in the diesel particulate filter (DPF) and one way to this, is to use the term " Lambda adaptive learning control " which is crucial for correct operation of the DPF.

One thing that caught my curiosity when I was reading this document was, how the lambda target value are determined. It is during the regeneration of DPF whereby the temperature of the DPF are monitored that the target lambda value are set.

Here is quote from this document "If the lambda value becomes too lean, the DPF temperature declines. If it becomes too rich, the DPF temperature rises excessively,causing engine-out PM emissions to increase" I do not say that every car manufactures are using the same algorithms or strategy, but why shouldn't do so, when its all about the emission the cars are emitted to atmosphere.

One last thing regarding the operation of the wideband oxygen sensor in a diesel engine versus a petrol engine. I think one should be very careful to compare those two, as they react totally different based on the the fuel, startegy/algorithms they both use.

Todays diesel engine are really complicated as there so much restriction on how much emission is allowed. Before, it was just a matter of how much diesel you could inject into the diesel engine and how to control the turbo boost that controlled the engine power output, there very were little focus on the emission back then.

Regards
Kim

KimAndersen
TwoWaves
TwoWaves
Posts: 246
Joined: Fri Nov 30, 2012 2:53 pm
Location: Denmark

Re: How to measure a Bosch LSU 4.9 wideband lambda sensor

Post by KimAndersen »

I'm back.

Just to continue my journey about to measure lambda values from a diesel engine.

This time have I used the VCDS software to measure lambda values on a dynamically road test as I did earlier in this case study. I was curious to know the correlation between real time readings and the values one gets from a EOBD software as VCDS also known as Vagcom.

One of the test I did with VCDS software was to make a log file of measuring value blocks (MVB 032) which shows the oxygen voltage signal together with the (MVB 040) which shows the calculated lambda value also known as actual value.

One thing that strikes my from the original or the latest measuring of the pump current through the oxygen wideband sensor is the lambda value during de´acceleration of the diesel engine.

Seen at this log file over lambda value during de´ acceleration is it clearly to see when I lift my foot off the
accelerator pedal position whereby the lambda value gets very high.
JETTA 2.0 TDI PD LAMBDA MVB 032 MVB 040 LOG
JETTA 2.0 TDI PD LAMBDA MVB 032 MVB 040 LOG
Of Course, it must be high as the engine is just pumping pure air through the exhaust pipe without any diesel being
added to the mixture. But a lambda reading at 25.5 is very high if you ask me, but again when i measured the pump current with my picoscope it showed a lambda reading at 1 during de´acceleration which is very low value.

I doubt, that those lambda values I'm getting from MVB 040 in VCDS software is the true values, but if its not lambda values,whats is it then !!!.

Is there anyone with some knowledge about these values in MVB 040 !

Regards
Kim

brett nz
OneWave
OneWave
Posts: 17
Joined: Sat Jun 08, 2013 10:50 pm

Re: How to measure a Bosch LSU 4.9 wideband lambda sensor

Post by brett nz »

Robski wrote:
brett nz wrote:Just a though, rather than needing a micro amp clamp why not just put a 100x coil in place and use normal clamp?
hey Brett,

how you getting on with your tiepie ?

One of the reasons I don't use that method is access to the harness/O2 connector to attach breakout leads, I might make one up for next time I have the need to.
Hi Robski, apologies for the lateness of reply only saw this when having a gander through, maybe my settings here need tweaked to let me know when posts added. Yeah tiepie going well. Just need more time on it. Only issues I have are with the loose nut behind the wheel. Would love to be using it a hell of alot more than i do. Find the 100x coil very handy, your right tho can be a bit of a pain to get all connected up with access issues. Find the coil great for picking up faults with the abs proximity sensors on nissans/subarus etc. Nice cheap handy bit of kit i reckon.

User avatar
Robski
Advanced User
Advanced User
Posts: 598
Joined: Tue Sep 14, 2010 6:36 pm
Location: Yorkshire

Re: How to measure a Bosch LSU 4.9 wideband lambda sensor

Post by Robski »

cheers Brett, nice to know to keep my next scope options open :wink:

User avatar
Fat Freddy
TwoWaves
TwoWaves
Posts: 335
Joined: Fri May 04, 2012 4:52 am

Re: How to measure a Bosch LSU 4.9 wideband lambda sensor

Post by Fat Freddy »

Good to see this thread revived. Keep meaning to add my 2 cents.
Would there be any value in pumping oxygen out of the broadband sensor measurement cell under these conditions when no fuel will ever be applied under these driving conditions?
Monitoring is one. Yes we can assume a result. But when engineers make an assumption is where a repair becomes hard and mistakes happen.
Of Course, it must be high as the engine is just pumping pure air through the exhaust pipe without any diesel being
added to the mixture. But a lambda reading at 25.5 is very high if you ask me, but again when i measured the pump current with my picoscope it showed a lambda reading at 1 during de´acceleration which is very low value.
If my brain is working this early from off spring inflicted sleep depravation. I would agree that one would expect the pumping of pure air through the engine but a Lambda of 25.5 is far too low in my opinion and theoretically should read ∞. The 4.9 is capable of reading from 0.65 to ∞. But assuming O2 at atmospheric is 21%, lambda 25.5 equals to an O2 concentration of 20.2% (if I have this correct :? ). Which for all intensive purposes is pure air.. I hope I've got that correct.

HTH
FF

Post Reply