# Industrial engine electronic management system development - HBA style

### Help Support HomeBuiltAirplanes.com:

#### PW_Plack

##### Well-Known Member
It might be possible to set up single throttle-body injection on an industrial V-Twin. Midwest Super Cub's 122-HP alcohol tractor pull engine, based on the Kohler 45 cubic-inch V-twin, was designed to run in a class which limited contestants to a single 1.125-inch-throat venturi-style carb. At the RPM we'd be talking about with a reduction drive, the intake manifold could likely be tuned to keep the mixture balanced at any RPM where significant power is being made.

Here's the dyno run video. Notice the EGTs are within 100º above 5500 RPM, where the power is being made in this case. Tuning of intake runner lengths would be required to bring this threshold down into our operating range, but not much if running 4000+ RPM and a redrive.

If the application is direct-drive, things could be a little tougher.

#### Hot Wings

##### Grumpy Cynic
HBA Supporter
Log Member
Okay, we're all done here. And it does ignition, too.
One day, a new record!!

Seriously, pros and cons of going with this ( or similar) embedded controllers vs Arduino, etc.
Looks like it could have been a contender. It's NLA. :depressed Will look for it's successor when I get some more time, if there is one.
Purpose built chips like this do have some real advantages but if we can't depend on a good supply....?

Edit:
Thinking about this some more - there isn't much difference in what this chip offered and what the OEMs are starting to supply.

Last edited:

##### Well-Known Member
Log Member
I am working on converting a Duromax 440 solution for Ultralight use. Would this EFI also work on a Single Cylinder?

#### jedi

##### Well-Known Member
Time out!
I like the project and am impressed with the expertise here. Sound like good work.
Although I have an ME degree and worked fuel systems and engines for more than one company that was many years ago and I am struggling to follow all the nomenclature and acronyms. Please keep that in mind as you post. I do not want to drag the group down to the lowest common denominator but fear that may be me. Any other lurkers out there trying to follow this? I can almost spell carburetor.

#### rmeyers

##### Active Member
If NLA means No Longer Available possibly there is a miscommunication. When I log into my NXP account it shows the reference design board as active and in stock.
Freescale used to be very good about supplying an availability roadmap. They would tell you exactly how long the chip/reference design would be available.

I also noticed that the gerbers for the reference design (board production CAD files) are also available for free download. That means that you can never be cut ff completely. It also means that design changes are relatively trivial.

#### Hot Wings

##### Grumpy Cynic
HBA Supporter
Log Member
They would tell you exactly how long the chip/reference design would be available.
.
Got my info from Mouser. They too list the develop0ment board - at a slightly higher price, But when you look up the individual part they have no stock and a not that it is scheduled for obsolescence.

Given the market target they may have decided that there just isn't enough to there and will leave it to be filled by others? If not, and they stay, the replacement will likely be a drop in.

Time out!
<< >>
Any other lurkers out there trying to follow this? I can almost spell carburetor.
I'll try to gather together some of this kind of information and find a few existing sites that might be able to describe the EFI hardware and process better than I can. I'm at the point where I'm comfortable but not good enough to teach - that is when you really learn.

We need members like you as much as the 'experts'. ASTM, for example, requires a certain ratio of producers and consumers with official voting status.

Edit:

I'm not cloud enabled other than a free version of DropBox. Open to ideas on how to put this kind of file where it can be updated as needed and viewed by those authorized so we don't clutter up HBA.

BTW I type carburetor correctly about 60% of the time..........some day my Word program is going to need a red pixel refill.

Last edited:

#### jbiplane

##### Well-Known Member
Got my info from Mouser
Dont stick too much to ECU board itself. You will encounter a lot of other problems like where to get good and cheap
- throttle bodies
- connectors
- fuel nozzles
- sensors...
Exist some variants, but MOQ=50 or more. One day you'll buy good part and next day garbidge from the same supplier.
I bought nice looking transparent fuel hose proudly marked "made in Germany" and it exploded by 3 bar fuel pressure.
Sensor calibration is other problem. We like VW sensors for stability, most of them dont need calibration at all.
In result ECU will cost ~30% of complete fuel injection kit and you will need calibrate all stuff together on each engine.

So I suggest to develop system which would be extremelly user friendly and usable. We waste 2 years on "ecotron"
and finally reworked Speeduino to much smaller size and energy consumption. I still not satisfied and want to have
metering pump instead of classical fuel pump+pressure regulator. I have somethere photos of 3w-international metering
pump internals.

The good reference could be bateryless EFI produced by chainsaw brands like Stihl and Husqvarna.

Last edited by a moderator:

#### pictsidhe

##### Well-Known Member
It's not the ECU that bothers me in an EFI system either. It's he other stuff. Weight, cost and reliability don't look easy to juggle for fuel pumps, regulators, filters etc etc.

