How to make the move to an aftermarket ECU

In reply to Paul_VR6 (Forum Supporter) :

The onboard wide band was one of the things I really liked about the FAST XFI.  The computer could hold desired lambda so tightly that you could tune by looking at the real-time correction factor in the data log.

I also have had.... issues... with getting my AEM and my MS2 to give the same reading.  Am hoping wideband over CAN gets me most of the way to where I want to be.

For DBW, I want to be able to hold the throttle to a fixed percentage when antilag is active, I didn't see a way to do that, unless the controller can convert idle counts to a large throttle position.  Want to keep the turbo spooled while braking into a corner.

In reply to Pete. (l33t FS) and Paul_VR6 :

Can you guys clarify what the issue for on board WB is?  Seems like a very good thing to me- especially the non stoich control part.

In reply to alfadriver :

MS does not have it, must use external controller.

In reply to Pete. (l33t FS) :

That, I knew.  Is there no chance that it will ever go on?  Seems like it would be a natural board extension.  

alfadriver said:

In reply to mke :

For what?  Real use.  It's one thing to make sure the correct fuel is being delivered.  It's another where your right foot has enough control.  If your drive cycle is truly binary, then the mid-throttle control isn't that important.  But having been a calibrator for just about 30 years now, the hardest part of making a car move well is slow, parking lot stuff.  WOT is easy, stuff to WOT is easy.  Driving in a parking lot where the passenger isn't a threat to throw up is a challenge.  It's one of my pet peeved about our ETC- where there's a steep torque demand right off of zero- making slow moves much more difficult.

My example of the Alfa is also a 500cc cylinder.  Not 4V perhaps.  45 DCOEs are marginal on the street, and bigger is a waste of time when you have to park.

I do hear what you are saying and I agree with nearly all of it.  The only place we diverge is comparing weber sizing to EFI TB sizing...the webers MUST be smaller then EFI TBs need to be.

But yeah, getting ti to come off idle and roll around a parking lot looking for a spot smoothly is hard, really hard.  That is why I am so adamant about getting a good MAP signal, I've just never been able to make it work the way I expect it to work with just TPS, at least not with decent size TBs. 

On my setup I also use a table to translate pedal position to desired TPS, so I can set say the first 20% pedal travel to be only 5% throttle movement or what even % I want, just change the numbers in the table.  This is like progressive mechanical linkage but a lot easier to adjust.

Last  I read all 12 MAP sensors and compare them to the global MAP signal then trim each cylinder's fuel accordingly so linkage related errors cylinder to cylinder don't bugger up the mixtures and the engine runs smooth at all throttle positions.  This could be done with cylinder O2 too and that was my original plan but this neater and uses sensors I already have and seems to work pretty well so far but it's still only had limited testing.

A bonus of the cylinder MAPs is I have a TB sync screen on the PC....The actuator is attached to cylinder 12 so on the sync screen I see the 5 MAP error of every cylinder compared to 12 so adjusting the other 11 TBs to match is a 5 minute job.....lots of things can be done when more data is available.

In reply to mke :

Can I say it's amusing that you have a system to sync the 12 cylinders?  I hope you are not using any cables anywhere- the struggle to get two cable throttles to open at the same time was an adventure when we got the first V12 put into an Aston Martin.  And it didn't take a lot of heat to mess that sync up.  The result of that would be one bank that would go into idle or fuel shut off totally differently, and caused a lot of stalls when the other bank couldn't deal with it.

The end result was to force the two modules to sync up- which was not trivial since it was before CAN existed.

The next car (the Vanquish) fixed that by improving the communication and being electronic throttle.

alfadriver said:

In reply to mke :

Can I say it's amusing that you have a system to sync the 12 cylinders? 

why?  It runs smoother when the TBs are in sync.

I have motorcylce TBs so they are setup like a motorcyle.  A single electronic actuator drives all 12 (replacing the original pedal cable setup), but its is connected to #12 with a rod connecting to #1 on the other bank.  7-12 and 1-6  are motorcylce style, one shaft drives the next with a spring loaded adjuster. Look at the screen, adjust 12 thru 7, 1 to 12, 1 thru 6 and its done.  Cheap and simple.

