esv wrote:If such a board becomes available (please, make that happen James and Jean ), will it be able to be used with the forthcoming Sequencer?
In the sense that the Sequencer has CAN, it could but I don't know that the firmware will be able to use it, at least not at first. This is mainly a MS2/Extra thing (but James is better placed to talk about this than I am).
As for making it happen, there's going to have to be more people interested. Up to now, you seem to be the only one really interested in the board. And I will have to have confirmation that James is interested (and has time) in going further on the firmware side before I will commit to making a board.
But once those things are done, the board itself will not take long to become a reality. It would just be a bunch of known circuits put together on the same board with some way to connect them to the CPU's I/O pins. It would be a very basic board with a bunch of I/Os and all the CPU pins available for either on or off board connection.
I don't think a 4 layer board would be necessary in this case. For a 2-layer board the same size as the MS board (4"x6") it would be about $35. If it can be made smaller (which is probable but may not be convenient), it would be cheaper.
As for shipping, I have been shipping bare boards through regular mail in padded envelops with very few problems. And the cost was between $3 to $5 depending where in the world ($3 for Canada and US, $5 for the rest).
If I can get 10 people interested, then it's worth doing. As I said, it would be better to have James word on what he intends to do on the firmware side.
At the moment we are in a somewhat "chicken and egg" situation. With no GPIO (just yet!) there is less motivation to write any CAN code, but without the CAN code written there is less motivation to make IO boards.
If Jean makes an I/O card to use the MS2 then I'll extend the code to make use of it. I am not promising to write all the fancy algorithms that people have requested in this topic, but as a minimum the code will allow the outputs to be flipped like the generic ports and the inputs to be read back into Megatune.
Comments and suggestions on what existing features of MS2/Extra could be extended to use CAN'd ports are welcomed. You might have noticed that realtime baro already has it.
I don't have any plans to put ignition on a remote MS2 - you can always use the future MS2-sequencer code to do this as Bruce mentioned on MSEFI.
racingmini_mtl wrote:I don't think a 4 layer board would be necessary in this case. For a 2-layer board the same size as the MS board (4"x6") it would be about $35. If it can be made smaller (which is probable but may not be convenient), it would be cheaper.
by my count you have 6 + maybe James interested at the moment.
if it was exactly half a ms board, you could get two into a single case if you wanted. that could be good for all sorts of setups that need a lot of io in a limited space.
what sort of $ would it be at 4x3? or do you not think thats possible to squeeze in?
I didn't want to put James on the spot. This answer is very good one and is sufficient for me to go ahead at least for the design part.
A 4x3 board would be about $25 but I'm afraid it would be tight to have the MS2 with 8 ADC circuits and 8 high current outputs. I'll have a look but I'm not sure at all. However, it would probably be possible to still have 2 full size boards in one case because the board could be put in a lower slot since there wouldn't be anything on the bottom.
I would actually prefer to not have things to tight and leave some space for a nice big proto area and the option of having a DB37 connector. I think those would be useful for such a board.
I would if you added a couple of "high side" high current drivers and use a DB connector instead of screw terminals and find a case to house it and the MS. My gearbox needs switched 12V to drive the solenoids not switched ground.
As per my previous post I would be interested in one but I would require a minimum of 2 high side drivers, 2 PWM outputs, a Hall input for speed sensor and 3 12volt digital inputs. I have already started working on a board with these features to control my gear box so I'm happy to help out with this project. I was trying to design a board simular to the duel MS1 board but with two MS2 cards. It looks as though it could be done by mounting the pots on the underside of the V3 board and cutting a whole in the top of the case to make room for the 2 MS2 cards. plus you would have to cut off the jumper pins on the MS2 cards.
I think you need to think about what really needs to be on this board and what size it will be. I think it should be made smaller with maybe limited on board circuits. It sould be very much based around DIY circuits that arn't offered by GPIO or a second MS2.
I'll have a look at putting those on the board. However, keep in mind that the goal is to have an I/O extender board so controlling a gear box may be somewhat out of scope. I don't want to compete with the GPIO board on that front.
If you go with your own solution, you would be better off using another case instead of doing all those modifications. Using an EAS-400 would give you almost one inch more height to play with.
As for size, I'm not sure which way to go. On one hand it would be nice to have a smaller board but on the other hand, if you can't do much with the board without having to go with yet another board, then what's the use. I'm leaning towards having a bigger board that can be populated as needed but if that's not what people want then I'll do the smaller board.
Your idea sounds good. Even with the circuits I requested it will still be an IO expander, I will just be using the IO to control the gearbox. I plan to use James's Transmission code anyway.
I would be happy to stop what I'm designing and help you with you design if it will do what I need. I have already started thinking it would be easier to have the extra board outside of the MS case. My point was just that you don't want to design something that could just be replaced with a second MS or GPIO. Also GPIO is not suitable for controlling my gearbox as I need the high side drivers.
Let me know if you want any help or want to compare notes. I have aready done some schematics etc.
