PLX SM-AFR Wideband Install (Help!)

[35R]

Could someone help me with this wideband installation?

I purchased the new SM-AFR PLX wideband sensor and associated module.

Reading from the "Connecting Power to the Unit" section; I dont quite understand the bits relating to power setup, maybe I should just go to a sparky?

I dont want the sensor and module permanently installed however;

i would rather something more portable, if you know what i mean.

The user manual can be found online here:
http://www.plxdevices.com/InstallationInstructions/SM-AFRUsersGuide.pdf

Any tips or pointers would help me a lot, thanks

[Scotty]

Dunc,

I think you would find any auto elect would find that easy.

As far as being able to plug it in as you wish, I would recommend just a couple of good plugs from repco etc which are wired from the +/- terminals to your desired location, and then do likewise with your sensor gizmo.

I hope this helps. I will forward the pdf to a mate to comment on.

[Sonny@AutoWerks]

Could someone help me with this wideband installation?

I purchased the new SM-AFR PLX wideband sensor and associated module.

Reading from the "Connecting Power to the Unit" section; I dont quite understand the bits relating to power setup, maybe I should just go to a sparky?

I dont want the sensor and module permanently installed however;

i would rather something more portable, if you know what i mean.

The user manual can be found online here:
http://www.plxdevices.com/InstallationInstructions/SM-AFRUsersGuide.pdf

Any tips or pointers would help me a lot, thanks


Not sure wat type of unit you have duncan,though if ur in the area pop in and i can hook it up for you

regards sonny

[35R]

Not sure wat type of unit you have duncan,though if ur in the area pop in and i can hook it up for you

regards sonny

thanks Sonny, might take you up on that if I can get over your way sometime.

[35R]

Like most things it turns out simple, just getting the time to work it out is the issue. I had a local electrician knock up a portable power supply using the cigarette lighter socket thinging, and then simply wiring the analogue and ground wires from the WB to the HPTuners dongle. Simple I guess, but an extra four wires all over the place, arghh! Didnt have time to do any logging, so perhaps I can log on the way to work.

[35R]

I did a little WB02 logging on the way to work the other day

These numbers are based on Narrowband tuning using fuel trims:

http://www.trackchat.com.au/media/data/500/medium/afr_nb.jpg

Some "off" numbers above of course (mainly only a few cell hits for those) and some of the loopy ones are gear changes I noticed, but numbers arent as far out as I imagined.

Anyway, now time to tune using the wideband. So reading up on the various WB02 tuning tutorials, it looks like I should follow these steps (keeping in mind car is already tuned mafless)

Change all points to 1.13 in the Open Loop F/A vs ECT vs MAP table (commands AFR of 13.0)
Change all points in the Closed Loop Enable Coolant Temp vs IAT table to 250* (Disables closed loop)
Change all points in the Power Enrich Fuel Multiplier vs RPM table to 1.0 (disables PE mode)
Use wideband and HPT histogram to verify AFR of 13.0
Make adjustments to the VE table accordingly to dial in a 13.0 AFR using the desired formula - current afr/13.0 = VE multiplier
Hand smooth VE if desired, repeat VE tuning until all cells in the histogram are 12.8 - 13.2.
Change all points in Open Loop F/A vs ECT vs MAP table back to stock (re-enable stoich commanded AFR)
Change all points in the Closed Loop Enable Coolant Temp vs IAT table back to stock (re-enables closed loop operation and fuel trim leaning)
Change all points in the Power Enrich Fuel Multiplier vs. RPM table back to stock (re-enables PE mode)
Use wideband and PE table to dial in desired WOT AFR. (Optimum HP at WOT)

[SSUte01]

I did a little WB02 logging on the way to work the other day

These numbers are based on Narrowband tuning using fuel trims:

Some "off" numbers above of course (mainly only a few cell hits for those) and some of the loopy ones are gear changes I noticed, but numbers arent as far out as I imagined.

Anyway, now time to tune using the wideband. So reading up on the various WB02 tuning tutorials, it looks like I should follow these steps (keeping in mind car is already tuned mafless)

Change all points to 1.13 in the Open Loop F/A vs ECT vs MAP table (commands AFR of 13.0)
Change all points in the Closed Loop Enable Coolant Temp vs IAT table to 250* (Disables closed loop)
Change all points in the Power Enrich Fuel Multiplier vs RPM table to 1.0 (disables PE mode)
Use wideband and HPT histogram to verify AFR of 13.0
Make adjustments to the VE table accordingly to dial in a 13.0 AFR using the desired formula - current afr/13.0 = VE multiplier
Hand smooth VE if desired, repeat VE tuning until all cells in the histogram are 12.8 - 13.2.
Change all points in Open Loop F/A vs ECT vs MAP table back to stock (re-enable stoich commanded AFR)
Change all points in the Closed Loop Enable Coolant Temp vs IAT table back to stock (re-enables closed loop operation and fuel trim leaning)
Change all points in the Power Enrich Fuel Multiplier vs. RPM table back to stock (re-enables PE mode)
Use wideband and PE table to dial in desired WOT AFR. (Optimum HP at WOT)


