Jump to content

Comparing and contrasting wideband O2 options on the market.


jigga009

Recommended Posts

I have decided to update a few bits of equipment on my car, and my 12 year old AEM UEGO may be the latest item to be switched out and put into retirement. It currently works okay, and has never given me any problems over the years that I have used it. Perhaps my only issue with it was the tendency to go through LSU 4.2 sensors at what I feel to be a non-OEM rate, and its inability to indicate when a sensor is getting lazy and might need to be put out to pasture. I am also concerned about the possible implications of voltage offset on the safe running of my engine, given that it works in all weather conditions, with different levels of electrical load. My aim for this thread, while to primarily discuss the Link and AEM offerings is also to hear about others that you might highly recommend that one take a look at. 

My initial thought was to go with one of AEM's X series wideband with LSU 4.9 and connect it to my G4+ xtreme via can bus. I am also looking at the Link unit, but in all my searching of the forum, have not really found a thread that really fleshed out everything there is to know about the  Can-lambda. Most of the threads I have seen on it are members advocating the use of the unit, without really saying why it is a better solution over other units out there. 

- Given that the Can-lambda retails for over twice as much as the AEM X series option (and isn't too far off from Motec LTC4.9 territory), I am curious as to what benefits, if any, there are with the Link device over something such as the AEM x series inline wideband.

- Both devices utilize canbus to transmit information. In the context of use with a Link ECU, Is there a significant difference (or something in particular that a prospective buyer might want to pay attention to) in the kind of information that both devices can transmit or receive over can when used with the Link ECU?

- I've read that the Link might have the ability to warn the user when the wideband sensor is almost worn out.. is this true? One of my pet peaves with my current 30-4100 AEM UEGO was that the sensor would get rather lazy after less than a year of use, giving me the impression that it is feeding my ECU some rather delayed information, and this is on a car that runs pump fuel exclusively.

-  I see also that the AEM has the ability to detect errors in the sensor also (encoded into the canbus messaging). Can someone knowledgable in this area confirm whether this messaging is something that the Link ECU would recognize and perhaps display on a race dash?

 I would ideally like my next wideband to alert me as to when I should be changing O2 sensors in the same manner as an OBD2 car. Not sure if this is something either the AEM, Link, or other wideband does, but it is the dream. Would also like something that doesnt go through sensors at an accelerated rate, especially on just pump fuel use.

- Is there any information on how the Link compares to the AEM from a response perspective? The AEM appears to be one of the more responsive wideband out there. Curious as to where the CanLambda comes in within the pantheon of widebands out there. 

- I have read that the Link has the ability to receive input from the Ecu so as to start operation when a certain RPM has been reached. Is this it something that could be achieved on any other wideband through the use of a relay hooked up to a GP Output on the ECU?

- While I am sure neither is 100% trouble free, I have come across a thread on the forum regarding the need for the installation of  capacitor between power and ground wiring of the Can-Lambda in order for it to operate properly. Is this still the case, or has a running change been made?

- for those with the CanLambda, what has your experience been like with your device? How often are you runnning through Bosch LSU4.9 sensors, and what is your use-case like?

 

Are there any other Canbus wideband setups that you would recommend, and why?

 

Please discuss :) 

Link to comment
Share on other sites

Adding some info I dug up on the can-lambda as well as a few questions regarding them:

 

Can lambda is apparently able to transmit to the ecu 

sensor temperature - is this logged for the purpose of adjusting the ecu map if temperature gets too high?

Lambda - no questions here..this is what we are looking to measure...

error codes  - again, no confusion here, as it tells the user if there is a problem.

- controller status - what purpose does this parameter serve that the error code above would not?

- pump current - what purpose would knowing pump current serve?

- heater average voltage - what purpose would this parameter serve?

 

- Given that there is a calibration available for the G4+ to run the AEM x-series, would this calibration allow the G4+ to receive error codes from the AEM?

