Wideband O2 related subpage of MembersPage/VasilisP
New ECU (v3/10581 AAN with NTC MAT input, oid 10875); config file uploaded as requested in the order comment (without any change: eg. not changed in any way; even the LSU4.9 was not changed, the paper provided with the order - or any LSU4.9 sensor sent out - instructs to check that 4.9 is configured)
Please note that [ECU with motronic55] can be ordered with "loom for LSU4.2" which is compatible with [LSU4.2] and very old fw. (but newer fw-s are better, and LSU4.9 also better, so the choice of most people). If you purchase LSU4.2 sensor, and write "please include small loom for AAN SSC5-LSU4.2" in the order comment, we'll include the loom for no extra charge: that allows to use new fw or reproduce the old setup (1.2.11 fw, config) with old fw which is not LSU4.9 certified.
Upload ini file - note: VT ini-s must be uptodate
When using new fw (eg. 1.2.31), VT must have all the new ini-s for it (preferrably the latest).
- Use the newest VT (nightly, or [dev])
- AND "Preferences / fetch latest ini files from web"
- that might be the step why uploading old config to new fw shows so good fuelpw match on bench (for many-many testcases), and it ends up slightly rich when uploading without up-to-date ini-s (which can still be tuned of course, but more work).
- of course we would need to compare the actual vemscfg (or vemslog) saved from the device (should be provided with any report, but it wasn't), and the old config properly uploaded to new fw and saved'.
- can you detail the "updated the config" ? Or perhaps provide the "updated" .vemslog or .vemscfg for review) ?
- When i say UPDATED i mean the VE tables which were last known to work. the HTA 12A config i asked you to use in order for the car to run propeply i expected you would have had the WBO2 4.9 checked, as the ECU comes with that WB sensor and not the old 4.2 which was in the config. I noticed that myself.
The 'update' was on VE and lamda tables that i had reviewed myself after recent checks.
Cool, thanks.
Before the engine starts with the key turned to the pre-ignition state, my rev meter jumps to above 2000rpm and my dash warning lights up the LOW OIL LEVEL. Both stay up unless i start the car.
- so it seems the low oil-level is unlit when engine is started and oil-pressure is present (good)
- the RPM tacho also shows correct value when engine is started, if I understand. Showing 0-7000 RPM is well within the normal function of the dash (be assured, dash won't be harmed electronically from 8900 RPM or higher, or from any 0-12V signal coming in the dash-tach input from ECU),
- and lighting up low oil pressure is very good indication the dash works as intended, and actually doing what it was designed for; with no risk of damage (these dashes worn mostly because of solder-corrosion and cold-warm heat-cycling, not because of any voltage input well within the normal voltage range).
I have only connected the WBO2 sensor and nothing more.
- Is this some kind of ground current reflected in the gauges?
- I need this fixed and corrected as it may cause longterm problems with my dash on parts that are no longer available
- There should be no high-voltage or anything similar that could damage the dash.
Injectors=... cc/min, type=... Fuel pressure= ... bar above MAP pressure
I noticed that all my map was too rich (ECU-s nowadays come with 1.2.31 fw, but in oid10875 order comment ancient 1.2.11 config was specified, and uploaded accordingly).
- Yes, the semantics in fw changed slightly a few years ago and not exactly same VE (or reqfuel) values are needed for same injector-pulsewidth (multiple benchtests showed within 1.5%, definitely much lower than 15%). Usually no difference when upgrading, but when a very old (eg. 1.2.11) config / tables are uploaded to new firmware (eg. 1.2.23 or later) => it might be somewhat rich (changes are on the safe side at least, but tuning is good... It is actually tunable as the main feature, so that it can always be tuned to the actual setup including injectors, fuel pressure regulator, head, turbo, firmware-generation).
- if you post an 1.2.11 vemslog and 1.2.31 vemslog on same engine (under similar conditions, eg MAT, CLT), we'll verify and investigate. If possible, point us to ...sec and ...sec with comparable input-vector but more than 1.5% difference in injpw/(MAP*VE*req_fuel * enrichments).
- EGO was reading around 85 on all map regions
- ... aha, so removing appr 15% fuel(pw). (also same effect if eg fuel pressure is 4 bar instead of 3 bar, or if injector is different. I assume same types, just the fw differed - actually very same engine if I understand... But than, as the SSC5 WBO2-connection is backwards-compatible, your existing old LSU4.2 sensor can be used with the old SSC5-LSU4.2 loom, with old wbo2 sensor calibration value, and even old 1.2.11 fw... with same behavior that you liked that every possible way that matters - obviously ECU serialnr differ and new ECU has a bit more protection against installation abuse, eg. on analog inputs )
- If multiplier is that consistent (including idle?), might be easier to compensate with req_fuel (*=0.85). Just in case, check warmup and afterstart enrichment (is warmup enrichment 100% at CLT=72C ? Maybe it's ~115% causing the excess fuel / richness ? The view / calc model has a fuel pw related selection (try right-click) where the calculation can be followed, all factors of it... very useful.