Dunc,

Whilst it wont hurt to do all the things listed above I personally feel that they are outdated and this process is not optimised for what you are capable of doing. Perhaps it is just my WB but I find that perhaps setting fuelling accurate to 13.0 at low MAP and RPM is fine but as soon as you revert back to Closed Loop and/or Lean Cruise etc ie 14.6x or leaner there will now be error and you will have to adjust again anyways

My recommendation is (remember its just what I have found, take it for what you will) since you are using the very very capable 2.0 histograms simply use the pre-configured AFR % histogram which is formatted exactly tha same as your VE table for LS1s. Ensure you are logging Commanded AFR hi res and of course your WB or it wont work. Leave PE enabled, leave Lean Cruise std (although IIRC its disabled on Monaros from the factory) ie leave all your commanded AFRs as you would normally - if you wish to make all OLFA 1 for simplicity you can but it isnt necessay, if you do be sure to return to what you had when finished or you will have cold start issues.

Do disable Closed Loop either by maxxing temp in Editor or simply replace an O2 with your WB, as soon as an O2 is unplugged it is forced into Open Loop plus you get the bonus of logging pre-cat (might be a bit fiddly with the Munro cats though....dunno) and do disable DFCO for the tuning part as this will throw out your decel AFR error.

The reason why those previous steps where taken so that the commanded AFR is known always (13.0 to be safe at all rpm) you would then log actual AFR and use a speadsheet to calculate error. Now the histo does it for you regardless of commanded AFR, the histo is simply the error of commanded AFR whether it be 12 or 20 and actual AFR plotted conveniently in VE format. Point is it is sophoisticated enough not to require a constant commanded AFR

Once populated copy and paste special to your VE table (Paste special => Add) and hand smooth as required.

Summary

Disable DFCO
Disable Closed Loop
OLFA to 1 (optional)

*If you haven't got the 1Bar SD enh OS you will be operating on the low octane table once the MAF fails so you can copy this to the high or better yet put the advanced OS in for safety.

Now you are in open loop, log WB pre-cat if possible - not only is it more accuarate there is far less delay ie it takes time for actual AFR to reach sensor. Most prominent during transient conditions. Drive around start in 6th (M6 obviously) and slowly increase throttle from as low as possible about 1100rpm up to about 2000rpm (due speed of course), do the same in 5th but from 1500 to 2500, again in 4th from 2000 - 3500 etc right to cover all loads and rpms. This method does not take long to populate the table and it is perfectly within the speed limit. If you have a dyno then by all means use it but most of us dont and nothing simulates the real world like the real world - drag strip is a good place for WOT if no dyno. If auto use the bi-directional controls to lock it in 4th, 3rd, 2nd etc to do the same.

Now the table is populated, make sure 3-5 minimum in each cell to eliminate spurious data, and copy and paste the raw error the first time. Subsequent times, it usually takes 3 or 4 times to get it real close, only copy over half the error and sometimes on the fourth I only copy over a quarter or selectly use a combination if I see an odd value (I use Excel for this as the editor and Excel are compatible) This prevents overshoot or you will find you end up chasing your tail.

TIP: Filter you histograms using and IAT filter or to the scanner written like: [SENS.11]<use apt value here eg 40. This is so that temp lean condition caused by heatsoak (tfc lights, stop to reflash etc) does throw your fuelling. I have found that for example when cruising IAT is 30 a quick stop will see it shoot up quite rapidly and cause fuelling to be lean. If I filtered only below 40 in this example dialling the VE was much more consistent. Setting the bias to 1 eliminate alot of this and it never goes dangerously lean just enough to make dialling in the VE difficult.

Once you are satisfied you are where you want to be 1-2% throughout and less than that up top ie within 1% you can start working on other areas. I will then re enable DFCO and Closed Loop and do a final test including WOT to triple check all is in order.

All things equal fuelling (VE or MAF) is sorted (note you will need to create a custom histo for the MAF Err %, let me know if you need one it only takes a few seconds to set up).

BTW those values above, what are they, it almost looks like either measured AFR from the WB(albeit a consistent log) or commanded AFR. I am assuming it cannot be fuel trims or they would all be the same number in Open Loop eg PE. Double take I re-read what you wrote and I am assuming that is measured AFR based on your tuning using the NBs, is that the case?

I am sure someone has something critical to say, but this worked like a treat for me on various cars in various weather

Hope this helps

Cheers,
Luke

[Sonny@AutoWerks]

Pretty much on the money there luke,good write up .

