Evinrude MegaSquirt Injection ECU

[BLisBack!]

Dynamic jet changing. In this vid you can see me changing the period with the hand held. Changing jets at the press of a button, very cool! You can mount the buttons anywhere but I found a piece of 1” pvc was the best as I could hold onto the pvc with the same hand as I steer with changing the period at WOT according to what the EGTs were saying. Works like a champ!

Most these high end system talked about on this thread can do this in some way, what makes EFI so awesome.

https://youtu.be/wg0wrq3_yOU

-BL

[PanRonnie]

When barometric pressure is 98Kpa your WOT manifold pressure is going to be 98Kpa depending on how restrictive your intake is
So when barometric is a 102Kpa again your WOT manifold pressure is going to be a 102Kpa or close to it
Count 1%of your total hp per 1Kpa and 1% for every 10Dgr celcius
If you tuned your alpaN system at 20dgr celsius at 98Kpa barometric you are 5% lean at 10Dgr celcius with a barometric pressure of 102Kpa
That why carb motors who do not have a barometric compensation run crap when you go up the mountain
And mercury and the rest have special carb jettings for lakes at altitudes
Your MAP manifold pressure is not going to stay the same when weather conditions change even humidity does a number hence i have weather station for my land and sea dyno
I can get a large deviation from my dyno run just by manipulating the weather data

[H2OPERF]

Im running baro/tps... I had MAP active originally when i first put the motor together, i tried using 1 cylinder then made a little manifold that had all the cylinders feeding the sensor but the MAP was still all over the place.. you could filter the crap out of it and get a semi good number but it kinda defeats the purpose plus it still wouldnt pull down that low at idle, now its straight baro tps. Emtron has 4 map slots with fuel /timing you can switch on the fly with a switch, rotary or pretty much any parameter you want to set up to make it jump , its really nice if you want to set up a cruising map thats slightly leaner or have a milder timing map for get home crap marina gas and obviously prop changes on the water. I also set up a 12 position rotary that pulls the map rich/lean for tuning or cruising, the possibilities are endless with modern EFI but it can go both ways.

[BLisBack!]

the possibilities are endless with modern EFI...

Absolutely.

I like that multiple, preset maps. Kinda like preset EQ settings on you car stereo. I can see that really being helpful. I might incorporate that into VSFI.

Maps take up a lot of RAM though, so you gotta pick and choose carefully. Its a true embedded system, not a desktop with tons of virtual memory. I have individual maps/cyl now but I'm thinking of changing that to a much simpler 15 cell rich lean map per cyl. Don't really need a full map.

Do you guys think there's a need for 100ths precision on the period? (i.e. if a cell in the map is displayed it would have 2 places after the decimal). From my testing tenths seems to be just fine. I know other systems go to 100ths though. With a small built in display like VSFI has, displaying hundreths crowds the map and makes it hard to read and the whole idea behind VSFI is to keep it simple.

-BL

[Chaz]

I did post the things that would be necessary as a replacement for a stock mercury unit with the exception of a IAT sensor and two cyl head temp sensors.
You seemed to ignore the necessary inputs. What is needed is a unit that uses the stock inputs ... but is adjustable in both individual fuel and spark tables. The higher the resolution, the better. The tuner can program cell blocks into zones, not be limited to a few small areas.

Any time you retard timing, you are running the heat out into the exhaust. Basing a theory on, no load @ high idle and a EGT gauge, will tend to have you see things bass - ackwards. The world of fuel flow meters and O2 sensors will open your eyes.

I have never met a carburetor that did not go rich beyond the torque peak. Bad for peak power, but on a two stroke it will tend to keep from sticking a piston. MS-3 has a WOT timer that you can program to add fuel after so many seconds , to avoid that situation. A map sensor is also good, if for nothing else than to add fuel when you chop the throttle @ high RPM . Since you don't seem to believe in the modern electric pump .. the pre-mix gas / oil gets chopped as well. 8000 rpm and idle quantity of fuel .. I'd rather the ECU make that adjustment, rather than me looking for a "up" button to start banging away on.

