Is anyone else interesed in a non-EDIS megajolt lite setup?
I'm going to be fitting my mid-70's Honda twin bike w/ electronic ignition and I like the idea of the megajolt lite jr, but the extra modules won't do. Being a twin bike with 180-degree crank and 2 carbs, I don't think MAP will be what I'm looking for either so what I am considering for my project is a TPS only setup w/ 2 built in coil drivers in a small and watertight package. I think the design of the microsquirt and the megajolt lite jr can be combined to provide this, but I'll need to do some more research and testing. For now, I'd skip the SMT and 35P ampseal of the microsquirt as it adds cost and complexity. If there are others interested I can see about making concessions for other components, but if I'm the only one who would be using it, I'm going to strip down everything to just what I want for the smallest package possible.(Might even forgo the TPS and run RPM only based timing... but that's if I get lazy)
Any interest?
MircoJolt Lite Jr
Moderators: JeffC, rdoherty, stieg, brentp
If you know your stuff it is
If you know your stuff it is possible to take out EDIS module and Megajolt out of boxes and form a single and compact unit that you can tuck under the seat, tank, or where ever you going to put it.
-
Oliver Sedlacek
- Posts: 68
- Joined: Mon Sep 13, 2004 8:45 am
- Location: The Chalfonts
- Contact:
Opinions
I think a non-EDIS megajolt would be quite a step forward, but I would make the following comments.
Designing coil drivers which don't interfere with a micro takes as much development effort as the whole megajolt.
Designing the missing tooth crank sensor circuitry is not trivial.
The 3D mapping is one of the key benefits of electronic ignition, so I would not drop it from the design, even if you don't end up using it on your installation. The TPS circuitry is so trivial, I don't see why you would remove it.
I would also stick with SMT because it's really so much nicer. Soldering SO devices isn't difficult.
Designing coil drivers which don't interfere with a micro takes as much development effort as the whole megajolt.
Designing the missing tooth crank sensor circuitry is not trivial.
The 3D mapping is one of the key benefits of electronic ignition, so I would not drop it from the design, even if you don't end up using it on your installation. The TPS circuitry is so trivial, I don't see why you would remove it.
I would also stick with SMT because it's really so much nicer. Soldering SO devices isn't difficult.
mrbell, I can't comment on
mrbell,
I can't comment on the eletronics as I'm just a guy who can follow directions and solder when it comes to this stuff, but I'd definitely be interested. I have a 1965 Honda CL72 (250cc twin) and a 1959 AJS Model 31 (650cc Brit twin)project that I could use them on. If you build it, I'll copy it. I hate to be a mooch but I really can't help on the electronic tech stuff. At least you'd have my gratitude and that of those who be sure to follow.
I can't comment on the eletronics as I'm just a guy who can follow directions and solder when it comes to this stuff, but I'd definitely be interested. I have a 1965 Honda CL72 (250cc twin) and a 1959 AJS Model 31 (650cc Brit twin)project that I could use them on. If you build it, I'll copy it. I hate to be a mooch but I really can't help on the electronic tech stuff. At least you'd have my gratitude and that of those who be sure to follow.
Jax2A
I considered assembling the
I considered assembling the EDIS and Megajolt into a single unit, but it seems a bit of a hack. I'm big on purpose built hardware.
As for designing all of the extra circuits, most of that work has been done. The only hard part is making sure they work with the megajolt processor. From what I can tell, it should.
I was also reading about the MAX232 chip and it certainly seems most common, but I've also seen other chips that don't require the external circuitry that the MAX232 does. Does anyone know why those aren't used? I have a feeling they're less available or more expensive.
If this actually happens, I have a feeling that hand soldering SMT components is beyond alot of people, or at least alot of people's equipment. I've certainly burned up my share of stuff trying that. There is a point that shrinking the circuitry only introduces problems. All of these components could theoretically fit on a board about the size of a silver dollar, but then thermal management becomes an issue and the plug is likely to be larger than that anyway. Standard components are easier to work with and find anyway.
As for designing all of the extra circuits, most of that work has been done. The only hard part is making sure they work with the megajolt processor. From what I can tell, it should.
I was also reading about the MAX232 chip and it certainly seems most common, but I've also seen other chips that don't require the external circuitry that the MAX232 does. Does anyone know why those aren't used? I have a feeling they're less available or more expensive.
If this actually happens, I have a feeling that hand soldering SMT components is beyond alot of people, or at least alot of people's equipment. I've certainly burned up my share of stuff trying that. There is a point that shrinking the circuitry only introduces problems. All of these components could theoretically fit on a board about the size of a silver dollar, but then thermal management becomes an issue and the plug is likely to be larger than that anyway. Standard components are easier to work with and find anyway.
i have thought about doing
i have thought about doing this using megasquirt and spark extra code in a stripped down megasquirt and using VB921 ignition drivers and the trigger wheel decoder from the newer version of megasquirt.
It would be nice to have a striped down version of the megatune software that only has the ignition and logging options and no fuel settings.
Edis units are already getting less available in the local scrapyards as the cars they were fitted to are getting older now and rarer in the yards, so another option would be good.