- When i calibrated the WBO2 for the first time the 4.2 was selected, I calibrated the lamda sensor and used 214 number to achieve a 20.9 air calibration.
After changing the wide band setting to match the LSU4.9 sensor and tried to recalibrate, it went to 255 for me to have 20.1 at air.
Ahh OK, so actually very close. As a car owner, keep the 255 config value. (and if you really care, compensate for the over-sensitive LSU4.9 sensor with lambdatarget=0.89 where 0.88 intended; 1.00 where 1.00)
- PLEASE, Can you share your measurement of the sensor Rcal = ... Ohm ? with DVM.
- With that simple info, for no charge to you, we were considering to send an LSU4.9 sensor that is verified not at the extremely-sensitive end of the Bosch gaussian distribution (and we'd verify a stock of 300+ similar sensors without heating, to see how many are that extremely sensitive, if any).
- i am a car owner, and im afraid i have no means to measure what you ask. I can only tune the ECU.
- OK. Although an Ohm measurement is the simplest thing that comes up during install (verifying injectors, igncoils, CLT/MAT sensors, WBO2 sensors, wires, all routine part of install+tune - even with a factory ECU if it comes to real-life diagnostic ). Anyway, glad to here you can tune the ECU, because that solves it all: you should be able to tune it as you wish.
- do you have another sensor (LSU4.9 or LSU4.2)?
- I have a 4.2 lamdda sensor lightly used. If i am to recalibrate should i clean it?
- is it possible the sensor was dropped (eg. to concrete floor) ?
- SENSOR WAS NOT DROPPED!!!!
- note that if WBO2 calibration is lower than optimal (eg. not calibrated, or maxed out), the WBO2 reading is still REPRODUCIBLE.
- It just means that lower deviation from 1.0 lamda is displayed (lower than the real 20.9% O2 value is displayed; and higher than real lambda value displayed:
| lambdatarget (=lambda reading) | desired lambda (real lambda) |
| LR | LR, or LR-0.01 below 0.9 |
| 0.79 | 0.78 |
| 0.80 | 0.79 |
| 0.81 | 0.80 |
| 0.82 | 0.81 |
| 0.83 | 0.82 |
| 0.84 | 0.83 |
| 0.85 | 0.84 |
| 0.86 | 0.85 |
| 0.87 | 0.86 |
| 0.88 | 0.87 |
| 0.89 | 0.88 |
| 0.90 | 0.90 |
| 0.91 | 0.91 |
| 0.92 | 0.92 |
| 0.93 | 0.93 |
| 0.94 | 0.94 |
| 0.95 | 0.95 |
| 0.96 | 0.96 |
| 0.97 | 0.97 |
| 0.98 | 0.98 |
| 0.99 | 0.99 |
| 1.00 | 1.00 |
| 1.01 | 1.01 |
(some multiply with zzz constant and say "AFR" that is even more unnatural).
Unless any followup info, use that -0.01 under 0.9 lambda formula to compensate the miscalibration (255 instead of 258) for the sensor apparently living on the extreme (RCal) end of the Bosch-LSU4 calibration Gaussian-distribution (being very sensitive, interestingly, so lower Ip current was needed at 20.9% O2.
Any air pressure, elevation, humidity data perhaps ? Would only be useful with Rcal Ohm data though)
- it does not even prevent closed loop feedback EGO/lambda mixture control. VT can be configured to compensate, or another sensor can be used of course.
I am now using my old VEMS and i can send this unit back for all necessary checks.
It can be tuned where the engine is.
Maybe send the sensor if anything (but not before measuring Rcal).
I also tried 1.2.30 version without any success.
That is no surprise, as 1.2.30 works exactly as configured just like 1.2.31 (it's just that 1.2.31 has some new convenience and trigger features).
The key to success is measurements (even if just a few and simple in this case, they are important).
We suspect
- the sensor might be faulty (we do NOT heat the sensor, the sensor is tested by the manufacturer - Bosch in this case).
- or perhaps the SSC5-LSU4.9 small loom might have 2 wires swapped, or perhaps all mirrored ? (we test the v3, and the manufacturer of the loom tests the loom, we just supply that from the shelf).
- we test the v3 carefully before sendout, but of course nernst or pump HW problem cannot be completely ruled out (or misconfiguration, which we should verify also from vemscfg / vemslog).
Besides the measurement, write the SSC5 pin number and the wire color for each pin (and the wire-color at the LSU4.9 sensor !), and verify connectivity and pin-correctness of the small SSC5-LSU4.9 loom according to the [LSU4.9 pdf].
- DC voltage between SSC5/pin4 pump- and GND ? (4V or so ?)
- SSC5/ ... ? loomwirecolor/sensorwirecolor ?
- DC voltage between SSC5/pin1 nernst and GND, if 1k (or other, 1..100k documented value) resistor is connected between nernst and GND ? Near 1/(1+27)*5V = 0.179V ?