Once again, 90% of nitrous users bring their own fuel to the party. It's the ECU's job to retard timing while the music is on. Something else you have failed to recognize.

There are a few other entry level ECU's that are nothing more than a 1980's GM throttle body injection. It would however be really nice to see a waterproof ECU that had the needed components to run a 2-stroke engine efficiently, without dwelling on things that are just somewhat of a given.

The Fuel Tech ECU, dashboard and date logger are one in the same ..

https://www.fueltech.net/products/ft600-efi-system

[BLisBack!]

What is needed is a unit that uses the stock inputs

Early on I looked into using the stock sensors. It was a good idea no doubt. After characterizing (lab testing) them they were found so bad I didn't want to build a system around them. So I went with all digital sensors, not this garbage the OEMs were using just because they were cheap or the OEM sensors needed a bunch of circuitry to filter or turn it into a digital like signal. Nowadays, the newer motors are probably not that bad. I find at times I'm using the same connectors and sensors they are like the eTec for example.

two cyl head temp sensors.

Right now VSFI version 7 has 4 temp sensors Port, Starbrd, Ambient and Crankcase (intake). Version 8 will have 8 Mil Spec grade digital sensors along with the 10 EGTs. As testing has shown the EGT is a must have for each cyl.

The higher the resolution, the better. The tuner can program cell blocks into zones, not be limited to a few small areas

Agreed which is why the VSFI map can be either a full 30x64 (TPSxRPM in Alpha-N mode) or any simplified map down to one cell! Its called "breakpoints". I have found that a 30x64 is simply too much, you just don't need it. Especially for drag applications. Most these high end systems are built around a big map because the OEM system which it was pirated from used a big map to cover all use cases anywhere in the world. But if you want it, you can setup a 30x64 map, crazy resolution, very awkward.

no load @ high idle and a EGT gauge, will tend to have you see things bass - ackwards. The world of fuel flow meters and O2 sensors will open your eyes.

Not quite sure what you mean here but VSFI tries to keep it simple (more so than the other EFI units out there) and a flow meter is really not needed. Although I could add it in as a trim. In one of the youtube vids you can see a whole list of "trims" under the fuel tab for various edge cases like individual cyl correction, overall enrich, primer, cold start etc. I think most the top end EFI units do the same as H2O had talked about the Emtron unit. You can do a washdown trim but you can also do that in the upper right of your map (high rpm zero throttle state) on a simple Alpha-N system.

I'd rather the ECU make that adjustment, rather than me looking for a "up" button to start banging away on.

This is a good idea. Its called "closed loop". That can be put in at anytime, right now, full manual control, just keeping it simple. If you load up the system with too much of that stuff it confuses most users and hence the need to go to EFI University

It's the ECU's job to retard timing while the music is on

Yeah that's a no brainer. They all do that. There are a half dozen "open drain" aux inputs than can be used for almost anything on most these high end systems.

It would however be really nice to see a waterproof ECU that had the needed components to run a 2-stroke engine efficiently, without dwelling on things that are just somewhat of a given.

Problem is you can't seal a 2000w digital CDI unit that operates off a battery not a power robbing megneto coil, and is also stackable to double or quadruple the power in supercharged applications. Thats the CDI that VSFI uses, and due to the power it needs a fan or two! I guess I could scale a version down for only 250v primary and a light duty cycle and seal it. Maybe in future versions. For now keeping it simple to cover all applications. OEMs need waterproof, most DIY guys just the boxes somewhere dry and it keeps the price down.

Good questions! Keep 'em coming

-BL

[PanRonnie]

I am a big fan of things modular
This way you can pick and choose what you want
It is also easier to upgrade parts or have a way to repair them rather then sending in the whole unit
I was also playing with DC-CDI build 2 units but if i remember correctly it took 2 to 3 milliseconds to charge the capacitor so you would need 6 at which point i lost interest because the cost were getting out of hand even when i did my own winding
Keeping things simple and affordable would be a big plus using as much original parts as possible

[BLisBack!]

