BMW M5 Forum and M6 Forums banner

481 - 500 of 501 Posts

·
Registered
Joined
·
387 Posts
Yeah, it's definitely not worth my time. I've looked into it...haha...
 

·
Registered
Joined
·
375 Posts
The E60 chassis is just too well integrated with it myriad of sensors and computers to play well with a standalone ecu for a trans.
I don't know about that. The general strategy would be to set the ECU to manual mode (which has no transmission electronics whatsoever), and then let the standalone handle the shifting. As far the rest of the chassis would be concerned, you're just amazing at shifting. It wouldn't be quite as well integrated as a proper DCT setup, but it should still be workable.
 

·
Registered
Joined
·
368 Posts
Workable and being able to use is a manual only mode is very different in my book than fully integrated where it'll shift itself needing real time info from the DME to work well. That's where you run into integration issues. I can't imagine anyone who wants this kind of swap for a street driven purpose and who will actually put their wallet down want's a hacked together job where they lose features, have lights on in the dash, and codes permanently thrown.

It's not like you can just change a few settings in winkfp to nicht aktive or plug and play E9X parts. I really don't see how to get a DCT fully working playing well with everything in the chassis wouldn't require reverse engineering/rewriting custom software for the DME and who knows what other computers. If I'd hazard a guess, that's where the multiple businesses who tried to tackle this before gave up.

Again, not saying this all can't be done. With enough money, time and stubbornness anything is possible. But you need to have an expert understanding of how to rewrite the car's software and the time to devote to what would be a passion project. The time, money invested, and ROI clearly don't make sense for the aftermarket to tackle, and that's saying something given the issues with the SMGIII and cost to diagnose/keep going. Otherwise we'd have a solution by now, and not still dreaming about it on the internet.

Put it in the same category as the fabled S85 twin turbo kit. Although I think that's actually more easily realized.
 

·
Registered
Joined
·
375 Posts
Workable and being able to use is a manual only mode is very different in my book than fully integrated where it'll shift itself needing real time info from the DME to work well. That's where you run into integration issues. I can't imagine anyone who wants this kind of swap for a street driven purpose and who will actually put their wallet down want's a hacked together job where they lose features, have lights on in the dash, and codes permanently thrown.
I mean that's precisely what many are doing when swapping DCTs into earlier generation BMWs (and non-BMWs for that matter). It does work. Won't be OEM level integration, but probably still better than the pos that is the SMG. You wouldn't have any permanent lights or codes or whatever, because as far as the car is concerned, it's strapped to a manual transmission.

It's not like you can just change a few settings in winkfp to nicht aktive or plug and play E9X parts. I really don't see how to get a DCT fully working playing well with everything in the chassis wouldn't require reverse engineering/rewriting custom software for the DME and who knows what other computers. If I'd hazard a guess, that's where the multiple businesses who tried to tackle this before gave up.

Again, not saying this all can't be done. With enough money, time and stubbornness anything is possible. But you need to have an expert understanding of how to rewrite the car's software and the time to devote to what would be a passion project. The time, money invested, and ROI clearly don't make sense for the aftermarket to tackle, and that's saying something given the issues with the SMGIII and cost to diagnose/keep going. Otherwise we'd have a solution by now, and not still dreaming about it on the internet.

Put it in the same category as the fabled S85 twin turbo kit. Although I think that's actually more easily realized.
Having a fully integrated setup is also feasible, just time consuming. Essentially you'd be using an MSS60 and editing it to run in V10 mode and altering all the maps etc to match the E60 M5. It's fairly difficult, there's probably only a couple members on this forum at best who are capable of it, but by no means is it impossible. There isn't really a business case for it, so it'd have to be a hobbyist venture.
 

·
Registered
Joined
·
368 Posts
Having a fully integrated setup is also feasible, just time consuming. Essentially you'd be using an MSS60 and editing it to run in V10 mode and altering all the maps etc to match the E60 M5. It's fairly difficult, there's probably only a couple members on this forum at best who are capable of it, but by no means is it impossible. There isn't really a business case for it, so it'd have to be a hobbyist venture.
Looks like you just reworded the second half of my last post more or less?

Not sure where we disagree except for the acceptability of a non fully working oem integrated system and what it would take to get such a project working but no need to get into the weeds on that. If it's for a track toy I get it or hell even a more classic car that isn't anywhere near as computer controlled, but for a nice street car Ie M5/M6, not a clapped out beater, it doesn't make sense to me unless the whole package works.