If we can take a step back and see what we actually require from a system, we will likely find we don't a complex one.
Mixture accuracy of car EFIs is very accurate to meet emission requirements. For an engine with no cat, we can slacken these requirements a lot and still have trouble measuring a power or economy change. I'm away from my books right now, but I think something like +/- 5% would be fine for us. That will allow simpler and cruder fuelling implementations. We can use methods that are inadequate for meeting EPA requirements, but will still provide some advantage over carbs. I suspect that altitude compensation being the main one.

#### blane.c

##### Well-Known Member
HBA Supporter

I think that although this thread is aimed at small industrial four stroke engines it will have spillover for two strokes with little modification so it will be useful for many.

I wonder if the system in the end will be flexible enough to size it for use on the "Pegasus" ? That would be cool.

#### blane.c

##### Well-Known Member
HBA Supporter
Since it seems that the power hog is the fuel pump? Maybe find and settle on that item to build around? Also find an efficient alternator that works for the system?

#### Hot Wings

##### Grumpy Cynic
HBA Supporter
Log Member
I wonder if the system in the end will be flexible enough to size it for use on the "Pegasus" ? That would be cool.
I'm hoping that is where this will lead. These kind of opensource projects often fork off into related projects. Adapting OpenAero EFI (Anyone have other name proposals?) to 1/2 VW should be almost as simple as changing the timing in the software and mounting the sensors. From there to the O-100 it's not much more than a higher flow injectors.

Our Russian contributor is right about the ECU only being a small part of this project. Once the electronics have been chosen and the software developed it still all has to be mounted on the engine and tuned. Each engine will be a little bit different.

Since it seems that the power hog is the fuel pump? << >> Also find an efficient alternator that works for the system?
Power hog is relative. The Kohler EFI engine comes with either a 20 or 25 amp alternator. That means that 20 amps is enough to run a small EFI and have enough left over to run some other electrical load. There are some here that may be desiring a plane with no electrical system. With this in mind we should try to keep the system efficient enough that it can be run safely from a reasonably sized battery.

But a lot of these kind of specifics, while needing to be determined, probably should be left for later in the process?

#### Vigilant1

##### Well-Known Member
I'm hoping that is where this will lead. These kind of opensource projects often fork off into related projects. Adapting OpenAero EFI (Anyone have other name proposals?) to 1/2 VW should be almost as simple as changing the timing in the software and mounting the sensors. From there to the O-100 it's not much more than a higher flow injectors.
And Ross reports that, while he has sold some SDS systems to full VW owners, the folks flying VWs have generally been reluctant to spend the money for this fully engineered and supported solution. So, maybe some people using VWs would be interested, too, if the architecture can be stretched to deliver 7 GPH. The VW heads could really benefit from port injection--many induction systems are far from balanced, and being able to richen-up the mix on one or two problematic cylinders as a temporary measure to get CHTs down in a climb would be very welcome. There are quite a few full VWs flying out there--many more than 1/2VWs and industrial engines combined.

Power hog is relative. The Kohler EFI engine comes with either a 20 or 25 amp alternator. That means that 20 amps is enough to run a small EFI and have enough left over to run some other electrical load.
And the B&S EFI systems come on engines with alternators as small as 7 amps (no starter or engine battery), so there's that.

The B&S 810cc EFI engines have a list price about $250 more than the same engines with a carburetor. While list price reflects market forces (not costs), it does at least indicate that the EFI parts & installation probably isn't costing them much more than about$250 above the price of a carb.

But a lot of these kind of specifics, while needing to be determined, probably should be left for later in the process?
Yes. Still, a feasibility assessment might be needed at an early stage. If the parts (fuel pump, injectors, CPU/controller, sensors, new throttle body, mounts for injectors, cockpit controls and display, etc) are going to cost \$1000, then . . . .

If we can take a step back and see what we actually require from a system, we will likely find we don't a complex one.
Mixture accuracy of car EFIs is very accurate to meet emission requirements. For an engine with no cat, we can slacken these requirements a lot and still have trouble measuring a power or economy change. I'm away from my books right now, but I think something like +/- 5% would be fine for us. That will allow simpler and cruder fuelling implementations. We can use methods that are inadequate for meeting EPA requirements, but will still provide some advantage over carbs. I suspect that altitude compensation being the main one.
A significant source of inflight engine failure is carb ice. Yes, it is avoidable, but even some very competent pilots have been stung. So, while carbs seldom fail, it's worth including carb ice when we weigh reliabiliity of EFI vs a carb.
Maybe the small/simple EFI quest will branch into two efforts:
1) Bare bones, simple and cheap: Minimal sensors, no attempt to provide info to external systems, reliance on pilot actions to fine-tune mixture to each cylinder based on EGT. Maybe a small dedicated display in the cockpit for fuel flow, RPM, etc.
2) Highly automated, fuel efficient, fully integrated with other aircraft systems: Full sensor suite, full exchange with other systems (CANBus, etc).

