R1150GSA Pinging

Discussion in 'GS Boxers' started by GS-Jake, Dec 18, 2006.

  1. cycleman2

    cycleman2 Been here awhile

    Joined:
    Jan 18, 2010
    Oddometer:
    608
    Location:
    Alberta, Canada
    It doesn't surprise me that the PC III connects something to the 02 Sensor. Once the engine is at operating temp, that is the sensor where the PCM gets its main signal with lesser inputs from the throttle position sensor as to amount of fuel to add.

    If you are going to take the 02 Sensors out of the mix then you'll have to provide an artificial voltage reading to the PCM when it is in closed loop mode, which is just about everything up to 3/4 throttle.
  2. fred flintstone

    fred flintstone Been here awhile

    Joined:
    Apr 12, 2009
    Oddometer:
    827
    Location:
    Above 7600 ft.
    Well it surprised the Dynojet people because for years they just recommended disconnecting the O2 sensor on older PCIII USB models to run properly "open loop". Low and behold after getting the wideband PCIII for the GS I was reading up trying to figure out what the harness connection into the OEM O2 socket was doing. I kept seeing the O2 simulators now being recommended for older PCIII to eliminate left over adaptation due to deafault mode for no O2 sensor detected.

    To be clear the old O2 sensor is thrown away on the wideband. The new sensor plugs into its own harness that goes directly into the PCIII. There is another branch off the PCIII harness that plugs into the empty OEM O2 socket off the main OEM wiring harness to the motronic. What exactly that does is the question. The O2 simulators I mention above for older non wideband PCIII's also plug into the empty O2 sensor socket in the main harness. It looks mainly like a resistor. It does not plug into the PCIII harness, so maybe something more than just a resistor is going on with the connection to the PCIII wideband on the GS; ie it is being used as an input to the piggyback or to the motronic somehow.
  3. Mr. Canoehead

    Mr. Canoehead Taste Gunnels!

    Joined:
    Apr 16, 2007
    Oddometer:
    2,198
    Location:
    Calgary, Too far to the mountains and too cold
    It would make sense for the PCIII to feed a constant voltage to the O2 sensor port on the Ecu so that the Ecu doesn't try to adapt or throw a code. Then the PCIII could control the mixture because the ECU would be blind to the changes being made. The Ecu would think everything was fine and stick to the stock map, which is consistent with my experience of it not fighting the PCIII.

    Originally the PCIII kit came from one of the BMW dealers (Santa Fe BMW?) and you just disconnected the O2 sensor and it always ran open loop. One of those kits came up on FM and I called the dealer and spoke to the guy who developed it. He said that they stopped producing it when DynoJet came out with a closed loop kit.

    Sent from my SGH-I747M using Tapatalk 2
  4. roger 04 rt

    roger 04 rt Long timer

    Joined:
    Nov 9, 2011
    Oddometer:
    2,715
    Location:
    Massachusetts
    That is the question. Powercommander makes a product called Wideband Commander which i would guess uses the same technology--but it's only a guess. (Dobeck, maker of Techlusion, has a Wideband option on its latest product, not available for our bikes. They announced a deal with Innovate: Dobeck and Innovate.)

    Since the PC III has a cable to the Motronic O2 input, and since the SW shows a closed loop area, I'm now starting to reconsider the way it functions with Motronic. It may be that, like the innovate LC-1, it lets the Motronic run a Closed Loop program part of the time. That part time is likely to be in the shaded area below. That being the case, the Motronic would then have what it needed to calculate Adaptation Values in the areas where the PC allowed Closed Loop.

    If this all sounds a little confusing it is. On the one hand, the PC can only add Open Loop fuel when the Motronic isn't Closed Loop. On the other hand, as much as I've measured the Motronic for my LC-1 install, I can't say precisely how the Motronic's Adaptation Values would interact with the PC. Given the Wideband Commander technology and the difficulty of designing a Closed Loop software program it makes some sense that the PC would let the Motronic perform CL. I'll look into this some more.