:)

regards sonny

[35R]

Dunc,

Whilst it wont hurt to do all the things listed above I personally feel that they are outdated and this process is not optimised for what you are capable of doing. Perhaps it is just my WB but I find that perhaps setting fuelling accurate to 13.0 at low MAP and RPM is fine but as soon as you revert back to Closed Loop and/or Lean Cruise etc ie 14.6x or leaner there will now be error and you will have to adjust again anyways

My recommendation is (remember its just what I have found, take it for what you will) since you are using the very very capable 2.0 histograms simply use the pre-configured AFR % histogram which is formatted exactly tha same as your VE table for LS1s. Ensure you are logging Commanded AFR hi res and of course your WB or it wont work. Leave PE enabled, leave Lean Cruise std (although IIRC its disabled on Monaros from the factory) ie leave all your commanded AFRs as you would normally - if you wish to make all OLFA 1 for simplicity you can but it isnt necessay, if you do be sure to return to what you had when finished or you will have cold start issues.

Do disable Closed Loop either by maxxing temp in Editor or simply replace an O2 with your WB, as soon as an O2 is unplugged it is forced into Open Loop plus you get the bonus of logging pre-cat (might be a bit fiddly with the Munro cats though....dunno) and do disable DFCO for the tuning part as this will throw out your decel AFR error.

The reason why those previous steps where taken so that the commanded AFR is known always (13.0 to be safe at all rpm) you would then log actual AFR and use a speadsheet to calculate error. Now the histo does it for you regardless of commanded AFR, the histo is simply the error of commanded AFR whether it be 12 or 20 and actual AFR plotted conveniently in VE format. Point is it is sophoisticated enough not to require a constant commanded AFR

Once populated copy and paste special to your VE table (Paste special => Add) and hand smooth as required.

Summary

Disable DFCO
Disable Closed Loop
OLFA to 1 (optional)

*If you haven't got the 1Bar SD enh OS you will be operating on the low octane table once the MAF fails so you can copy this to the high or better yet put the advanced OS in for safety.

Now you are in open loop, log WB pre-cat if possible - not only is it more accuarate there is far less delay ie it takes time for actual AFR to reach sensor. Most prominent during transient conditions. Drive around start in 6th (M6 obviously) and slowly increase throttle from as low as possible about 1100rpm up to about 2000rpm (due speed of course), do the same in 5th but from 1500 to 2500, again in 4th from 2000 - 3500 etc right to cover all loads and rpms. This method does not take long to populate the table and it is perfectly within the speed limit. If you have a dyno then by all means use it but most of us dont and nothing simulates the real world like the real world - drag strip is a good place for WOT if no dyno. If auto use the bi-directional controls to lock it in 4th, 3rd, 2nd etc to do the same.

Now the table is populated, make sure 3-5 minimum in each cell to eliminate spurious data, and copy and paste the raw error the first time. Subsequent times, it usually takes 3 or 4 times to get it real close, only copy over half the error and sometimes on the fourth I only copy over a quarter or selectly use a combination if I see an odd value (I use Excel for this as the editor and Excel are compatible) This prevents overshoot or you will find you end up chasing your tail.

TIP: Filter you histograms using and IAT filter or to the scanner written like: [SENS.11]<use apt value here eg 40. This is so that temp lean condition caused by heatsoak (tfc lights, stop to reflash etc) does throw your fuelling. I have found that for example when cruising IAT is 30 a quick stop will see it shoot up quite rapidly and cause fuelling to be lean. If I filtered only below 40 in this example dialling the VE was much more consistent. Setting the bias to 1 eliminate alot of this and it never goes dangerously lean just enough to make dialling in the VE difficult.

Once you are satisfied you are where you want to be 1-2% throughout and less than that up top ie within 1% you can start working on other areas. I will then re enable DFCO and Closed Loop and do a final test including WOT to triple check all is in order.

All things equal fuelling (VE or MAF) is sorted (note you will need to create a custom histo for the MAF Err %, let me know if you need one it only takes a few seconds to set up).

BTW those values above, what are they, it almost looks like either measured AFR from the WB(albeit a consistent log) or commanded AFR. I am assuming it cannot be fuel trims or they would all be the same number in Open Loop eg PE. Double take I re-read what you wrote and I am assuming that is measured AFR based on your tuning using the NBs, is that the case?

I am sure someone has something critical to say, but this worked like a treat for me on various cars in various weather

Hope this helps

Cheers,
Luke

Luke,

thanks for that rather grand contribution, i will try and digest it over the next few days; MUCH appreciated.

[35R]

Luke,

thanks for taking the time to explain all above. I didnt realise it was that simple now (essentially disabling closed loop, and logging AFR Error). I had some dramas getting the new PLX SM-AFR to work, but once it was, it was relatively easy.