#### Hot Wings

##### Grumpy Cynic
HBA Supporter
Log Member
<< >>

A significant source of inflight engine failure is carb ice. Yes, it is avoidable, but even some very competent pilots have been stung. So, while carbs seldom fail, it's worth including carb ice when we weigh reliabiliity of EFI vs a carb.
Maybe the small/simple EFI quest will branch into two efforts:
1) Bare bones, simple and cheap: Minimal sensors, no attempt to provide info to external systems, reliance on pilot actions to fine-tune mixture to each cylinder based on EGT. Maybe a small dedicated display in the cockpit for fuel flow, RPM, etc.
2) Highly automated, fuel efficient, fully integrated with other aircraft systems: Full sensor suite, full exchange with other systems (CANBus, etc).
There isn't, or shouldn't be, much difference between a bare bones EFI and one with all the extras. Our base hardware/board can be the same with plugin modules for things like wide-band O2 and K type thermocouples. One of the nice things about the Arduino architecture is that the various boards (shields) can just stack on top of each other, within a family group. We can copy that method if it works for us. The software can be modified as needed by commenting out lines in the code to deactivate modules. The development PDF that was linked to above uses this method for software development.

#### Vigilant1

##### Well-Known Member
Here's a question/observation possibly related to the project:

The Vanguard line has some EFI engines as well as some with carbs. In every case, the highest HP rating in a particular displacement is an EFI engine. But, there are several cases where several EFI engines are offered in the same displacement engine, and they have different HP ratings. They usually have the same compression ratio.
Here are three engines in the 810cc displacement. All are listed as having 8.4:1 CR:
Model 49R9: 26 HP carburetted
Model 49E7: 26 HP EFI
Model 49E8: 28 HP EFI

1) How is this accomplished in the EFI programming (same fuel to air ratio, but just put a cap on the GPM of fuel delivered for the lower HP variants?)
2) Why do this? My guess is just so they have an engine that is a readily available engine/drop-in replacement for any use an OEM might have, to include needing to gain price differentiation between different models.
3)Any ramifications for the present project? For example, if fuel-to-air ratios can be kept right and if the net amount of fuel is just capped to limit HP, it would seem easy to adjust that cap upward if we are programming our own EFI. Obviously, this would lower all the reliability/robustness margins that industrial engines offer and the user would be in uncharted territory. Sample applications:
A) If the engine has been installed for awhile and cooling has been unexpectedly good and the mechanicals seem to be holding up well, it would be possible to incrementally increase available HP upward with experience (maybe experience gained over thousands of hours).
B) If something equivalent to the "war emergency" mode of some WW-II engines were available on demand, it could be a useful optional feature. "Yes, I realize this will lower my engine's life, and that I may throw a rod immediately. I want the power now."

And, just to make things more confusing, B&S offers an 810cc in the "Commercial Turf" line (non-Vanguard), carburetted, 27 HP, and slightly higher CR (8.5:1)

#### blane.c

##### Well-Known Member
HBA Supporter
I wonder if all the different hp is made at the same RPM? If RPM is increased to raise HP and relatively the same twisting force is applied to the crankshaft is that a better approach within reasonable limits, or the RPM remains the same and a greater twisting force is applied to the crankshaft to gain hp? This is because I am concerned about these cranks and propellers to begin with and I am not sure what is a worse situation for them, more grunt coarser propeller or more centrifugal gyroscopic loading finer pitch propeller?

#### blane.c

##### Well-Known Member
HBA Supporter
I guess with two engines you could try both ways and see which crank separates first?

#### Vigilant1

##### Well-Known Member
I wonder if all the different hp is made at the same RPM?
Yes, the standard for quoting HP for these engines is 3600 RPM. If you'd like to see the HP curves, there's a link to a HP graph PDF in each of the tech pages I linked above. They typically look fairly flat at the end, like they wouldn't give you much more HP without going a lot higher in RPMs. That could be a function of induction limitations (to include EFI settings or carb restrictions) or something else.

#### blane.c

##### Well-Known Member
HBA Supporter
A de-enrichment to stoichiometric and water injection can improve hp substantially for a short period of time, water needs appropriate mixing with alcohol to prevent freezing depending on environment/altitude. You just flip a switch, the de-enrichment and water injection happen simultaneously. It will put a greater twisting moment on the crankshaft guaranteed.

#### blane.c

##### Well-Known Member
HBA Supporter
Yes, the standard for quoting HP for these engines is 3600 RPM. If you'd like to see the HP curves, there's a link to a HP graph PDF in each of the tech pages I linked above. They typically look fairly flat at the end, like they wouldn't give you much more HP without going a lot higher in RPMs. That could be a function of induction limitations (to include EFI settings or carb restrictions) or something else.
Probably camshaft.