Ruggedized ECU

[jbelanger]

This follows from the thread in the wish list section: http://forum.jbperf.com/viewtopic.php?f=12&t=50.

This will cover the development of a new ECU based on the Microsquirt module using the Cinch SE enclosure with the 48-pin connectors together with a board to carry the module and having an auxiliary CPU (mini IOx) to control additional I/Os and features based on the I/O Extender code and features.

The features and pin usage as of now are as follows:

• 4 pins for high Z injector drivers (2 on the module and 2 on the new board)
• 4 pins for logic coil drivers
• 6 pins for 2 differential VR inputs (VR+ and VR-) which can be configured as opto/Hall inputs with one pin per input used for pull up or shunt depending on how it is connected
• 4 pins 2 RS-232 serial ports (Microsquirt module and mini IOx)
• 2 pins for the CAN bus (for both Microsquirt module and mini IOx)
• 2 pins for main power supply: 12V and ground (may need to double that to reduce noise)
• 1 pin MAT
• 1 pin CLT
• 1 pin TPS
• 2 pins for 2 O2 sensors
• 1 pin Vref
• 2 pins for idle control: either 3-wire valve or 2-wire valve plus one spare output
• 1 pin fuel pump
• 2 pins MAP+Baro
• 1 pin bootload
• 1 pin Tach output
• 1 pin table switching
• 1 pin launch control
• 1 pin boost control
• 3 pins NOS (1 input + 2 outputs)
• 2 pins programmable outputs (may need to remove these if additional 12V and ground needed)
• 2 pin analog inputs (may need to remove these instead of programmable outputs if additional 12V and ground needed)
• 3 pins ground: 1 sensor ground + 2 high current grounds

In addition to this, there will be an on-board micro-SD card and a real time clock for logging purposes (once I finish this part of the IOx code...). The on-board mini IOx will take care of some of the aforementioned I/Os as well as logging.

The inputs will be configured to use the Microsquirt module circuit directly except for the tach inputs which will use a MAX9926 interface and the few optional inputs which will use an active low input circuit. The outputs will use the standard Microsquirt module circuits except for the ignition outputs which will use the pushpull circuit discussed recently on the MS3 forum (to allow the use of most, if not all, "logic level" COPs) and optional outputs which will use an appropriate driver for the feature. The mini IOx serial port could be used for reading the digital data from an LC1 chain (or other WBO2 serial controller protocols yet to be implemented such as PLX and TechEdge); this will free one ADC input.

I'll update this as things progress. If you have any preference on the aforementioned options in case some I/Os need to be cut, post you comments.

The planned number of units for the initial release is from 6 to 8. If you have an interest in having one, please mention it here (24c and tmsmini have expressed their interest). Depending on the interest, there may be some future development and more units or these may be the only units produced. I'll confirm a price target as more data becomes available.

Jean

[dontz125]

If using the 'vented' enclosure, could Const Baro be made an on-board feature? The MPXHZ6130A is pretty small, and this would free up a connector pin.

Will the mini-IOx (with on board micro-SD & clock??) be offered as a discrete product?

Can the 3 NOS pins be used for other functions if not needed for NOS?

[jbelanger]

If using the 'vented' enclosure, could Const Baro be made an on-board feature? The MPXHZ6130A is pretty small, and this would free up a connector pin.

I thought of that but I don't know if the vented feature means there is some time delay between atmospheric pressure and case pressure. If this still allows a significant pressure differential and/or it takes a while to equalize pressure then that would make having an on-board baro sensor dubious at best. But I'm not dismissing it and will look at the available space when I am closer to a final layout. One advantage of dedicating a pin for it is that the ADC can be used for a different function if a separate baro sensor is not wanted/needed.

Will the mini-IOx (with on board micro-SD & clock??) be offered as a discrete product?

The mini IOx is an integral part of the board carrying the module, the components and the connector so it will not be possible to make a direct separate product out of it.