But then it doesn't matter that they stay anything like in perfect sync because each cylinder fuel matched the MAP reading for that cylinder....when I first fired it up they were +/- 30% on the MAP readings and it ran fine (once I fixed my coding error so it was error removing instead of error doubling, Doh!).  I adjust them to +/-5% and call it done.  I do have it setup to light a malfunction light when they are out more than 10% just to let me know its time to re-sync them but that's just a nice to have.

I agree the things you're describing are problems, but a well designed control system solves them in short order.

alfadriver said:

In reply to Pete. (l33t FS) and Paul_VR6 :

Can you guys clarify what the issue for on board WB is?  Seems like a very good thing to me- especially the non stoich control part.

I have used the onboard O2 all of ONCE in the 36 installs I just counted where it was an option in the ECU. Most of those needed 2 channel wb so for consistency used two external controllers anyway. The next batch were using NGK sensors with either the old AFX or AFR500 controllers (from 1-4) and none of the ECUs chosen supported the NGK natively (Bosch chip). The next few were using 4-or more more O2 over CAN anyway for per cylinder readings. 

Even with all the MS ecu's I sell, I rarely need to include a WB as the owner already has one in the car so why use a second sensor or add a display just for that.

So it seems like the onboard O2 is so far down the list I usually don't pay it any mind. Maybe it's different for mild builds/turbo conversions, etc. 

In reply to Paul_VR6 (Forum Supporter) :

Not onboard O2, but onboard WB.  Is there anyone putting a WB controller directly attached to a MS system?

It's been almost 20 years since I've worked with a car that just used the basic O2 sensor, so I'd require a WB as part of the MS system.  IIRC, there were a handful of WB kits that were the basic board that theoretically would fit inside a controller box.  Just a matter of integrating that circuits into the whole box.

In reply to mke :

funny in the way that it reminded me of the struggles we had with the two cable throttles.  Since they were on the opposite sides of the engine, a mechanical linkage would have been, well.. almost impossible.  So there was a common crank that then change the pedal input to the two cables to the throttles.

The issue wasn't a/f from bank to bank, it was synchronising the transition from decel to idle.  Before CAN, the modules didn't have an easy way to talk to each other.  

Should we ask to have this set up as a seperate site?  
I for one would love to have one site to go to for all EFI questions and I'm getting the idea I'm not alone. 

In reply to Paul_VR6 (Forum Supporter) :

My experience has been the opposite, I have used the onboard wideband in 100% of the installs that used it (lots of FAST, one Hydra something or other for a Subaru)

I really like the NTK sensor.  It handles leaded fuel just fine.  Most of the cars I had a FAST on were fed a steady diet of C16.

alfadriver said:

In reply to mke :

 Since they were on the opposite sides of the engine, a mechanical linkage would have been, well.. almost impossible. 

My problem was the opposite, finding enough room to physically fit 2 rows of TBs... I ended up moving the ports in the head to make them fit 

alfadriver said:

In reply to Paul_VR6 (Forum Supporter) :

Not onboard O2, but onboard WB.  Is there anyone putting a WB controller directly attached to a MS system?

It's been almost 20 years since I've worked with a car that just used the basic O2 sensor, so I'd require a WB as part of the MS system.  IIRC, there were a handful of WB kits that were the basic board that theoretically would fit inside a controller box.  Just a matter of integrating that circuits into the whole box.

Sorry I wasn't clear I meant onboard wideband when I was saying o2 (havent used a narrowband for anything in a long, long time as well).  I did put an SLC-oem controller inside a ms case once but all it did was eat up all the spares pins and need an added connector for the spark outs and additional i/o. I couldn't find any advantage other than a slight packaging gain which in most installs isn't an issue. 

I guess an ECU that comes with a built-in WBO2 controller or 2 is a cost advantage but if you have to add it as an extra cost option and it takes pins from other stuff then not so much an advantage.  I do like having to mount less "stuff" though so even for extra cost  might consider it.  My ECU has 2 which I am using on the 2 banks like a normal person so when I bought it I liked that it had them and considered it a selling feature.  

