Any general discussion around the firmware, what is does, how it does it etc.
By Rocket
#63077
Recently, there is little information about the progress of the project in the public space. What is the future of speeduino? In what direction is the project moving? Will the mega2560 remain the main processor or will it gradually be pushed out? Has the mega2560 reached its limits in the context of implementing new functions? Maybe the "dual core" (two mega2560 uc for different functions for example one for injection and main functions other for ignition and other related functions) concept will be considered?
User avatar
By jonbill
#63078
Afaik, the mega is still "it" for now and hasn't completely run out of room for new stuff.
Eventually it will be superceded I imagine, but no eta for that.
I've not seen any discussion of dual mega as the way forward. stm and teensy seem to be it.
By miker
#63079
The mega is old tech, but it still works, has scope for enhancement and it's what I develop on. The future is teensy or stm. Both are supported in the code and have boards made for them. If we'd not had the pandemic and the chip shortages I imagine 50% of the boards would be teensy based now. However we have and they're not.

Id wager well not see dual mega as it would mean a completely different architecture and codebase which wouldn't work with the existing chips. To support it we'd split development which would slow down any new features.
User avatar
By PSIG
#63083
I think the question is a good one. This question moves from technical, to theory and philosophy.

I do not think the question is whether or not the Mega2560 processor is "good enough", as it has proven to operate many projects rather well, and will still do so when you pull your project from the barn in 40 years to show your Grand-kids. From that position more hardware advancement is not required to satisfy that segment of the community and applications. But not all issues are satisfied with speed and features. In-fact, none of them, without a world-class foundation to build from. Speeduino is very close to this level. :D

I do think that expansion to envelop more projects, specialty applications, and future-proofing is a smart idea. While I would roughly estimate more than 80% of conventional EFI applications could be served acceptably with the Mega, that still leaves a fast-growing 20% that aren't or can't. This is the segment of users and applications that could be addressed moving forward. However, I do have an issue with expending substantial resources and efforts that direction - just yet.

I have said many times over many years, that Speeduino does not need to do everything - it has to do what it does very well. To plant a flag on awesome, it has to be that, and growing more features and capabilities without optimizing the basic functions to awesome-level will be less than spectacular. You need a crack-free foundation to build a taller building. Speeduino deserves to be more than a yawn or mention in the field of EFI. This IMO is what is holding this project back from stardom at the moment, and preventing it from 'floating to the top'. It's OK, and yeah it works, but not universally recommended everywhere as "get Speeduino - it's awesome and if anything can run it well, Speeduino can." Any system looking to be world-class (or even survive) needs that status, and Speeduino is not exempt. I'm not here to find in 10 years they are saying "It's too bad about that Speeduino thing fading out." This is future-proofing.

Speeduino is a firmware, not an ECM or hardware. Perhaps it's time for a tweaked v2. That firmware should be the best thing we can use to do the job. It should work as best as possible, for what it does, every time. Room for fairly minor improvements to the foundations has been stated all along, and I believe that is where energy should be focused, and then on more 'cool stuff', including processors and features. From that perspective, it will not matter if it is running in a Mega, STM, or other processor, and that forward hardware and feature progression will then come naturally and quickly. All IMO, and I'm always open to different perspectives. 8-)
By Rocket
#63094
I've been using speeduin for probably about 5 years, with mega uc without any major problems. As far as I have been watching the discussions, the main limiting factors are the amount of memory and timers. Two separate processors could greatly expand the possibilities. The ignition processor could do the heavy work with the trigger and ignition calculations and the necessary data could be transferred via serial3 to the second processor which could perform the iinjection function and other related functions, thus it would be possible to extend engine support to v12 engines as well. With custom firmware, I also have seen a sequential 6-cylinder ignition. The appeal of the Mega2560 is its accessibility, simplicity, cheapness and reliability. What I couldn't say about teensy. Regardless of the availability and price, have alternative processors already proven their reliability? I haven't tried a 3.3v system with level shift yet. Aren't they emf sensitive? Earlier I saw that several processors were tested, is there already a clear leader? I personally miss a few little things that I could implement in the official code for the mega2560 processor, probably the only function that simplyy could not be implemented in mega is sd card datalogging. But overall, I am completely satisfied with the mega2560. Where is the main focus in firmware development now? I don't follow discord, I don't know what's going on behind the scenes, there has been no updates for a while.

Will this have an updated version about this featu[…]

See my earlier post . The Base Tune sets t[…]

Vr Conditioner Noise when cranking

I did the fix, but it's not work. I put my setti[…]

Perhaps some different points and pictures. Instr[…]

Still can't find what you're looking for?