- Is the Can Lambda able to free-air calibrat the wideband sensor? Asking because my current UEGO does not either, amd I read that it is needed in order to keep the sensor readings accurate as the sensor ages.

 

Link to comment
Share on other sites

I will give some comments.

The Link CAN Lambda was designed specifically to work with our ECU, it has bi-directional communication so not only is data received from the device but the ECU also sends data to it.  This allows the proper warm-up strategy to be applied during start up, EMAP compensation can be applied correctly for situations where there is back pressure and if a sensor is reporting an error condition the ecu can disable CLL or in the case of multiple sensors, ignore the failed one from "lambda average" calculations.  The stuff like temperature, heater voltage, etc are mostly only used for diagnostics. 

Most of the 3rd party devices will have no receive - they only transmit data so they dont know if the engine is running or how long it has been running for to apply the correct warm-up strategy (potentially shorter sensor life).  Many of the 3rd party devices will send some sort of error or status message but in many cases it is not in the format that our ecu needs it to be able to do stuff like disable CLL.  In some cases you will be able to log these errors but unlikely act upon them.

I have used a few of the X-series AEM's on budget installs and they have worked ok for me to date but I do hear/see some reports of short sensor life with them more often than others.  To get the extreme response speed they certainly dont use a Bosch chipset or control the sensor in the way Bosch designed it to be so sensor life is possibly a side effect that.  Some claim they are not using a genuine Bosch sensor either.

 

4 hours ago, jigga009 said:

- I have read that the Link has the ability to receive input from the Ecu so as to start operation when a certain RPM has been reached. Is this it something that could be achieved on any other wideband through the use of a relay hooked up to a GP Output on the ECU?

Yes, an ecu controlled relay is a good idea.  It still wont give the proper heat up strategy but it is better than nothing.  

 

2 hours ago, jigga009 said:

- Is the Can Lambda able to free-air calibrat the wideband sensor? Asking because my current UEGO does not either, amd I read that it is needed in order to keep the sensor readings accurate as the sensor ages.

No.  Free-air cal is not particularly useful in my opinion, it was something made famous by Innovate as they didnt use the Bosch fitted calibration resistor.  Air has about 21% O2, exhaust gas has none when less than 1 Lambda.  So it only calibrates on the lean side of the curve and even then it is like trying to calibrate your precision 1volt voltmeter using a 2000V source...

 

4 hours ago, jigga009 said:

Are there any other Canbus wideband setups that you would recommend, and why?

Ecotrons ones have a good reputation - I have never used one myself but only ever hear good comments about them.  If you are not in a big hurry I know 14point7 are releasing a CAN bus "Spartan 3" soon, I tested a prototype a while ago.  I use his analog ones for most of my budget installs and have very good success with them.

 

 

 

Link to comment
Share on other sites

Thank a for the response Adam.  

I have also seen threads on other forums where a lot of what you say comes up.

I have been curious as to AEM's dogged insistence that one use their sensors and only their sensors with their wideband. They claim that it is a Bosch LSU 4.9 sensor, but the sensor included with their unit appears to be devoid of the usual markings that would identify the sensor as such. Perhaps users if the AEM could chime in to confirm? And you are right, AEM definitely are not using a Bosch chipset in their x series wideband. Neither does Innovate. I think it was Alan From 14point7 that mentioned something to the effect of the Bosch chipset being the reason why a lot of the widebands on the market making use of that same chip are relatively "slow" to respond. With that said though, Alan did drop a lot of  knowledge, and now I know that just because a manufacturer says they use a Bosch chipset does not mean that they are controlling the sensor in the manner that Bosch advocates. There is code that has to be written to control just how exactly the chip is run to control the sensor.

This thread on this forum here proved to be a very interesting read for me: http://forum.diyefi.org/viewtopic.php?f=6&t=2267

as they discuss a lot of the widebands available on the market, and even take a few of them apart to discuss what really makes them tick.