Hi Jean I was looking at your web site and I had an idea. What if you did a 2” x 4” board or simular with a 40pin dip socket and the other basics then had some 1”x4” add on boards like your duel VR board with high current drivers, high side drivers, input conditioning, hall inputs, etc. That way some of the existing expansion boards can be used as well. (Duel VR, “Error board”, etc). This way it would be highly configurable for end uses.
The MS2 daughter card has the end resistor (120 1/4W) as SMD part so as part of an I/O extender it needs to be the end node (ie, anything else that needs the CANbus will go between it and the main MS2). With the injector setup 2 of the timers/hardware PWM are not available on PWM1 & PWM2 (Port T2 & T4), and hopefully, you can still use IAC1 & IAC2 (Port T6 & T7) as Output Compare or PWM right through the PWM motor chip by disabling it with IACEnable (Port B4). On the 40 pin DIP, Pin 3-5, 10 & 15 are usable for Port E0 (via JP4), Port E1 (need to solder a jumper wire) and maybe Port E7 if trace cutting works on that pin.
KGB wrote:The MS2 daughter card has the end resistor (120 1/4W) as SMD part so as part of an I/O extender it needs to be the end node (ie, anything else that needs the CANbus will go between it and the main MS2). With the PWM motor driver and the injector setup not all the timers/hardware PWM are available - don't know if it's easy to get around it and regain some hardware timers
There are ways around that but it does involve the need to put some jumpers on the MS2 daughter board.
I have been thinking of using this as the basis of the board instead of the MS2: http://www.technologicalarts.ca/catalog ... cts_id=343. This would have the advantage of making all the CPU pins available and making the CAN bus terminating resistor optional. With the on-chip serial monitor, the existing MS2/Extra (or any other MS2-based) code could be loaded on the chip. It would also be about the same price as an MS2.
However, there is the ethical issue of using B&G based hardware design and firmware. I'll have to discuss this issue as well as the issue of using James firmware.
Hmm, that's interesting, but the ethical issues and it being "different" are likely more trouble than it is worth. If you really want access to all the pins on the CPU then make a custom board with a larger pinout CPU from the same family directly mounted to the board and skip the plug-in altogether.
I see the "quick and dirty" board using the MS2 card, a few circuits and some screw terminals as beneficial and a method for people who want to extend their systems and hopefully also develop some of the CAN extensions. Anything more and you'll be in conflict with the GPIO without adding any value.
If the quick board was half circuit and half proto area that would be even more DIY-able. Anyone who didn't want the proto section could cut it off anyway.
Don't be afraid of SMD either - aside from ICs with lots of pins, the other devices can be hand soldered without to much bother. i.e. use the VND5N07 and SMD base resistors. You can pack more onto a small board than using TO220 packages.
As Jean posted on MSEFI, the need to remove the 120R CAN termination resistor from the MS2 card if more than two devices are used in the CAN network is a bit of a pain, but hardly a show-stopper.
jsmcortina wrote:then make a custom board with a larger pinout CPU from the same family directly mounted to the board and skip the plug-in altogether.
Surely the ethical issues would be exactly the same. Isn't the rule "no b&g code or any derivative thereof can be run on hardware that isn't b&g" or similar?
If the quick board was half circuit and half proto area that would be even more DIY-able. Anyone who didn't want the proto section could cut it off anyway.
good thinking :-)
Don't be afraid of SMD either - aside from ICs with lots of pins, the other devices can be hand soldered without to much bother. i.e. use the VND5N07 and SMD base resistors. You can pack more onto a small board than using TO220 packages.
devojet wrote:Hi Jean I was looking at your web site and I had an idea. What if you did a 2” x 4” board or simular with a 40pin dip socket and the other basics then had some 1”x4” add on boards like your duel VR board with high current drivers, high side drivers, input conditioning, hall inputs, etc. That way some of the existing expansion boards can be used as well. (Duel VR, “Error board”, etc). This way it would be highly configurable for end uses.
that seems like a very good idea too, nice and modular and therefore flexible.
jsmcortina wrote:then make a custom board with a larger pinout CPU from the same family directly mounted to the board and skip the plug-in altogether.
Surely the ethical issues would be exactly the same. Isn't the rule "no b&g code or any derivative thereof can be run on hardware that isn't b&g" or similar?
I wasn't really suggesting going the different chip route, more saying that for a minimal board the re-use of the existing MS2 chip can solve a lot of potential headaches even if the available pins are somewhat compromised. No design work and huge code re-use.
I am NOT an official source on what does or doesn't go, but I would have thought that a simple board to sit under the MS2 would be just fine.
Definitely ask Al and Bruce. I think either idea sounds more flexible and DIY (the non B&G chip as well as a higher pin count CPU from the same family) than a MS2 chip, but even that sounds like a very useful way to get more IO at a palatable price.
Since the MS2 forms the basis of the device (you need an MS2 or GPIO, or sequencer) to be able to use it, and we are not porting the B&G code to another processor, it may be OK. It is similar to having a PIC or AVS chip as a basis for taking the can signals and passing them through as an extender, just a lot less work on the programming side as the CAN code is already done. A chip that is not compromised by having some pins blocked from use is preferable.
), will it be able to be used with the forthcoming Sequencer?