....then I added 4 NBO2  in the quadrant collectors (its a street car intended to live most of its life at or near stoich so they seemed good enough to be useful working on low power issues) but I keep thinking of replacing then with WB and that would make the 2 I have redundant and the internal controllers wasted since I'd want the 4 to match.....but everything you need to buy adds up my wife keeps explaining to me 'Hon, I need to order an $80.....fine, I don't really NEED that right now I guess......"

Only considering the big ticket items is one my best tools for talking myself into really stupid projects "wait, I can buy a V12 for $4k?  I'm was going to spend more than that on the new blower setup so it would be basically free to switch to a V12!". .....yeah, not so much  

That's maybe an extreme example (but sadly it's true) but taking the time to plan out the project a little before starting is the the best way to pick out an ECU and not need to replace it or add a ton of upgrades and extra stuff later to make up for a poor choice you made based only on cost at the start of the project.  In all my years I don't recall ever hearing anyone say "Man!,  I wish I'd bought an ECU with less features!", its always people wishing they'd bought the more feature full unit when it was only another $100 or 2. Something always come up, something you forgot, or something not working how you'd hoped...something..... and its way cheaper to be prepared with an ECU that has extra I/O pins and features.  I get budget but "pennywise and pound foolish" comes to mind when I here talk of the lowest cost options.

In reply to Paul_VR6 (Forum Supporter) :

So then other question of adding pins to compensate for the extra board is not really a question, then.  

Which brings up a different question- are there WB systems out there that the MS can tell to turn on?  So there is communication between the two systems?  I'm very much out of the loop for aftermarket WB systems- last time I bought any was 1998 or so- and those ECM units are still being made.  

Sorry about the tangent.

My "fantasy" MS project could be influenced by which one of I can have some control over the WB module.

(Not that anyone cares- but my "fantasy" project is to EFI my SPICA Alfa, and add a catalyst.  Right now, the simple MicroSquirt fits my requirements, but it's still only in my mind.  Things can change)

In reply to alfadriver :

Most/all stand alone wideband controllers can output an analog 0-5v signal, this gets fed to the narrowband O2 input and the firmware is configured to interpret that correctly as lambda.

Some controllers, such as my AEM, can communicate via CAN, and MS3 can read this data.  (MS2 apparently only has enough computing power to broadcast, which I discovered AFTER I bought the AEM)  This eliminates a couple layers of digital to analog and back conversion and the errors that get picked up along the way.

Microsquirt is MS2 based, if I remember right.

I like having everything in one module, no separate devices to have to find homes for and run wiring to.  Part of the WRX plan involves using two of those nifty looking 70mm CAN-based gauges just so I don't have to run extra sending units or (gack!) mechanical capillary tubes.  Will be monitoring oil pressure with the MS3 for datalogging purposes so that is already going to the computer.  Basically the only additional gauge I will need is a fuel level gauge, which I really could probably set up for the computer to read and control, but at some point it's just doing something for the sake of doing it

In reply to Pete. (l33t FS) :

So I take it that the CAN based A/F meters are not really capable of taking instructions via CAN?  Having done a lot of work on when you can fire up a O2 sensor heater, I have a pretty good idea of the time needed to minimize the water droplets in the exhaust.  I guess one can use a MS output to control a relay to power the a/f meter.  But it would be cool to tell it to turn on.