Can the 3 NOS pins be used for other functions if not needed for NOS?

Yes. All the pins that are shown for a specific extra feature will be available for other functions the same way they are on other MS product through the TS user interface. However, each pin will have a single specific circuit so the input or output functionality will be fixed. So in the case of the NOS pins, you will have one active low input (on PA0) and 2 low side drivers (on the mini-IOx).

Jean

[24c]

...Can the 3 NOS pins be used for other functions if not needed for NOS?

Are you thinking power valve, EXUP, and/or quick shifter?

[24c]

I thought of that but I don't know if the vented feature means there is some time delay between atmospheric pressure and case pressure. If this still allows a significant pressure differential and/or it takes a while to equalize pressure then that would make having an on-board baro sensor dubious at best.

This link about the breather vent might help, key bit snapped below.

breather_vent.png (28.45 KiB) Viewed 33053 times

[dontz125]

I've thought about that aspect some more. There shouldn't be an issue with response time (we're talking baro pressure here; unless you're driving off a cliff, the pressure isn't changing THAT fast), but there may be deadband issues, some residual dP that can't overcome the stiffness of the membrane - it's meant as a breather, after all, not an instrument port.

While it would protect the engine and correct for large swings in altitude, I would suggest that there will always be some untunable irregularity. This may or may not be significant.

[jbelanger]

Mike, thank you for the link and quote. That seems to indicate it would be sufficient. However, I have the same feeling as Don on this and am afraid it could bring some problems that would be difficult to solve. Moreover, having a connector pin for the ADC input gives more flexibility for those who don't need or want a baro sensor.

Jean

[24c]

However, I have the same feeling as Don on this and am afraid it could bring some problems that would be difficult to solve.

OK Jean, but I would still consider the case with a breather vent though, as it improves the heat dissipation, according to their site info.

Personally, it doesn't matter to me, because I would just take the initial reading for say baro off the MAP sensor, and if I was into altitude shifts/jumping off cliffs fit a second baro & a chute!

Mike

[jbelanger]

Good point and I'll consider it. The issue might be availability since it seems to be more difficult to get the case with the breather vent. It is also significantly more expensive. In the case of the ME enclosure, it doubles the price, at least from the distributor who has both part numbers in their catalog (neither are in stock at this time though).

Jean

[jbelanger]

I have made quite a bit of progress on this. And due to a change of enclosure from the SE to the ME which does not require a special tool to open or close the box, I have been able to move some things to jumpers inside the box and make a better use of the pins. But the smaller size of the board has proven to be challenging for layout and routing.

Taking the original list in the first post, the updated features and pin usage as of now are as follows:

• 4 pins for high Z injector drivers (2 on the uSM and 2 on the new board): each driver can drive one or 2 injectors
• 4 pins for logic coil drivers with a jumper for a 5V or 12V output signal: each driver will easily handle 1 or 2 coils
• 4 pins for 2 differential VR inputs (VR+ and VR-) which can also be configured as opto/Hall inputs using a jumper for a 5V or 12V pull up or to set the resistor as a shunt
• 4 pins 2 RS-232 serial ports (uSM and mini IOx)
• 2 pins for the CAN bus (for both uSM and mini IOx)
• 2 pins for main power supply: 12V and ground
• 1 pin for sensor ground
• 2 pins for high current outputs: 12V (dirty flyback return) and ground (high current outputs, ignition outputs, additional injector outputs)
• 1 pin MAT
• 1 pin CLT
• 1 pin TPS
• 2 pins for 2 O2 sensors (uSM O2 and spare ADC 1)
• 1 pin Vref
• 2 pins for idle control: either 3-wire valve or 2-wire valve plus one spare output (2 mini IOx outputs)
• 1 pin fuel pump
• 2 pins MAP+Baro (uSM MAP and spare ADC 2)
• 1 pin Tach output
• 1 pin table switching (uSM PE1 through an active low input circuit)
• 1 pin launch control (uSM Flex through an active low input circuit)
• 1 pin boost control (uSM IDL)
• 3 pins NOS (1 input (uSM PA0 through an active high input circuit) + 2 outputs (on mini IOx))
• 2 pins programmable outputs (on mini IOx)
• 3 pin analog inputs (on mini IOx)
• 2 pins I2C bus extender (mini IOx could be used with external SLC OEMs)