I am a big fan of things modular
This way you can pick and choose what you want
It is also easier to upgrade parts or have a way to repair them rather then sending in the whole unit
I was also playing with DC-CDI build 2 units but if i remember correctly it took 2 to 3 milliseconds to charge the capacitor so you would need 6 at which point i lost interest because the cost were getting out of hand even when i did my own winding
Keeping things simple and affordable would be a big plus using as much original parts as possible

Transformer is too small. Mine are huge but very fast. Its a tradeoff. I found boost mode was better than flyback/pulse too. But that depends on the transformer style too.

Here is a dual power supply CDI unit that was in the field for 10yrs; It has the biggest transformers I use. I also use a smaller one by about 30% which I usually build my CDIs with.



Here the smaller one is making 600v primary in just over 100us, 400v (which is as high as most coils can take) is about 60us.



-BL

[BLisBack!]

Also, bigger transformers mean more current, longer spark. But there is a limit. Ignition coils can be overloaded and become saturated and do nothing. I talk about this on the website.

If you up the primary voltage to the transformer you can get away with a much smaller transformer. This is the way the OEM systems work as they work with the charge coil voltage which is much higher than 12v. Very similar to the way new inverter welders work (just bought one so I know ) they use a much smaller transformer because they switch the input voltage into the main transformer to a higher freq and voltage.

-BL

[Chaz]

Actually I made statements, not asked questions. I spoke on what I was looking for in a system. It has taken you two or three shots now to say one thing I was stating as necessary .. you claimed it as a given .. and called it something else.

Good luck playing with your little toy .. as it appears that your not really trying to communicate with a customer base and develop something that fits their needs.
However, I give ya credit, you do have the nerd part .. down pat ..

Chaz = no longer loosely following ..

[BLisBack!]

You guys will like this one, 'cept maybe Chaz ;

VSFI testing of the eTec algorithm and dual channel CDI power supply. Tested with High Voltage active (sparks sent through actual spark plugs in the background). Its meeting timing with absolutely no jitter, timing drift and not loosing sync, injectors are at 100% duty so no pulse, and the rev limiter is set to 10,500rpm and you can see it cut in/out on the timing light. "Meeting timing" means the ECU is tracking the sensors on the engine simulator and all coils and injectors are firing when they should be, nothing is out of sync. This is a MUST DO test to ensure the software and power supply are keeping up with the real time requirements. Like Ronnie posted earlier the power supply has to be very fast. Pulling about 40amps at that rpm, about 400v to the coils, battery charger is at max in the background

At 10,500 thats 175 revolutions a second, each rev is just over 5ms and each cyl only having 95us to do 16 degrees of timing delay and fire the cyl.

https://youtube.com/shorts/XzzxyeI2R04?feature=share

-BL

[H2OPERF]

.016 X rpm X cyls = Hz... the number starts rising above the capability of many aftermarket units pushing you into multiple boxes and/or not using all the channels.

[BLisBack!]

.016 X rpm X cyls = Hz... the number starts rising above the capability of many aftermarket units pushing you into multiple boxes and/or not using all the channels.

Exactly. Then on top of that there is the thermal considerations, power dissipation to worry about. As rpm goes up, power consumption and heat goes up in a huge way. You really have to watch you don’t exceed duty cycle.

I remember the first time I got assembled boards for that CDI power supply 10yrs ago. I knew it could handle 10k rpm so I slapped it on the simulator at 8k and walked away. 20 min later I came back to a horrible smell of electronics burning. Those transformers had no air flow over them and melted clean through the board! Seriously exceeded the duty cycle for zero air flow. This is why OEM flywheels are vented so much.

-BL

[H2OPERF]

Water cooled would be the best for marine use, make back plates with passages to draw off the heat. All my stuff is on the ph and no issues so far.. The efi technology cdi suff is amazingly small and seems to handle the job they have internal temp sensors you can use for logging or indication. Found the ones I have used new prices are crazy but their stuff is top shelf. There are 2 boxes one above each head.

[BLisBack!]

Awesome. I like it!

-BL