Even if someone tackled the DCT swap I'd wager that fair prices parts and labor to complete super low volume one off swaps would exceed the value of an E60 not to mention being so much more expensive.difficult than a full blown manual conversion or to go through and repair an SMG at a shop.
 
  • Like
Reactions: Emfiver

·
Registered
Joined
·
285 Posts
The thing is an SMG3 in good health and driven by someone who has some understanding of how they work is actually not bad.

They are pretty horrible in automatic mode but I drive mine 99.99% of the time manually and in s5 or s6.

They aren’t the greatest for parking on a hill or other precision parking tasks but there are certain styles of user input that helps. You have to be fairly deliberate and precise with commanding throttle. Issues start when people are too timid on throttle (so smg is still slipping the clutch and car not really moving) then they apply too much throttle right as the smg decides to fully close the clutch. At that point you get a sudden jerk forward and the user then typically overcompensates by lifting too much off the throttle and the whole cycle starts again. You just need to learn to speak to it (with the throttle) in a language it can understand.

When you give it firm, decisive input it behaves pretty well. Once you get a feel for how fast it will engage/disengage the clutch based on throttle input you can basically make it do what you want. By ‘manually’ shifting the gears you remove one more variable. Because all of the clutch issues get compounded when you ask it to guess which gear you want too. It may think you want to short shift to 3rd right as you’re ready to punch it in 2nd and you then get a bonus upshift/downshift likely followed by traction control intervention.

So I’d agree if you want to drive a lot in automatic mode the smg3 sucks. But honestly I hate driving cars without at least choosing gears myself. I drove my DCT car 100% in manual and drive my m140i (zf8) 100% in manual too. So to me driving the e60 in s5 with the paddles (or the selector) is totally fine. And full throttle up shifts from 2nd to 3rd (or even 1st to 2nd) with dsc fully off and the tail hanging way out are magic. You can hold it in a lovely, big, fat slide without the upshift unsettling the balance (or attitude) of the car more than a hair. That’s near impossible in a manual and dct/dsg’s sometimes freak out in those kind of shenanigans.

I think if you really hated the smg and wanted a more reliable/affordable long term option then a manual swap is the best bet.

The dct swap to me doesn’t offer a massive gain and the cost to do it ‘right’ would far exceed the value of the car.
 

·
Registered
Joined
·
432 Posts
i agree you need to know how to drive manual to properly operate an smg. i still find it as bad as a real manual in stop&go traffic. as a car, this is where it fails, in a parking lot, in traffic, etc. i'm over 50% throttle only for a few seconds during a drive. it shifts like a pro when TPS is reading high. mayb we can fool it to think TPS is always 100%, but i'm not a programmer.

i will disagree dct swap has to 'far exteed" the value of the car. bolting it in isn't the problem. it's all in the programming. someone will figure it out, hopefully one of us.

i still think the best route is to convert smg-3 into a manual so we can throw out the hydraulics and operate smg3 with a third pedal. it may need some shift fork or gear work due to the weird shift pattern.. or some trick shifter box like s1sequential that will allow us to manually throw the gears in smg3.

i personally got used to the smg-3 during my ownership. it has been reliable to me. just change out the hydraulic fluid every 50k'ish along with trans fluid/filter. even if she blows up tomorrow, this is the best car i ever had for the money i spent and she owes me nothing.
 

·
Registered
Joined
·
368 Posts
i will disagree dct swap has to 'far exteed" the value of the car. bolting it in isn't the problem. it's all in the programming. someone will figure it out, hopefully one of us.
Between buying the dct, getting a custom driveshaft I imagine, who knows what little odds and ends, trans controller, interior trim for the shifter so it doesn't look like @$$, and then adding in the hours in labor to not only get it in but troubleshoot after? I'd be super surprised if it would come in under 25k.

Obviously this assumes paying whoever figured this out/a professional shop to do it for you, and not even tackling any while you're in there jobs/upgrades.
 

·
Registered
Joined
·
432 Posts
m3 dct is $2000 retail. custom brand new shaft is $600 from driveshaft shop. say $1000 conservative. used TCU is not expensive. where u getting the extra $20,000? the shop's kids college funds?

it's all in the programming. the hardware is all readily available. IMHO
 

·
Registered
Joined
·
1,217 Posts
They aren’t the greatest for parking on a hill or other precision parking tasks but there are certain styles of user input that helps. You have to be fairly deliberate and precise with commanding throttle. Issues start when people are too timid on throttle (so smg is still slipping the clutch and car not really moving) then they apply too much throttle right as the smg decides to fully close the clutch. At that point you get a sudden jerk forward and the user then typically overcompensates by lifting too much off the throttle and the whole cycle starts again. You just need to learn to speak to it (with the throttle) in a language it can understand.

