o2bad455

OldSkullTuning looks like a promising source as a time-saver, especially if they can provide both a comprehensive 450 cdi engine XDF and a comprehensive 450 transmission XDF. But personally, I don't think anyone should charge for XDF definition files since that's an ongoing group effort for open-source tuning software (e.g., TunerPro), and XDFs are technically ineligible for copyright protection due to being a mere compilation of facts. I see that I actually have several Smart engine XDF definition files in one of my old Smart folders from 2020, but none were accurate or complete enough to do what I'd wanted in TunerPro for the Bosch EDG15C ECU (0.8L cdi engine), and I'd never even found one for the 6-speed 450 transmission (still considering a manual swap anyway - I saw a 5-speed Toyota trans swap into a 450, but it used the Toy engine too and the fit wasn't pretty). From memory, I think I got one of the XDF files from the DIY community on the Nefmoto site, another from a batch of dozens that I'd stumbled upon online, and the third one I'd converted from a working OLS def file using an alpha release conversion tool that I'd found on either NefMoto and/or GitHub, but the conversion was either incomplete or imperfect (don't know if it can be touched up or fixed). I don't remember much else, other than that I was contemplating a full version of WinOLS to simply use the unconverted OLS def file instead of converting to XDF for TunerPro, but my desire to stay open-source kept me from trying that. See attached (OLS zipped to meet upload file size limit). Having some experience with TunerPro, my next step was going to be to learn enough about WinOLS to do a manual conversion from OLS to XDF myself, but I never found enough documentation to get the job done. Smart_ForTwo_cdi_nm_csc_351431.xdf 0261205005.xdf Smart_CDI_0281010161_351431.xdf Smart CDI 0281010161 351431.7z

Let us know if you find a proper XDF for the 0.8L cdi. It seems like I spent months (on and off) looking for an XDF, then trying to convert from OLS to XDF, but ultimately set it aside a while back. It sure would have been nice to hit the ground running with TunerPro and NefMoto or the like after some previous success with a different car.

Perhaps it is a North American thing related to N.A. OBDII. Since VCDS for VW was mentioned in the O.P., I'll admit I once used an official Rosstech VCDS VW/Audi interface (that I had to use one day for no other reason than that it was all I had on hand for generic OBDII testing/logging) that definitely jumped the Smart cluster odo, at least at first. But I recall that Uwe (the head of RossTech) advised me to disable "boot in intelligent mode". That change at least slowed the Smart odo jumping, but it was several years ago and I don't recall if it completely stopped it. I haven't used that particular interface on a Smart again since then.

I found this handy old ChipTuning shareware utility (the rar contains an exe for ECU v0.12), and it still works on Win 10! If my understanding is correct, this little utility may more accurately tell the software version that a modified map is actually based on, which can apparently be different from that "advertised" or reported by kwp2000. All stock maps should be and report the same number, so this utility and kwp2000 both reporting the same version is at least a good sign. Name: Bosch ecu nr.rar Size: 55660 bytes (54 KiB) SHA256: F0E49E6C7623098913234040BD9D8E7DA474C1CE26AC0942742E8A551B79923B Cheers! Bosch ecu nr.rar

Good question and I have no answer. But if similar odo corruption (cluster and SAM) was happening on both US gas 450 and Canadian diesel 450 due to the same OBD cable incompatibilities, and US dealers could read the ECU and correct the SAM and cluster based on the ECU's uncorrupted reading, then I don't see why it wouldn't also be possible on the Canadian cars even if the Canadian dealers are, shall we say, either less knowledgeable or less indulging. In any case, I'll probably try to do a full backup (not just the map) of my ECU sometime soon, and seek help wherever I can find it to read the true odo reading from that (assuming it hasn't also been corrupted by the SAM and/or cluster). I'll let you know if any success or definitive failure, either way.