The bootload pin has been removed in favor of a pair of jumpers: one for the uSM and one for the IOx. Neither of them should be needed under normal circumstances since both will have code with a jumperless feature (standard on MS2/Extra and the regular IOx code). And the on-board micro-SD card and a real time clock for logging purposes are still there and still require me to finish this part of the IOx code.

It should be noted that while the board will have circuits that will allow all the aforementioned features, the extra features will not be hardcoded but will remain as flexible as they are in the MS2/Extra code. And the additional outputs provided by the mini IOx are all using hardware timer pins so it will be possible to use PWM on them with almost any frequency without losing precision like you do on a software PWM on an MS2 or uSM output.

So there is still a bit of work to do to finish this and I still need to make sure it will all fit inside the ME enclosure. But if you are interested, let me know as soon as possible because this will be a limited release at least initially. And if you see something missing or wrong, let me know and I'll see if that makes sense in this context.

Jean

[24c]

Excellent news...thanks for the update.

Like the idea of using a jumper rather than bridging solder pads, easier to configure or change.
Smaller enclosure is good for motorcycles/quads etc.
Differential twin VR inputs work well for me, as you know.
Realistically what do you think is a likely release date? No rush, as rather everything was well thought out (no doubts here), but it'd be nice to have a countdown.

[jbelanger]

Actually, my main constraint for a release is the enclosure availability. It is back-ordered at this time so I don't have one in my hands to do the final checks but that should be corrected within a couple of weeks if I got the correct information. And it's also likely that when I'm ready to go ahead, there will be another delay to get the enclosure for the release batch.

So in the best of cases (which is honestly think is not very likely) it would be available in about a month. In the more likely case that there are more delays, it could be 2-3 months. And if I find out that there are more space constraints than the enclosure data sheet indicates, then it would be more problematic and might even require going back to the SE enclosure with the additional constraints of a box that can't easily be opened and closed.

Jean

[tmsmini]

Still interested as family issues have slowed progress on my own assembly of the MS I purchased along with your components.
Let me know and I will send funds when needed.

Terry

[jbelanger]

Will do.

Jean

[tmsmini]

Any update on the status?
Terry

[jbelanger]

I have a few boards here but I'm waiting for the enclosure. The last word I got was that it was due to be available by the end of the month. I'll assemble a board with a module in the coming days and I'll see if I can do some tests.

Jean

[aarc240]

Was there any consideration of more than 4 cylinders?
This would have been of great interest to me had it been capable of 6.

[dontz125]

Jean's design is based on the MicroSquirt Module, which can be configured with 4 spark and 4 ignition channels. If you want to run wasted-spark and semi-sequential (or batch) injection, I would suggest it can handle an 8-cyl engine.

[jbelanger]

As Don says, this is based on the Microsquirt module so it can't have more than 4 channels for both ignition and injection but that also makes it capable of handling 6 and 8 cylinder engines in wasted spark and semi-sequential injection. Each injector driver can handle 2 injectors and the ignition drivers are logic level driver than can handle 2 COPs, CNPs or external ignitors.

The only Megasquirt product currently available that can handle 6 or 8 cylinder engines sequential injection is MS3.

Jean

[SordFish]

I didnt see knock input mentioned, was it for a reason? I've not used it yet but its something I was looking into for my next build.

apart from knock it sounds perfect as long as its not too expensive I'll put my name down for one too.

The only thing bothering me is ms2 is getting on a bit now and the new features of ms3 are calling, what are the odds of an ms3 module showing up anytime soon?

also any chance of a peek of the boards?