When you give it firm, decisive input it behaves pretty well. Once you get a feel for how fast it will engage/disengage the clutch based on throttle input you can basically make it do what you want. By ‘manually’ shifting the gears you remove one more variable. Because all of the clutch issues get compounded when you ask it to guess which gear you want too. It may think you want to short shift to 3rd right as you’re ready to punch it in 2nd and you then get a bonus upshift/downshift likely followed by traction control intervention.
It is the low speed stuff that is really sketchy, like creeping up to a parking barrier on a slight incline, or just parking in the garage nosing up the last few inches. For me I drive the SMG and LFB all the time which is why it will be quite a bit faster for me on the track, but also LFB to ensure I don’t drive through the effing wall, as I don’t fully trust the thing because you do have to be firm with the gas in a way that you never would in a manual or conventional auto. I have a 6MT also and it’s so much more engaging. Every time I get in the SMG, I naturally depress my left foot on the non existent 3rd pedal. It does mess with my mind, since it’s the same seating position and same car as the 6MT. I lived in SF for 8 years with a manual trans as a DD, it does really test your hill holding and coordination skills, especially at the stop lights where you can’t see over the front of the hood because it’s so steep and you have a little roadster.
 

·
Registered
Joined
·
368 Posts
m3 dct is $2000 retail. custom brand new shaft is $600 from driveshaft shop. say $1000 conservative. used TCU is not expensive. where u getting the extra $20,000? the shop's kids college funds?
I have serious doubts that all the hardware is plug an play with just reflashing whatever software tweeks it takes. As with all swaps it's the labor that gets you if you're doing the swap itself it's your time, if you're paying a shop it's your wallet. Not to mention the trouble shooting, as if you could bolt everything right up, flash your car like an off the shelf tune and drive away.

To have it all working fully integrated you don't think there's going to be an awful amount of disassembly to run wires/harnesses to look clean inside the car and out of the elements under the car? Any custom fabrication for brackets/adapters let alone any modification required? What about the amortizing of time and resources it took to figure it all out?

If you're getting the job done at cost alone, I guess I could see around 10-15k for all parts and labor. Don't know what kind of shop, which has the skills and knowledge to be first to figure it all out, would charge you at breakeven prices.

Compared to what professional shops charge to do tried and true "easy" conversions, I think 25k to be an early adopter is a fairly conservative estimate.
 

·
Registered
Joined
·
375 Posts
I have serious doubts that all the hardware is plug an play with just reflashing whatever software tweeks it takes. As with all swaps it's the labor that gets you if you're doing the swap itself it's your time, if you're paying a shop it's your wallet. Not to mention the trouble shooting, as if you could bolt everything right up, flash your car like an off the shelf tune and drive away.

To have it all working fully integrated you don't think there's going to be an awful amount of disassembly to run wires/harnesses to look clean inside the car and out of the elements under the car? Any custom fabrication for brackets/adapters let alone any modification required? What about the amortizing of time and resources it took to figure it all out?

If you're getting the job done at cost alone, I guess I could see around 10-15k for all parts and labor. Don't know what kind of shop, which has the skills and knowledge to be first to figure it all out, would charge you at breakeven prices.

Compared to what professional shops charge to do tried and true "easy" conversions, I think 25k to be an early adopter is a fairly conservative estimate.
Wiring is easy. It's relatively modular on newer BMWs and the electromechanical stuff is all self-contained, so you really only need a few bus lines going from the transmission to the rest of the car. Transmission bolts onto the block. Transmission brace bolts onto the chassis. That's all already been verified long ago. Driveshaft would probably have to be custom, but that's fairly cheap.

The software is the hard part. If you're figuring out the the software on your own, I'd venture to say the swap can be done in the $5k range.
 

·
Registered
Joined
·
368 Posts
Since it's so cheap and easy to do since apparently the only hang up being software, I anxiously await ratdog or whomever else to prove me wrong. I won't need any prompting to admit I'm wrong .

Until then I'll be waiting with the rest of you.
 

·
Registered
Joined
·
375 Posts
Since it's so cheap and easy to do since apparently the only hang up being software, I anxiously await ratdog or whomever else to prove me wrong. I won't need any prompting to admit I'm wrong .

Until then I'll be waiting with the rest of you.
You act like software is just an after thought and trivial to take care of. It’s not. That’s 100% what’s been holding back everyone. The mechanical bits are easy.
 