Correction regarding the OBDII Connector: CAN-bus uses pins 6 and 14 as standard, while pin 8 was used at manufacturer discretion (see below). Some cars (at least some BMW) used pin 8 as a second K-line (in addition to pin 7), but Smart used pin 8 as a +12V switched source! Some cables (e.g., some generic VAG-COM cables I've seen) used pin 8 as an extra carrier of the pin 7 K-line signal (possibly for redundancy, better signal propagation, noise immunity, or really anyone's guess why). Here's the standardized OBDII connector pin-out: Pin Signal Description 1 Manufacturer Discretion 2 J1850 Bus+ SAE J1850 PWM and VPW 3 Manufacturer Discretion 4 CGND GND 5 SGND GND 6 CAN High CAN ISO 15765-4 and SAE J2284 7 ISO 9141-2 K-LINE K-Line of ISO 9141-2 and ISO 14230-4 8 Manufacturer Discretion 9 Manufacturer Discretion 10 J1850 Bus- SAE J1850 PWM only 11 Manufacturer Discretion 12 Manufacturer Discretion 13 Manufacturer Discretion 14 CAN Low CAN ISO 15765-4 and SAE J2284 15 ISO 9141-2 L-LINE K-Line of ISO 9141-2 and ISO 14230-4 16 +12V or +24V Battery power But here's the Smart 450 pin-out (per Evilution re 700cc petrol, hopefully also applicable to 800cc diesel): Pin 1 SAM unit connection 9, pin 3. Diagnostics Pin 2 No connection Pin 3 ECU pin 102. TNA Pin 4 Negative earth Pin 5 Negative earth Pin 6 No connection Pin 7 ECU pin 110. K-line, Diagnostics Pin 8 Switched live, fuse 20 Pin 9 ESP unit pin 11. Diagnostics Pin 10 No connection Pin 11 Steering assist unit, connection B, pin 2. Diagnostics Pin 12 Restraint systems/Airbag controller pin 9. Diagnostics Pin 13 No connection Pin 14 No connection Pin 15 No connection Pin 16 Permanent live, fuse 11 It also seems to me that pin 1 might be causing an issue since many diagnostic cables use it to detect switched power but Smart has it connected to the SAM (where Smart's pin 8 is switched power). It's possible that the odo jumping issue is actually happening at the SAM due to detection attempts on pin 1, and then being updated to the instrument cluster since those two (SAM and cluster) supposedly update each other to the greatest reading of either (while the ECU supposedly remains neutral while maintaining it's own private copy of odometer count), at least from what I've read so far.

Our Smart's finally back after isolating from Covid for a couple of years. Does anyone have the little shareware file called bosch_ecu_number.rar (presumably something similar once decompressed)? Once I've backed up our map, I'd like to confirm that both the kwp2000+ and the bosch_ecu_number utility report the same map number, but all the old links to that little utility seem to be broken. Thanks in advance!

Ah, that makes sense, but I think there's a bit more to it. I can confirm that when I used my VCDS Hex-Com+CAN cable it definitely jumped our Smart 450 cdi odo. So let's attribute that to CAN. But I also tried two "dumb" VAG-COM cables (no longer remember which was which), one of which jumped it and one of which didn't (or at least I didn't notice if it did). I believe I read that some of these dumb cables have pin 7 internally jumpered to pin 8 while others leave pin 8 open (something about being dual-K capable or not). I think only the ones with pin 8 open are safe for Smart 450 cdi. I also remember using a BMW cable with a physical switch on it (which had to be one way for a late 2004 BMW and the other way for a late 2008 BMW) that definitely jumped the Smart odo in one switch position but didn't seem to jump it in the other switch position (no longer remember which switch position). I just looked up that switchable BMW cable and saw that the switch seems to connect either pin 7 in one position or pin 8 in the other position, but not both. So presumably it was jumping the odo only when pin 8 was connected to something, but not jumping the odo when pin 8 was disconnected/open. In summary, CAN or not, I believe if pin 8 of the OBDII connector is connected to anything other than entirely open/disconnected, that there's a risk of it jumping the Smart odo. That might not be the only reason, but it definitely seems to be a factor. EDIT: Also, Smarty450's excellent KWP2000+ Guide (attached) indicates that some but not all KWP2000+ cables have pin 7 internally jumpered to pin 8, so pin 8 must be separated (e.g., clipped internally) for use on Smart. KWP2000 Smart CDI Guide.pdf