    In the meantime here is an article that compares Innovate and Wideband Commander: WIDEBAND SHOOTOUT
  5. roger 04 rt

    roger 04 rt Long timer

    Joined:
    Nov 9, 2011
    Oddometer:
    2,715
    Location:
    Massachusetts
    Guys, Just sent a detailed email to Dynojet, waiting to hear back. Will post results. Can any of you tell me how many wires are in the cable going from the PC III to the Motronic O2 sensor input. It is most likely one wire but may be two. RB
  6. fred flintstone

    fred flintstone Been here awhile

    Joined:
    Apr 12, 2009
    Oddometer:
    827
    Location:
    Above 7600 ft.
    Hi roger,
    I think you'd have to pull the tank off to do that, and it is below zero deg F here in my non heated garage plus I just filled up the tank after confirming everything worked on this mod. Then this winter storm moved in. I'll at least remove the seat later today and look at it.

    AJ
  7. fred flintstone

    fred flintstone Been here awhile

    Joined:
    Apr 12, 2009
    Oddometer:
    827
    Location:
    Above 7600 ft.
    Just to get even more confusing, the unit that I have and I think Canoehead (Ian) has are not Wideband Commanders as shown in the above article, they are called "PCIII USB including Wideband Sensor". It is the exact model 919-611 that is pictured in the PDF sheet here:

    http://www.powercommander.com/downloads/919/install/611/eng919-611.pdf
  8. roger 04 rt

    roger 04 rt Long timer

    Joined:
    Nov 9, 2011
    Oddometer:
    2,715
    Location:
    Massachusetts
    Yup. Understood that. Waiting on some feed back from PC for 919-611 (R1150GS) and 918-611 (R1150RT). It's likely the technology and circuitry for the Wideband Commander and the Powercommander III USB are similar if not the same. But will know more in a day or two.

    Can you take a look and see how many wires are in the cable that goes to the Motronic O2 input? Wondering if it is 1 or 2.

    One of the things I discovered during my LC-1 project is that the Motronic input is two wire, offset by 140 mV but that it can be driven by a single wire if the voltages are offset.
  9. Mr. Canoehead

    Mr. Canoehead Taste Gunnels!

    Joined:
    Apr 16, 2007
    Oddometer:
    2,198
    Location:
    Calgary, Too far to the mountains and too cold
    It is cold and snowy here as well and the mighty GSA is tucked away, waiting for some Chinook winds but IIRC, there are two wires. The reason I remember it this way is that the plug from the motronic has 4 pins and I puzzled over why there are only two coming from the PCIII. Then I realized that two of the pins are for the O2 heater circuit which the PCIII doesn't use. I actually thought about rigging up a circuit to power the sensor heater.

    Fred is right, I have the same model as he does.

    Sent from my GT-P7510 using Tapatalk 2
  10. roger 04 rt

    roger 04 rt Long timer

    Joined:
    Nov 9, 2011
    Oddometer:
    2,715
    Location:
    Massachusetts
    No need to power the heater circuit on an 1150. I've tried it with and without a load resistor to simulate the heater. There is no error code if you leave the heater off--and nothing getting really hot! (On some motorcycles and cars you do have to provide a load to the O2 heater circuit, but not on the 1150.)
  11. Mr. Canoehead

    Mr. Canoehead Taste Gunnels!

    Joined:
    Apr 16, 2007
    Oddometer:
    2,198
    Location:
    Calgary, Too far to the mountains and too cold
    I was thinking more about getting the sensor online faster, particularly in cold weather.

    Sent from my SGH-I747M using Tapatalk 2
  12. roger 04 rt

    roger 04 rt Long timer

    Joined:
    Nov 9, 2011
    Oddometer:
    2,715
    Location:
    Massachusetts
    The Wideband sensor that comes with my LC-1 and your PC III USB includes a very fast heater circuit, less than a minute. The cable to Bosch LSU 4.2 Sensor that they all seem to use is 5-wire harness that includes a heater circuit run by the LC/PC.

    Because the sensor is so fast, I've notice that the Closed Loop swings are about half as small as the narrowband, meaning you get tighter afr control. As far as I see $170 for an LC-1 is about the best performance money you can spend. And even though the PC III IS more, it's money we'll spent too--I'll post a table soon that would preclude dyno tuning for most. I just need some details from Dynojet.

    RB
  13. Mr. Canoehead

    Mr. Canoehead Taste Gunnels!

    Joined:
    Apr 16, 2007
    Oddometer:
    2,198
    Location:
    Calgary, Too far to the mountains and too cold
    The PCIII doesn't heat the sensor at all. Those pins are missing on both sides of the PCIII. It would have been easy to pass the circuit from the ECU to the new O2 sensor but they didn't bother. :confused:

    I have an LC-1 sitting on the shelf that I bought for use on my Lightning engined Expedition. I bought an extra bung and plug for my Mazdaspeed but I never thought to try it on the bike!