·
Registered
Joined
·
368 Posts
You act like software is just an after thought and trivial to take care of. It’s not. That’s 100% what’s been holding back everyone. The mechanical bits are easy.
That's probably my fault for my heavy sarcasm not coming through clearly. Typed words and all that not conveying inflection/tone/etc...

Given our exchange on this thread I'm puzzled where you'd think I would believe that given my multiple previous posts stating that the software side is the major hang up for even professional companies who can pay people to work on things full time. I really don't want to come off too insulting, but I made that point painfully clear multiple times, I have to assume there's some breakdown in communication with me not getting my point clearly across to you. It makes me feel like you didn't even fully read what I've written which makes internet arguments with strangers much less "fruitful." If I had feelings, they might even be hurt.

I think the vast majority of people who come here every once and a while asking about a DCT underestimate both the hardware and software hurdles to do this swap right, (including you). In fact I think it's such a great hurdle that until someone wants to actually show their work in something that isn't for all intents and purposes a gutted "race" car, it's a pipe dream conversion. As in practically impossible, not literally.

There's no need for me to rewrite my previous posts to explain why, you can just re read what I've already written if you want further clarification on what I think and why regarding this. At this point I feel like going further would be more or less going in circles.

Should rat dog want to prove everyone who doubts this wrong they easily could given that they've been honest and actually figured it all out.

Until then, if it ever comes. (psst, spoiler alert, I don't think it will, want to make sure my sarcasm comes through clearly.)

Cheers
 

·
Registered
Joined
·
285 Posts
i agree you need to know how to drive manual to properly operate an smg. i still find it as bad as a real manual in stop&go traffic. as a car, this is where it fails, in a parking lot, in traffic, etc. i'm over 50% throttle only for a few seconds during a drive. it shifts like a pro when TPS is reading high. mayb we can fool it to think TPS is always 100%, but i'm not a programmer.

i will disagree dct swap has to 'far exteed" the value of the car. bolting it in isn't the problem. it's all in the programming. someone will figure it out, hopefully one of us.

i still think the best route is to convert smg-3 into a manual so we can throw out the hydraulics and operate smg3 with a third pedal. it may need some shift fork or gear work due to the weird shift pattern.. or some trick shifter box like s1sequential that will allow us to manually throw the gears in smg3.

i personally got used to the smg-3 during my ownership. it has been reliable to me. just change out the hydraulic fluid every 50k'ish along with trans fluid/filter. even if she blows up tomorrow, this is the best car i ever had for the money i spent and she owes me nothing.
Yeah it does take a bit of finesse to get the best out of it. I believe part of the reason you need to be confident and quick with the throttle is that convinces the ecu to release the clutch in a single quick motion. If you’re tentative on the throttle it seems to start modulating the clutch which gives us unpredictable amounts of engagement and results in the low speed jerking.

I get the hardware not being too expensive. But the integration/control part (software, wiring, programming/coding) seems pretty significant. When I said it would likely exceed the value of the car to ‘do it right’ I should clarify that to me doing it right means retaining the bmw dme, no dash lights, working DSC and no hacked up interior. Basically it feels and behaves like it came with the dct. Fairly good operation in auto mode and fast, direct shifts in manual mode. As mentioned overcoming the software issues to get everything ‘talking’ to each other is a big task. If a shop does figure that out they would want to recoup all that time and money they spent. So while there might be $5k in hardware and maybe $3k in labor for the physical conversion the integration part would likely be another $5k or even more. I’m pulling those numbers from thin air but I could be way off. As an early adopter I could see a bill at a workshop getting to $20k pretty fast at $100-$150 per hour to get it solved. Maybe far exceed was too much, but it does seem a long way from being economically viable still.

I do love the idea of converting the smg3 to a manual. That would be awesome. As you pointed out the shift pattern likely would prohibit it being a ‘H’ pattern but perhaps some kind of aftermarket shift solenoids could be used, possibly air/c02 powered with a fairly basic TCU and a manually operated clutch. I had a car with a hollinger sequential and it wasn’t the ideal ‘daily driver’ but it’s certainly drivable on the road. Hollinger now offer a powered shifter kit for their transmissions. There would probably still be a fairly big challenge it getting the car to behave as normal without controlling the clutch or trans but someone smart could maybe simulate the inputs/signals it expects to see. It’s also probably a pipe dream but removing the smg pump, plcd, shift solenoids, poorly insulated wiring harness and all the other common smg and clutch failure points would do wonders for the cars longevity and reliability. Mechanically the transmission seems pretty good.