I agree! KWP2000+ interface with pin 8 (+12V on Smart) disconnected and almost any tuning software. For tuning software, I prefer the NefMoto suite since open-source and there's an active community, but it's proven hard to get proper definition files for the Smarts. I have OLS files presumably including Smart definition files for WinOLS, but would need full WinOLS to extract those for use with NefMoto's TunerPro, which definitions could then be further developed for better use in remapping using the open-source NefMoto suite. I've tried to extract XDF from OLS using the trial version of WinOLS, but that didn't work. Alternatively, I think I'd be all set if I could only figure out how to get the KP from the OLS, since there's then a mapdump program by nyet to convert that KP into an XDF definition file suitable for TunerPro... Sorry I'm so late to this thread, but you or others with like questions can search "KWP2000" in this forum if you hadn't earlier or are still wondering. There was plenty of initial interest on this forum in self-tuning the Smart cdi, and some are certainly still working on it. But unfortunately, there's also been a bit of disinformation about originator rights which has stunted sharing and open development at times. Apparently like you, I've both benefited from and contributed to ongoing projects with other makes, but this was our first foray in Smarts. IMHO, although the ECU programs that read our cdi data maps are likely copyrighted in some jurisdictions (and portions possibly patented in others), the maps themselves are just data (analogous to subject matter in some court cases where certain phone-books and other data listings were ruled ineligible for copyright). ECU maps are merely data, useful in the right context but non-artistic and non-functional on their own. I don't speak for anyone else here, but I think most reasonable minds have reached the understanding that a particular map's originator (if known and undisputed) should be recognized. I and some others seem to agree that, at least as a professional courtesy, data maps should not be identically (e.g., no detectable difference - not just one checksum which sometimes has to be maintained for other reasons) redistributed without permission if the originator is still doing business by selling those particular maps. But, there's no perpetual right if they're not; and even if they still are, incremental changes can sufficiently differentiate the new maps in these contexts. For those maps that aren't published with originator info or where the originator is unclear (an inherent issue since map data is basically a result of optimizing a complex system for different constraints among different users), and/or for which it's been discovered that such maps were actually the work of someone other than the purported originator from the get-go, it's generally considered hypocritical to claim others can't learn from the existing maps and adapt them for ourselves or even others' usage with few changes, or at least arguably in these cases as-is. We're not merely "standing on the shoulders of giants", but supporting each other as contributing members of society, even when just incrementally. That's what humans are good at (civilization)!

The 450 with Toyota swap seemed to have been running from many clues, including custom clutch pedal having been painted and then apparently worn off from wear. The stock rear plastic fenders had been widened by an inch or so per side, but I didn't have a tape measure with me. It seemed to have the stock DeDion rear axle so I assumed stock rear track at the time, although it's possible that axle was adapted from a 451 since I didn't measure and didn't even contemplate the possibility until you asked. It also had customized upper eye-bolt shock mounts, although not pretty. Definitely Smart three-lug drums. The car did have 451 seats and other random niceties in it, so the builder probably had access to a 451 parts car. My greatest concern was that the angle of the stock-looking Toyota oil pan looked too steep (I thought I took a pic of that but actually can't find it - perhaps my cellphone cam failed to click - I do have about 30 other pics of it). Not familiar with Toyota engines so not sure if oil starvation might have resulted from the poorly oriented oil pan (and presumably stock pickup screen). But who knows, maybe that's why it ended up at the boneyard! Definitely solvable, though, and I wonder if even peppier Toyota engines might mate up to that compact Toyota trans...