Alan from 14point7, who makes the Spartan widebands chimes in, as do employees from AEM and Ballanger Motorsports who still produce the AFX wideband chime in at an unofficial capacity to discuss how exactly their devices work.

An individual from Ecotrons chimed in also on this thread here http://forum.diyefi.org/viewtopic.php?f=14&t=1033 , but is not met with the warmest reception, given the seemingly deceptive means through which the individual signed up onto the forum to advocate the use of their wideband under different screen names. There is some talk that the Ecotrons units are Chinese-based, but not much more info on that. Still curious as to what they are offering though, given that they are based not too far away from me.

A lot of popular widebands are discussed in the initial thread I linked to, but as has been the case in my searching so far, Link's device is not mentioned anywhere, so tough to learn more about it.

I have definitely considered 14point7 for a wideband, given that they are based down the road from me, but I still do see slight niggles regarding build quality and some failure right out of the box that cause me to hesitate. Their prices certainly are attractive, especially for their multichannel wideband unit, but I am unsure as to whether they are actually using stock Bosch sensors that I can pick up from a parts store. They mention on their website that their sensor resistors are "calibrated by a third party".... meaning what? Also, I can't seem to find much info on how frequently users might be going through sensors either.

Ecotrons appear to now how a wideband that uses the Bosch ADV sensor as the next evolution from the LSU 4.9. Do we know of what advantages, if any, this new sensor has over the 4.9?  Just as the 4.9 is supposedly more reliable (although Alan from 14point7 appears to disagree on this) than the 4.2, is thr ADV sensor supposedly even more so? 

Bosch have a Lambdatronic wideband, which one would assume to be the perfect wideband, given that they make the sensor, know how it should be warmed up, etc.  and might actually last as long as the sensor would last in a normal car, but again, very little user info on the thing out there. I can't even find a price  posted anywhere for the unit.

Does the Link CanLambda still require users to solder a capacitor between the power and ground lines of the unit in order for it to work properly? http://forums.linkecu.com/topic/6663-can-lambda-problem/

Link to comment
Share on other sites

 

I recently purchased a Ballenger AFR500 - has options for variety of sensors including the tougher NTK unit.

It also outputs a series of voltages to the ecu during warmup which gives calibration data at different points

Seems to be the wideband of choice on the Guild of EFI Tuners page before stepping up to the full on professional equipment.

Link to comment
Share on other sites

Another thing to consider if your using the Can (as i  found out the hard way )  is the data rate  for communication , its not often listed in the specs  but critical if your using more than 1 devise

AEM X series  is 500kb 

Link CAN Lambda is 1000

I have a mounted but unused AEM if you  need one ;)

Link to comment
Share on other sites

4 hours ago, Ducie54 said:

I've been using some dual Emtron CAN widebands. Can't fault them so far. 

How long have you been using it? And can you go into why you went with Emtron? What's your burn rate of sensors been like?

1 hour ago, Greg W said:

Another thing to consider if your using the Can (as i  found out the hard way )  is the data rate  for communication , its not often listed in the specs  but critical if your using more than 1 devise

AEM X series  is 500kb 

Link CAN Lambda is 1000

I have a mounted but unused AEM if you  need one ;)

Is this to mean that you cannot have as many devices using the Can when using the Link compared to the AEM? Or is it that all devices have to be transmitting at the same rate?

very interesting point to consider...Thanks!!!

4 hours ago, Rob W said:

 

I recently purchased a Ballenger AFR500 - has options for variety of sensors including the tougher NTK unit.

It also outputs a series of voltages to the ecu during warmup which gives calibration data at different points

Seems to be the wideband of choice on the Guild of EFI Tuners page before stepping up to the full on professional equipment.

The Ballenger is another tempting unit. Used to be huge with the high power Supra guys back in the day. They don't appear to have CAN communication though?

my reading also indicated that the NTk sensors, while hearty, are sluggish in the response department. Have you found this to be the case?