    Al and I are kindred spirits when it comes to bike and car mods.

    Sent from my SGH-I747M using Tapatalk 2
  14. roger 04 rt

    roger 04 rt Long timer

    Joined:
    Nov 9, 2011
    Oddometer:
    2,715
    Location:
    Massachusetts
    That doesn't make sense to me because a Bosch LSU 4.2 won't work if not heated. The PC III doesn't use the Motronic heater circuit, it has one from the PC III.

    You should weld a second bung on the 1150 exhaust. Then you could log data and really see what's happening. You really have a lot of toys.
  15. fred flintstone

    fred flintstone Been here awhile

    Joined:
    Apr 12, 2009
    Oddometer:
    827
    Location:
    Above 7600 ft.
    If only we could use our powers for good instead of evil.

    Speaking of which all this great discussion of wideband sensors talking to motronics has me wondering the following:

    If the steptoe jumper mod for catless really is for no cat, then all the voltage oscillation to keep a cat happy should be missing, yes? And then I was wondering what purpose adaptation of any kind would have on an OEM catless map and if there were one would it differ from a catted adaptation. Maybe just for e10 vs no e. It sure would be nice to have a way to read the full motrocinc map(s) as opposed to guessing.

    Very curious to hear what the dynojet folks come back with on all this. I'd be very surprised to hear all the motronic adaptation is somehow left on. Pretty sure that all the closed loop happens inside the PCIII, if not getting an adaptive loop inside the motronic to work under those cicumstances would be challenging. And getting different fuel tables working in the PCIII with closed loop + adaptation happening inside the motronic would be even more difficult.

    Since this is in the GS section I wonder if any other 1150 engine BMW folks are running the PCIII wideband and have info. Searching the web I did not find much talk. I think now that used prices of these PCIII's have gotten to the point that they are cheaper than LC1 or comparable to booster plugs and similar, maybe some more folks should look into them. It was literally plug and play for me, 1 hr job tops including tank & exhaust removal/reinstall for an amazing improvement with substantial open ended future tuning capability.

    Ian I can't remember if you are running a Y-pipe or not.
  16. roger 04 rt

    roger 04 rt Long timer

    Joined:
    Nov 9, 2011
    Oddometer:
    2,715
    Location:
    Massachusetts
    Hi FF,
    You raise very good questions. The maps inside the Motronic are visible if you have an O2 data logger like the LC-1. I've run various coding plugs and ridden my bike in closed and open loop. I don't remember if I tried the 87a-only map. I can't find any signs that there is a NO-CAT map for the 1150 and I've been looking for one. All the maps I've tried seem to be for 14.7:1 with Closed Loop enabled. Later today I will specifically test 87a-only to see if the Motronic goes into Closed Loop and will report back.

    For the PC to Motronic connection there are two possibilities. 1) They send a simulated narrowband signal to the Motronic. 2) They don't send that signal. For condition 1) the Motronic will Adapt and there is no mechanism for the PC to stop Adaptation. For condition 2) the Motronic will default to limp home mode and send a 10% fueling variation pulse stream. I'm start to believe that case 1) is more likely but I am waiting to hear from Dynojet. I agree that case 1) seems like it would present a challenging Adaptation scenario. Lets wait and see.

    Going back to the 87a-plug, Steptoe never described what it did, only suggested that it was easy, cheap and worked to improve some 1150s. I haven't read anywhere that it was a no cat map. It may be that it affects spark advance more than VE/fueling. I highly doubt it's an E10/no-E map since Adaptation handles the ethanol issue.

    Good discussion, thanks.
  17. fred flintstone

    fred flintstone Been here awhile

    Joined:
    Apr 12, 2009
    Oddometer:
    827
    Location:
    Above 7600 ft.
    Hi Roger,
    Quick response, in the middle of something. See posts #28 & 29 in this thread for the 30-87a jumper description for non-cat. It is labelled as R1100 in #29 diagram but later on Steptoe and maybe Poolside elaborated that it also applied to no cat on 1150 models and in a more noticeable way.