I found a 2006 Smart 450 Coupe with Toyota 1.5L 16V (109 HP stock) and 5-speed MT swap in the Laval, QC Kenny U-Pull boneyard (row 118A). It looks like it was basically a running mock-up, and I think the engine tilt angle could be improved a bit, but I've decided against grabbing it myself. It's already been stripped of many stock parts, but the custom clutch pedal, manual transmission cable shifter, modified Smart engine cradle with complete Toyota VVT-i engine, ECU, wiring, manual trans, and custom half-shafts, as well as a Toyota instrument cluster (damaged?) are still there. For those who don't know this yard, they won't let you buy the whole car (or all that's left of it) under any circumstances (half a car at most, IIRC), but the whole swap including engine cradle, clutch pedal and shifter could probably be had quite easily. It's been there since May 10 so may get recycled soon if someone doesn't rescue it. I'd definitely do it if we had room for a second Smart (ideally a Coupe that I'd cage), but I've decided to try my hand at tuning our stock Cabrio (with KWP2000+ and probably NefMoto) and finishing some other car projects instead. EDIT: I did take some 2.16 MB pics, but can't upload due to 2 MB per file size limitation.

Not sure, but I think it's determined entirely by the pinout. I've lost track of which scanners did what, and stopped caring after our 40K kms purchase jumped significantly after some scans plus real-time OBD port logging while I was trying to sort out an over-boost issue. But for those of us who've unintentionally jumped our odometers by scanning through the OBD port, I've read that US MB dealers can read the true mileage from the ECU even after the instrument cluster's been corrupted. Does anyone know if/how we can do that ourselves? Here's what I'd read: "There is some north american compliance issue with the ODBII port that can cause the odometer to arbitrarily roll forward when a cheapo ODBII scanner is used on the car. One such cheapo scanner is the one used for the Missouri emissions test. When I took the car in for a scan, it had 67,000 miles. When I left, it had 80,000 miles. There's not much I can do about this problem other than document it and warn you against using ODBII scanners. The correct mileage is stored on the car's computer as verified by a Mercedes dealer scan. It's just the display that's off."

Stock is generally quoted as 41 bhp, which I believe is at the flywheel before other losses (some fixed, some not). Most seem to agree that the power can be about doubled (e.g., Brabus) without seriously affecting engine longevity, but that low-rpm torque should be artificially limited. Ours is fast enough for the flats unless there's a headwind, but slows substantially up hills. I'm sure any reliable improvement would help! Having had some Bosch ME7 ECU OBD2-port tuning success (luck?) a few years back with a previous ride using NefMoto and TunerPro, I've been itching to try my hand on the Smart's Bosch EDG15C ECU. The maps are one thing, but finding and/or developing the definition files will be key to any repeatable success (i.e., not just one size fits all or copy/flash/copy/flash). Definition files are the comprehensive files that define at least what the data maps mean, and ideally also what the program variables are, even better the program steps themselves! Unfortunately, they're rather difficult to devine from scratch, and there hasn't been a lot of open community development for this particular ECU platform.

Which engine/ECU? I noticed that the software I received with my ebay cable included a folder with 44 different DLL files including one specifically for the 0.8L cdi (identical to the one Smarty450 already provided) as well as one for the 0.7L petrol. Attached is the DLL for the 0.7L petrol engine. If it's a different petrol engine, it might possibly be covered by one of the other ECU units in that DLL folder (although if we're to believe the file dates, none are newer than 2005). Kwpu003.dll

My KWP2000+ kit arrived from ebay and I was pleased to find it identical to the one in the guide except that pin 8 was not connected. Unable to install in Windows 10 Home 64-bit due to lack of any 64-bit driver. Using your 32-bit drivers, I was able to install in 32-bit Windows virtual machines including Win 7 Pro and Win 10 home running under a Win 10 64-bit host OS. . Our Smart's already in winter storage so not sure when I'll actually read it, but does anyone know how to convert an ".ols" flash file into an ".obj" file? Attached is the file in question. Smart CDI 0281010161 351431.7z