I was going through my logs from Wakefield today, and noticed the timing down vs where I thought it should be. I eventually tracked it down to the IAT park Advance correction table. It looks like i still have the factory numbers in there (they seem rather aggressive with 9* taken out at the pointy end).

What would you say the accepted practice is for this table, how pessimistic is it? I think it is probably robbing me of some power, perhaps I shoudl try cutting it in half and try that. Just looking at some other aftermarket tunes, they seem to have cleared the table completely.

[SSUte01]

Luke,

thanks for taking the time to explain all above. I didnt realise it was that simple now (essentially disabling closed loop, and logging AFR Error). I had some dramas getting the new PLX SM-AFR to work, but once it was, it was relatively easy.

I was going through my logs from Wakefield today, and noticed the timing down vs where I thought it should be. I eventually tracked it down to the IAT park Advance correction table. It looks like i still have the factory numbers in there (they seem rather aggressive with 9* taken out at the pointy end).

What would you say the accepted practice is for this table, how pessimistic is it? I think it is probably robbing me of some power, perhaps I shoudl try cutting it in half and try that. Just looking at some other aftermarket tunes, they seem to have cleared the table completely.

Yep, it essence matching commanded AFR to actual AFR is the key to fuelling. Matching the optimum fuelling to the optimum spark is the fun bit after all the idle and other nuances are sorted.

IAT Spark Correction table is by far a necessary protection feature even more so on FI setups but still important on NA setups. I do not recommend zeroing this table, it just isn't smart.

Basically it allows you run optimal timing in any temperature - in theory at least.

Did you happen to log Main Spark (absolute timing after all adders/multipliers) usually the very first histogram that will tell the real story as to what timing you are running and if it isn't what you expect there could be some not yet found adder playing a part.

Timing is very unique to each setup but there are accepted ranges. To answer your question I looked at an old tune of yours and I assume it has changed since then however it was a little too aggressive. I view Cyl Airmass (g/cyl) as, whilst not strictly correct, load. So since this IAT spark adder is ECT vs g/cyl you want your most protection at higher loads and higher temperatures.

To begin halving is not a bad idea, but without seeing the table I would be guessing. I personally have IAT correction zeroed under 30* IAT due to the fact I see 30* IAT when ambient is more like 20* and 50+ when ambient is 35*. From there I subtract a small amount of timing at higher loads in 30* and increase with IAT and g/cyl. I do not see above 0.90g/cyl normally but have the correction from 0.9 to bottom the same anyway.

I have attached a correction table that works well on mine. Big thing to remember is it will depend on your main spark table in those areas how much protection is required. Here's mine. Apologies for the cut down version, but I couldn't get the x axis and keep it under 97kb

Cheers

[35R]

Thanks Luke. Yeah i logged main spark which alerted me to the fact (maxed out at 22* it seems) which was low versus my spark table. I am only guessing really but I think it is the IAT adjustment tables. I too was going over an old tune today (stock mafless) and I noticed the IAT tables essentially zero'd out (or close to it). I started from scratch again (with the head and cam) so I guess thats why the factory numbers are back there. I'll do some experimenting over the next few days and report back. Thanks for the advice.

[SSUte01]

Thanks Luke. Yeah i logged main spark which alerted me to the fact (maxed out at 22* it seems) which was low versus my spark table. I am only guessing really but I think it is the IAT adjustment tables. I too was going over an old tune today (stock mafless) and I noticed the IAT tables essentially zero'd out (or close to it). I started from scratch again (with the head and cam) so I guess thats why the factory numbers are back there. I'll do some experimenting over the next few days and report back. Thanks for the advice.

IAT correction may be playing a part, that's it's job when setup. I would be willing to bet that if there is a consistent amount taken of WOT spark say 4-6*, that it is Burst Knock - you can log this as a separate PID to find out if it is a contributor. I must admit I do not fully understand the rationale of Burst Knock, but as I understand it, it isn't true knock but rather used as a preventative measure to real knock created by transient conditions ie stomping on the throttle. I cannot speak for others but I personally zero this table out as it is one may adder that is not required. It causes timing to be pulled when it just shouldn't like low load and rpm when an engine mechanically sound/ fuel OK just wont knock. True KR will still function however it has been set up. Some LS1 Knock sensors are nototriously under sensitive but thats getting off track.

Burst Knock lives under Engine => Spark Control => Spark Retard => Burst Knock Retard. Dont worry about the enable delta cyl air, but zero the 'vs RPM' table - I believe this will make quite a difference. Just make sure your Main tables Hi or Lo which ever you run off are still safe.

I have never had an issue with a car that has the spark setup correctly and safely and with Burst KR zeroed out.

I would like to hear others' thoughts

Cheers