Pete. (l33t FS) said:

I like having everything in one module, no separate devices to have to find homes for and run wiring to.  Part of the WRX plan involves using two of those nifty looking 70mm CAN-based gauges just so I don't have to run extra sending units or (gack!) mechanical capillary tubes.  Will be monitoring oil pressure with the MS3 for datalogging purposes so that is already going to the computer.  Basically the only additional gauge I will need is a fuel level gauge, which I really could probably set up for the computer to read and control, but at some point it's just doing something for the sake of doing it

As another option, I'm now running all my original analog dash gauges (CLT, Oil Pres, Oil Temp) using PWM outputs on the ECU and it seems to be working fine, but obviously eats ECU output unless you use a CAN expander and at that point CAN gauge are probably easier unless you want the dash to look stock which is what I was going for.

alfadriver said:

In reply to Pete. (l33t FS) :

So I take it that the CAN based A/F meters are not really capable of taking instructions via CAN?  Having done a lot of work on when you can fire up a O2 sensor heater, I have a pretty good idea of the time needed to minimize the water droplets in the exhaust.  I guess one can use a MS output to control a relay to power the a/f meter.  But it would be cool to tell it to turn on.

 

I think the most common setup is controller on when ignition is on?  That is how I've always done it and then the controller ramps the heater to bring the sensor up to temp slowly over about 30 seconds? Something like that but jsut follow the instructions that come with it and you should be fine.  Burning a GPIO out to turn it on when you want it on is certainly an option thought and is "communicating" via a digial signal right ?    All the controllers I've seen other than MoTec (I think, they are all in on CAN everything) have an 0-5V output you just connect to an analog input on the ECU, even the CAN capable unit also have the analog out, at least the ones I've seen.

alfadriver said:

My "fantasy" MS project could be influenced by which one of I can have some control over the WB module.

(Not that anyone cares- but my "fantasy" project is to EFI my SPICA Alfa, and add a catalyst.  Right now, the simple MicroSquirt fits my requirements, but it's still only in my mind.  Things can change)

just my opinion, so take it for what little its worth, waste spark is fine but I just  wouldn't bother doing a new install without sequential fuel....its just a lot easier to get the low power stuff sorted and smooth when you can say adjust the idle mixture in each cylinder.  So if MS is your must have ECU then a 2 or 3 is where I'd suggest looking.....but are also LOTS of other options and most if not all use better processors then the MS lineup.

I mentioned rusEFI...the 4cyl unit with 4 fuel, 4spark, DBW, CAN, and quite a lot of I/O  is $320 in a sealed case with sealed connectors.  All possible because it has a fast modern processor....it just seems way more bang for the buck.  It might be worth a look....

In reply to mke :

The problem with just turning them on with the ignition is that you risk breaking the heater.  Especially at lower temps.  We turn on our UEGO's almost immediately after crank, but at 20 F, there's a delay of about 20 seconds at the exhaust temp passes through the condensation temp.  And that's to avoid getting a drop of water hitting the sensor, which can easily break the heater.

Not sure how many recalls we've issued for broken sensors- both the front main one  and the catalyst monitor sensor.  

On the other hand, for instrumentation UEGO's, I've still not managed to break one.  So it could be worth the risk.

Personally, WRT the warm up, I'd want to find the fastest one I can find.  Waiting for 45 seconds to get it working sucks.  We are currently down to 8 seconds in production, and will be working down to ~5 seconds over the next few years.

Sequential fuel is practically a must when you have short intake runners.  I have tried batch fire on a BGN engine with its tiny length runners and it was bad.  Some cylinders got fuel, some spat it back up into the plenum instead, where the other cylinders thereby got too much.  Literally running on four cylinders until I switched over to sequential.

I note that every batch fire system I am aware of had very long runners between the plenum and injector location. Reversion won't push it back into the plenum.

In reply to alfadriver :

I think you are illustrating the difference between OE and aftermarket. For the aftermarket, having a fully self-contained WB controller that fires up with ignition works just fine 99.9% of the time. When you have an install base in the hundreds or at most low thousands, that's not a problem. When you add another couple of orders of magnitude and 100k mile warranty coverage, well, that 0.1% is significant.  So there is little or no call for a WB controller that takes instruction.

In reply to alfadriver :

With MS3 (not sure about MS2), you could set up a custom medium current output to power a relay that turns on your WB controller.  This could be based upon coolant/exhaust temp, and there may be a way to make it work from time since start.  I thought about doing this with the TR6, but just went with key on to better tune warmup.  We'll see how long the Bosch sensor lasts.