Link to comment
Share on other sites

10 minutes ago, jigga009 said:

Is this to mean that you cannot have as many devices using the Can when using the Link compared to the AEM?

It just means you cant mix and match. If you put on 1x device running at 500kbit, all other things on that same canbus also have to be running at 500kbit. If you look at the CAN setup screen in the link ecu's you have to set a data rate for that connection - this needs to match all the devices on the bus. If you have a model with 2x canbus's you can run them at different data rate (as most OEM's do for radios etc vs driveline system), but its kind of a waste to use up both CAN bus interfaces just to support different lambda sensors.

Link to comment
Share on other sites

6 hours ago, Greg W said:

Another thing to consider if your using the Can (as i  found out the hard way )  is the data rate  for communication , its not often listed in the specs  but critical if your using more than 1 devise

AEM X series  is 500kb 

Link CAN Lambda is 1000

I have a mounted but unused AEM if you  need one ;)

Many devices have configurable bit rate.  The Link CAN Lambda can do 100, 125, 250, 500k & 1M so it should work with nearly any other device.  The 14point7 one I believe will be configurable bit rate.  The AEM X-series is not configurable as it was originally only designed to work on their "AEMNet".  Emtron (HV Electronics) is configurable, Motec is configurable, Ecotrons is not.

 

12 hours ago, jigga009 said:

I have definitely considered 14point7 for a wideband, given that they are based down the road from me, but I still do see slight niggles regarding build quality and some failure right out of the box that cause me to hesitate. Their prices certainly are attractive, especially for their multichannel wideband unit, but I am unsure as to whether they are actually using stock Bosch sensors that I can pick up from a parts store. They mention on their website that their sensor resistors are "calibrated by a third party".... meaning what? Also, I can't seem to find much info on how frequently users might be going through sensors either.

I believe that is possibly bad wording to not give away his source. I believe they come from an OEM application, they have all the correct markings on them and none of the signs of the typical clones.  They have a longer cable on them than the common aftermarket LSU4.9.  I have installed probably 10 of his Spartan 1 & 2's over the last few years and have not had one fail or play up.

 

12 hours ago, jigga009 said:

Bosch have a Lambdatronic wideband, which one would assume to be the perfect wideband, given that they make the sensor, know how it should be warmed up, etc.  and might actually last as long as the sensor would last in a normal car, but again, very little user info on the thing out there. I can't even find a price  posted anywhere for the unit.

The Lambdatronic LT4 is their highend one, the retail is around $4000.  They have a more consumer level device, the LT2, for about $1000.  https://www.streamlineautomotive.com.au/lamdatronic-lt4-lambda-to-can-module

 

Really, it seems like you may be over thinking things.  Any of them mentioned in this post will do a fine job.  The Link one will give you a few more fail-safes in terms of CLL control if you intend to run CLL full time and possibly longer sensor life due to the more correct heating strategy.  The 14point7 one duplicates our CAN Lambda stream so that will have the same fail-safe functionality.   The AEM one is faster than all others I have used but will not disable CLL when there is a sensor error and there may be a possibility of shorter sensor life depending on who you believe.  I know a couple of professional tuners here and in Australia that use the Ecotrons ALM as their main tuning tool and rave about how good they are and I have never heard a bad thing about them so they seem good.  I know Innovate have a CAN bus one coming out soon (LC3) but I would avoid that at all costs based on experience with their previous products.  

 

 

Link to comment
Share on other sites

The Ballenger is another tempting unit. Used to be huge with the high power Supra guys back in the day. They don't appear to have CAN communication though?

my reading also indicated that the NTk sensors, while hearty, are sluggish in the response department. Have you found this to be the case?

 
 

What I have found is the ecu logs the information faster than you will see it on the guage, so if you were relying solely on the guage to see what the engine is doing you would miss some issues with fueling.