Paul.
It would be nice to have a striped down version of the megatune software that only has the ignition and logging options and no fuel settings.
Edis units are already getting less available in the local scrapyards as the cars they were fitted to are getting older now and rarer in the yards, so another option would be good.
Paul.
-
Oliver Sedlacek
- Posts: 68
- Joined: Mon Sep 13, 2004 8:45 am
- Location: The Chalfonts
- Contact:
Trigger wheel interface
If the megasquirt has a decent trigger wheel interface, then that's obviously the way to go. I would check that it is accurate enough for ignition work, as the timing requirements for fuel injection are much more relaxed.
Hrm... So I was considering
Hrm... So I was considering starting from the mjlj and adding the necessary circuits(and removing some unnecessary ones), but it now occurs to me that the wheel decoding is done by the EDIS module. I wonder if the mjlj processor has the horsepower to handle that. It may be better to start w/ the full MS and strip things out.
It would be nice to see the
It would be nice to see the Megajolt progress in this way, so as not to be dependent on seperate Edis units. Im sure it could be done but the processor would need to be upgraded. i went to the breakers yard today and found that Edis units are definitely getting rarer, so it would be a good thing to develop for future use especially as many people dont need or want fuel injection but do require a mappable ignition system.
The megasquirt interface and software is fully capable of supporting ignition.
Paul
The megasquirt interface and software is fully capable of supporting ignition.
Paul
So, Jax2A... want to sell me
So, Jax2A... want to sell me that CL72? hehe
-
Oliver Sedlacek
- Posts: 68
- Joined: Mon Sep 13, 2004 8:45 am
- Location: The Chalfonts
- Contact:
VB921
Have you got a source for VB921s?
mrbell/CL72
I'd sell it, but only for crazy money. It's far from original. It was in boxes when I got it. Threw it together just to have something to ride but built it as cafe style. Still needs some cosmetic work, but my interests went to other projects. I have son that's almost five, I'll ride it occasionally but I'm hanging onto it for a project for him. I do have a bunch of CL72 parts I'd sell though. Tank, fenders, seat with one small tear and other small pieces.
Jax2A
MJLJ Version 'next'
I've been thinking quite a bit about the next version of the MJLJ, and non-EDIS control is high on that list.
One could use the Megasquirt successfully, but would be a somewhat inelegant solution with all of the unused components. Plus, a dedicated unit could be made more compactly than the existing Megasquirt.
So with that, what would be the initial list of requirements?
Here are some candidates. I'm sure others will have their own feedback.
Non-EDIS support, compatible with same 36-1 trigger wheel
10x10 or 12x12 ignition map
on-the-fly MAP switching
Discrete coil drivers (4, 6 or 8?)
Coil on Plug support in addition to coil packs
Knock control
User outputs
Launch control
Turbo anti-lag
Coolant temperature integration
More robust serial API, provide programming DLL for people to write apps against unit
One could use the Megasquirt successfully, but would be a somewhat inelegant solution with all of the unused components. Plus, a dedicated unit could be made more compactly than the existing Megasquirt.
So with that, what would be the initial list of requirements?
Here are some candidates. I'm sure others will have their own feedback.
Non-EDIS support, compatible with same 36-1 trigger wheel
10x10 or 12x12 ignition map
on-the-fly MAP switching
Discrete coil drivers (4, 6 or 8?)
Coil on Plug support in addition to coil packs
Knock control
User outputs
Launch control
Turbo anti-lag
Coolant temperature integration
More robust serial API, provide programming DLL for people to write apps against unit
-
Oliver Sedlacek
- Posts: 68
- Joined: Mon Sep 13, 2004 8:45 am
- Location: The Chalfonts
- Contact:
My thoughts
Brent,
I think some of the features you've listed are quite ambitious, and it highlights the fact that a project like this should be easy for the users to develop in their own directions. With that in mind, what I'd like to see is a basic non-EDIS system, with the following features.
A processor which can be programmed without additional programming hardware
Plenty of spare RAM
Plenty of FLASH
Code written in C
Three or four analogue inputs
Three or four digital inputs
Two VR speed inputs
Two, three or four coil drivers
Some programmable outputs which are high side drivers, with a decent current capability
If the basic box has just megajolt+EDIS capability it will still be of interest to many users. It would be nice if there was a structured way in which users could then submit enhancements without fragmenting the development.
I think some of the features you've listed are quite ambitious, and it highlights the fact that a project like this should be easy for the users to develop in their own directions. With that in mind, what I'd like to see is a basic non-EDIS system, with the following features.
A processor which can be programmed without additional programming hardware
Plenty of spare RAM
Plenty of FLASH
Code written in C
Three or four analogue inputs
Three or four digital inputs
Two VR speed inputs
Two, three or four coil drivers
Some programmable outputs which are high side drivers, with a decent current capability
If the basic box has just megajolt+EDIS capability it will still be of interest to many users. It would be nice if there was a structured way in which users could then submit enhancements without fragmenting the development.
How about 60-2 tooth
How about 60-2 tooth wheel.....