It is the low speed stuff that is really sketchy, like creeping up to a parking barrier on a slight incline, or just parking in the garage nosing up the last few inches. For me I drive the SMG and LFB all the time which is why it will be quite a bit faster for me on the track, but also LFB to ensure I don’t drive through the effing wall, as I don’t fully trust the thing because you do have to be firm with the gas in a way that you never would in a manual or conventional auto. I have a 6MT also and it’s so much more engaging. Every time I get in the SMG, I naturally depress my left foot on the non existent 3rd pedal. It does mess with my mind, since it’s the same seating position and same car as the 6MT. I lived in SF for 8 years with a manual trans as a DD, it does really test your hill holding and coordination skills, especially at the stop lights where you can’t see over the front of the hood because it’s so steep and you have a little roadster.
I’m with you there. Tight parallel parking on an incline can be ‘exciting’ to say the least. A bit of LFB is a good strategy as when you’re trying to park close to a car/barrier and get a little surge forward there is no time to jump off throttle and onto the brake with your right foot no matter how fast you are.

I do think a lot of the complaints about the smg come from people who haven’t driven a manual (or don’t very well) or who have poorly functioning smg’s (or both). I can see someone who is used to a conventional auto or dct being frustrated with it. Particularly if they drive it in auto mode a lot. As I said I never drive mine in anything but ‘manual’ but the couple of times I tried auto I can see why people would be unimpressed. It really is a car or manual mode the car is guessing what you want it to do with the clutch and in auto it’s now trying to guess what gear you want also. It often seems to guess wrong. Telling it which gear you want removes one variable and that does help.

I didn’t have the choice of a conventional manual here but even if I did I think I’d take the smg. It has its drawbacks for sure but it does feel like it’s the ‘right’ box for the car.

sunghyun7 said it’s the best car he’s had for the money and I feel the same way and I’ve been blessed to have some awesome cars often for awesome prices. I bought an R35 GTR on its launch in Japan for $65k usd. A beautiful R32 GTR for $10k usd. But the m5 is probably the most bang for buck even with all the $ on maintenance.
 

·
Registered
Joined
·
375 Posts
That's probably my fault for my heavy sarcasm not coming through clearly. Typed words and all that not conveying inflection/tone/etc...

Given our exchange on this thread I'm puzzled where you'd think I would believe that given my multiple previous posts stating that the software side is the major hang up for even professional companies who can pay people to work on things full time. I really don't want to come off too insulting, but I made that point painfully clear multiple times, I have to assume there's some breakdown in communication with me not getting my point clearly across to you. It makes me feel like you didn't even fully read what I've written which makes internet arguments with strangers much less "fruitful." If I had feelings, they might even be hurt.

I think the vast majority of people who come here every once and a while asking about a DCT underestimate both the hardware and software hurdles to do this swap right, (including you). In fact I think it's such a great hurdle that until someone wants to actually show their work in something that isn't for all intents and purposes a gutted "race" car, it's a pipe dream conversion. As in practically impossible, not literally.

There's no need for me to rewrite my previous posts to explain why, you can just re read what I've already written if you want further clarification on what I think and why regarding this. At this point I feel like going further would be more or less going in circles.

Should rat dog want to prove everyone who doubts this wrong they easily could given that they've been honest and actually figured it all out.

Until then, if it ever comes. (psst, spoiler alert, I don't think it will, want to make sure my sarcasm comes through clearly.)

Cheers
From your writing, including the very post I quoted, I get the impression that you're overestimating the hardware and underestimating the software. The hardware is solved. Anyone with a garage and some tools can bolt in a DCT transmission and wire it up and have it look fairly OEM. The wiring is just a mater of cross referencing a few things in the WDS and hooking it up. That aspect is far easier than you're implying. It just won't move.

The software on the other hand is not solved. One can have an okayish setup if they're willing to use a standalone DCT controller and run the ECU in manual mode. I do think the car would probably still drive better than a factory SMG3 M5 since SMG is one of the worst transmissions to ever exist. But if you want it to drive as well as a factory DCT BMW, that's going to take a lot of work. The pieces are all there, BMW did do a lot of work for us already. It's practically impossible for most people on this forum since most people on this forum don't even know what a hex editor is. But someone who actually knows how to reverse engineer software and has the will to spend their time recalibrating M3 software for the M5 can get it to work.
 

·
Registered
Joined
·
132 Posts
Looks like wiring would be the easiest part, even if you had to use a few resistors for paddle signal voltage division.

945621
 
481 - 500 of 501 Posts
Top