Link to comment
Share on other sites

I have installed several can lambdas and its about 50/50 whether they work or fail with heating too long, undervoltage error and require a capacitor.  There is no correlation with the vehicle wiring as we always install them as close as possible to the primary power source with 2.5mm^2 wire. 

Bizarrely, the bosch specification requires the heater voltage to be limited to 8.5 Volts on power up and ramped up at a maximum of 0.4 Volts/second so I would have thought these conditions could be easily met by any installation. 

Screenshot_20190213_193542_com.google.android_apps_docs.thumb.jpg.406b5a0353dbb60cc12fda1d5d7ead1f.jpg

 

 

 

I also have an ecotrons can lambda which I will be trying out soon, the only problem I foresee with that is the lack of encapsulation, I will have to print an enclosure. 

It does also support variable CAN speeds. 

Screenshot_20190213_192831_com.google.android_apps_docs.thumb.jpg.2c39c36f50ca0f9208cd3f62184b4ee5.jpg

 

 

The only thing lacking from both of these interfaces is a visual indication of the AFR, but I'm also working on that. 

Link to comment
Share on other sites

7 hours ago, Rob W said:

The Ballenger is another tempting unit. Used to be huge with the high power Supra guys back in the day. They don't appear to have CAN communication though?

my reading also indicated that the NTk sensors, while hearty, are sluggish in the response department. Have you found this to be the case?

 
 

What I have found is the ecu logs the information faster than you will see it on the guage, so if you were relying solely on the guage to see what the engine is doing you would miss some issues with fueling.

Thanks for the heads up. It is something that I suspect would drive one nuts, as my current OG AEM UEGO does the same thing when the sensor has some miles on it. The gauge would respond in what seemed like an eternity in response to the throttle pedal. I do currently have an AIM dash, and have the ECU sending the wideband data to it via CAN, so I'm able to skip dealing with a sluggish display. With that all said, the accuracy (or lack there of) of the AEM UEGO is a little concerning, and  part of my reasoning to update it.

Do you find that your unit suffers from voltage offset issues? From my digging, I did not think that they introduced CANBUS communication on those NGK/AFX units?

22 hours ago, Adamw said:

 

Really, it seems like you may be over thinking things.  Any of them mentioned in this post will do a fine job.  The Link one will give you a few more fail-safes in terms of CLL control if you intend to run CLL full time and possibly longer sensor life due to the more correct heating strategy.  The 14point7 one duplicates our CAN Lambda stream so that will have the same fail-safe functionality.   The AEM one is faster than all others I have used but will not disable CLL when there is a sensor error and there may be a possibility of shorter sensor life depending on who you believe.  I know a couple of professional tuners here and in Australia that use the Ecotrons ALM as their main tuning tool and rave about how good they are and I have never heard a bad thing about them so they seem good.  I know Innovate have a CAN bus one coming out soon (LC3) but I would avoid that at all costs based on experience with their previous products.  

 

 

Thanks Adam. Yes, I do like to measure thrice and cut once! Ended up ordering a Link Canlambda unit, so hopefully that should deal with voltage offset issues and possible questions of whether my wideband is actually doing what it is supposed to be doing.

Given the data parameters that the Link wideband is able to send to the ECU, which parameter most directly indicates when the O2 sensor is on its last legs?

For all of the discussion regarding warm up strategies and the like with the Link, are you able to comment as to just how closely the CanLambda follows Bosch's recommendation for warm up and use of the LSU 4.9?

Have you found Link's warm up strategy to yield noticeable differences in how long the CanLambda can maintain an LSU 4.9 sensor compared to other brands out there, or even older widebands such as the OG AEM UEGO (which also used the Bosch chipset, but with an unknown warm up strategy)?

Link to comment
Share on other sites

6 hours ago, jigga009 said:

Given the data parameters that the Link wideband is able to send to the ECU, which parameter most directly indicates when the O2 sensor is on its last legs?