- SSC5/ ... ? loomwirecolor/sensorwirecolor ?
- DC voltage between SSC5/pin2 heater+ and GND ? (12V or around?)
- SSC5/ ... ? loomwirecolor/sensorwirecolor ?
- DC voltage between SSC5/pin3 heater- and SSC5/pin2 heater+ ? Can you see the sensor internally glow to cherry-red and get warm, and slightly smell while the voltage climbs up above 6..9V when WBO2 calibration is started ?
- SSC5/ ... ? loomwirecolor/sensorwirecolor ?
- optional measurement if NOT already diagnosed by the above. DC voltage between SSC5/pin5 pump+ and SSC5/pin4 pump-, with 100 Ohm between pump+ and nernst; and 100 Ohm between nernst and pump-?
- SSC5/ ... ? loomwirecolor/sensorwirecolor ?
Refused to measure, but most likely all should be fine... It seems it's just the actual LSU4.9 sensor is at the most sensitive extreme end of the Bosch Rcal calibration Gauss-distribution (wbo2 calibration ~258..259 not 255; Rcal maybe ~226 Ohm ?).
- Marcel, as i mentioned to the email. It is most likely that VEMS has sent the wrong sensor to me when ordered the 4.9. I read the order there is no attachment of .2 or .9 on the invoice and i dont understand why.This may also be partially my fault as i mentioned in the order details that the ECU was meant for 4.9 but as i see the vems motroinc ordering page, it says with LSU4 but doesnt specify if it is now the 4.2 or 4.9. Instead of ordering JUST LOOM and separately the 4.9 which i should have done, this ECU probably came with the 4.2 as standerd. I think this needs to be more clear as the ordering page and you should include the option to buy WITH LSU 4.2 or 4.9. Because on the motronic page on vems shop, it says: with LSU4 highly recommended. This is a link and if you press it, it takes you to the 4.9 lsu shop page. That is why I got misled that if I add this on the ecu purchase I will automatically be sent the ecu with a 4.9 sensor and loom This has caused the lamda problems i have obviously
- In addition, as i connect my previous ECU i have no dash short circuits, and no oil warning coming up, even though as you say it may not have any implications, i do not want to take the chance and ruin an original dash of an audi RS2 which is obsolete and if broken probably costs more than VEMS ECU to replace.
- Therefore i would like to send both the lamda probe and ECU back to you for inspection and checks of the grounds and i would like this ECU to be tested on a vehicle before you send it back. If everything is ok, please send the correct probe (4.9) and connector cables to the ECU. IF you cannot find or cannot fix the internal problem that causes my ground issues, then id like a refund. I am happy to cover shipping costs back and forth. As the purchase of the lamda could be partially down to my misled ordering, i am willing to buy the 4.9 IF the ECU hardware can be fixed to not short circuit my dash
- thanks
Hello VasilisP,
To determine what is going on please provide a bit more information, please answer the following questions for me:
- are you using a new (or old please specify) lsu 4.2 or lsu 4.9 probe during the free-air calibration and is it in free-air not in exhaust ?
- The first 4.9 lsu calibration (with the wrong selection on vemstune, 4.2 instead of 4.9) was done with a sensor completely new. The second calibration where the 255 was reached without being able to calibrate was done after the same 4.9 sensor was used with the car idling and myself looking at the trigger with a strobe light. So basically very little use.
- is your wiring from ssc5 to wideband connector verified and known good ?
- My wiring is as sent from VEMS shop:
| SSC5 | WBO2 | color | wirecolor inside sensor |
| 1 | 1 | black | ... |
| 2 | 3 | thick red | ... |
| 3 | 4 | brown | ... |
| 4 | 5 | blue | ... |
| 5 | 6 | thin red | ... |
Checked and verified
And found wrong ? It looks wrong to me !
It seems to be made to LSU4.2 pinout (it's a miracle that it works at all), not LSU4.9 ... See [aan_doc.pdf]
Are you sure ? Is it the new LSU4.9 cable with small WBO2 connector ?
- unfortunately Bosch changed the pin numbering => do NOT use "WBO2" when talking about the connector, use "LSU4.9" or "LSU4.2"
- Can you photo so that wirecolors / orientation and size is seen ? (without misunderstanding about LSU4.9 connector pinout)
It appears then that i was sent the wrong parts.
- when trying to calibrate your sensor, does it get warm ?
- It does
- could you make a vemslog of wideband calibration warmup cycle (let it run for 4-5min) and add the vemslog to your page here for review.
Using the provided information i will provide hints on how to proceed.
Best regards, Dave
- Apparently and after an email from Marcel, it seems that VEMS shop has sent me the wrong lamda probe. when i read about the connector being different, i thought the size was slightly smaller, but when included in the package i get a connector to go with everything i thought i was sent the correct parts.
See FileArea (how to upload files). Generic instructions on MembersPage
I dont have vemslog with the heating cycle but i can make one. Thx.