    More later.
  18. fred flintstone

    fred flintstone Been here awhile

    Joined:
    Apr 12, 2009
    Oddometer:
    827
    Location:
    Above 7600 ft.
    Do you mean you can read the entire 3-d map inside the motronic? I don't think so. I think you may be able to log some I/O parameters and O2 sensor voltage but getting the map itself off the unit needs some expensive hardware to read. Just to get terminology and function straight, here is a nice write up:

    http://www.bikeboy.org/fuelinjection.html

    By map switching via jumper mods, I assume it means there are different 3-d tables of spark vs rpm vs TPS already loaded into the motronic, that are jumper select-able (8 of them) . That is the map(s) I'd like to read. Then there are the trim tables. The trim tables are where adaptation takes place [the base maps are hard coded (to my understanding)] and where the closed loop operation of the motronic is carried out.

    I THINK that the PCIII with a plug into the stock O2 harness is designed to disable most or all of the adaptation going on in the trim tables, AND all of the closed loop. BTW If what is said in the link is true then the PCIII closed loop operation area is much larger than the OEM. I have not played with the PCIII enough to see if you can make closed loop area larger or not.

    We have not said this explicitly but clearly the OEM O2 sensor is also used to move parameters around in the trim tables so that open loop operation is affected (for e10 for example). I might also add here that the reason one wants/needs a different base 3-d map is that the adjustments in the trim tables are limited to a max % up or down, that is they use O2 sensor output to tweak the setting up to a fixed amount above or below the 3-d map. So if a 3-d map is way off actual VE characteristics, due to mods (like no cat or cams pistons etc.) it could not move to optimal an won't be able to hit say 13.8 or whatever the target was. This is likely a very slow process relatively to closed loop corrections.

    What would be interesting is if the PCIII still uses the richer AFR target of 13.8 to adjust the open loop trim (i.e. still running some open loop adaptation), or even some subset of it, yet runs its own (maybe larger than OEM) closed loop area native on the PCIII.

    Another way of saying this is maybe there is a fast closed loop and a very slow closed loop native operation of the motronic. The fast closed loop is real time depending on dynamic conditions (analogous to the gray area on the PCIII map) and the slow is gradual updating of the "adaptation" of trim tables for the "open loop" operation of the controller.

    Surprising how little there is out there about how this all works, either the motronic or the PCIII for that matter.
  19. fred flintstone

    fred flintstone Been here awhile

    Joined:
    Apr 12, 2009
    Oddometer:
    827
    Location:
    Above 7600 ft.
    One more point and I will shut up (and get back to work!). The PCIII works by essentially adding a completely separate set of trim tables, at the injectors. It must read TPS and rpm directly, and then with some internal knowledge of the motronic map ads or subtracts % to the injector pulse width. This is why I think it needs to disable most/all adaptation, the two tables would be cross talking. Once adaptation is disabled, the fixed PCIII trim table would act as a updated 3-d base map, and then closed loop operation could work on that. It makes no sense to try to implement closed loop operation otherwise but what do I know. BTW I suspect this is what happened on the older PCIII's, and why they came out with the O2 simulator plugs. They either (wrongly) assumed open loop adaptation was disabled when it wasn't or it was so slow they did not see it happening on their product development dyno sessions.

    Those O2 plugs arrived for my other bikes, they are just a plug with two resistors across the four O2 sensor leads. If we get any kind of thaw going I am eager to install them and see what happens.
  20. roger 04 rt

    roger 04 rt Long timer

    Joined:
    Nov 9, 2011
    Oddometer:
    2,715
    Location:
    Massachusetts
    On my lunch break I built a Coding Plug jumper to try out the 87a-30 mod. The motorcycle was cold and I connected to the LC-1 datalogger port for AFR and attached the GS-911. I ran from cold to fully warmed up and logged all GS-911 and LC-1 data. Here are the results:

    1) At the four minute mark, the Motronic went into stable Closed Loop operation in the same way it does with the Pink Plug (30-87-87a.

    2) The shape of the cold-start through warm-up enrichment phase was identical to the Pink plug. However, the 30-87a jumper was 10% LEANER than the Pink plug. Surprise!

    3) I did not evaluate spark timing between the two yet.

    Although I am running these tests on an '04RT, I think that it is fair to say that the 30-87a map is leaner that 30-87-87a on any '04 R1150.

    When I get a half hour, I'd like to make some observations of your (Fred F) prior two posts which raise some interesting topics.

    RB