I dont think there is anyway to know be honest.  The reported Lambda will go to zero and the "Lambda 1 error" parameter will display an error code if anything such as pump current etc goes out of tolerance but that is usually the point they are considered "failed". 

 

6 hours ago, jigga009 said:

For all of the discussion regarding warm up strategies and the like with the Link, are you able to comment as to just how closely the CanLambda follows Bosch's recommendation for warm up and use of the LSU 4.9?

If it is connected to a Link ECU it adheres to all of Bosch's rules.  

 

6 hours ago, jigga009 said:

Have you found Link's warm up strategy to yield noticeable differences in how long the CanLambda can maintain an LSU 4.9 sensor compared to other brands out there, or even older widebands such as the OG AEM UEGO (which also used the Bosch chipset, but with an unknown warm up strategy)?

I dont have a lot to offer here as I have never really had an issue with sensor life with any controller I have had.  Except for a couple of innovate devices I had about 20years ago - they would sometimes report a sensor was dead in a matter of hours...

But I also dont use one for a daily driver so my use is possibly more gentle then someone who lives in a cold climate and does cold starts a couple of times a day...

On my dyno I have a single ETAS controller (Bosch's research division) for the dyno DAQ and if Im tuning a Link ECU I will have a Link CAN Lamda connected direct to the ECU so that I can use quick tune etc.  The ETAS can use any sensor type but I have had a LSU4.9 on it for about the last 3 years.  As it doesnt have any awareness of engine RPM and is powered by the dyno system, it obviously doesnt follow the correct heating strategy.  Quite often when Im tuning that thing will be powered up for the whole day with the sensor just burning away regardless of if the engine is running or not - and I dont think I have ever needed to replace the sensor (leaded fuel is banned in motorsport here now so that helps).  Neither with my Link CAN Lambda which would have probably 2 years of regular tuning work on it (I think that sensor was originally a 2nd hand one also).  I always have a new spare sensor in my draw to try if Im ever suspicious but havent killed one or noticed drift or slow down for a long time.

Link to comment
Share on other sites

We see lots of wideband controllers here, the worst for sensor failure are AEM, Innovative seem to be better, there is a noticeable difference as you really don't see sensor failing that often on the LINK CAN Lambda, we put the sensor failures down 100% to the heating strategy not shocking the sensor and allowing time for residual water / condensation to be cleared.

The other thing is that the AEM sensor supplied with the kit does not appear to be bosch, but although you could argue that is the reason they fail, replacement genuine bosch dont seem to last that long either. That is the main reason we try and get people to buy the Link sensor and we always ask about the placement of their Lambda bung.

PLUS you get EGT with the Link CAN Lambda an invaluable device that has saved my customers $1000's in turbo rebuilds.

Link to comment
Share on other sites

7 hours ago, ClintBHP said:

We see lots of wideband controllers here, the worst for sensor failure are AEM, Innovative seem to be better, there is a noticeable difference as you really don't see sensor failing that often on the LINK CAN Lambda, we put the sensor failures down 100% to the heating strategy not shocking the sensor and allowing time for residual water / condensation to be cleared.

The other thing is that the AEM sensor supplied with the kit does not appear to be bosch, but although you could argue that is the reason they fail, replacement genuine bosch dont seem to last that long either. That is the main reason we try and get people to buy the Link sensor and we always ask about the placement of their Lambda bung.

PLUS you get EGT with the Link CAN Lambda an invaluable device that has saved my customers $1000's in turbo rebuilds.

Interesting tip about getting EGT with the Can Lambda. Had no idea about that feature. I had previously run EGT sensors on the car, but stopped running them a few years ago when a probe snapped off and took out a turbo.

 

Edit - looking at the manual for the Can-Lambda, I'm can't find reference to the ability to measure EGT. Are you referring to probe temperature, as a proxy for exhaust gas temperature?

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...