Quantcast

Industrial engine electronic management system development - HBA style

HomeBuiltAirplanes.com

Help Support HomeBuiltAirplanes.com:

Vigilant1

Well-Known Member
Lifetime Supporter
Joined
Jan 24, 2011
Messages
5,754
Location
US
I guess with two engines you could try both ways and see which crank separates first?
The Micromaster or a similar project could make a pretty good testbed for these engines. Also, maybe going different routes with each engine will gain some insurance against having a failure mode common to both engines (e.g. if we had one engine with a carb and one with EFI, we'd have some "free" insurance against total loss of thrust due to total electrical system failure and from carb ice, too). OTOH, this can be taken to extremes and we wind up with a complicated mess.
 
Last edited:

blane.c

Well-Known Member
HBA Supporter
Joined
Jun 27, 2015
Messages
4,492
Location
capital district NY
It is a coin flip? Leave one relatively stock experiment with the other, experiment with both differently, ouch. Tuff decision, it would be easier to experiment with 1 engine and have two relatively stock for spares.
 

PW_Plack

Well-Known Member
Joined
May 25, 2015
Messages
154
Location
West Valley City, UT
I don't know about Briggs, but I've been looking at the Honda GX690 for a while, and got the shop manual. They make three different versions of this and call them GX630, GX660 and GX690, but they all have the same bore & stroke, same CR (9.3:1). The camshafts and carbs are the same part numbers, although they have different jetting; this may indicate some models are only available with more restrictive air cleaners. That doesn't appear to be the case with the two Briggs 810cc EFI engines.

I wondered whether it might be something as simple as lot-sorting the head castings. Some of the castings in these industrial engines are not pretty, and it might be possible to see a + 2HP difference just from how clean the ports are. If this was the case, I'm not sure they'd admit it!

It may also be that the different models are simply there to more closely match the ratings specified in equipment they're re-powering. So if an OEM has a design that's been vetted using a 26 HP engine, the 28 HP engine gets a nod and a wink, and a new rating, to save added paperwork.
 

pictsidhe

Well-Known Member
Joined
Jul 15, 2014
Messages
8,812
Location
North Carolina
The vanguard's have two different cams, 3 different heads, 3 capacities to vary the hp. I've not looked at carbs and exhausts, but I'd put money on different exhausts altering power too.
Briggs got themselves in some legal trouble some years back for altering the power of an engine range with just a sticker change, so they don't do that anymore...
 

rv6ejguy

Well-Known Member
Joined
Jun 26, 2012
Messages
4,078
Location
Calgary, Alberta, Canada
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.
Running at stoich with or without water injection will make less power and you don't need the added weight and complexity of this on a simple engine. Turbocharged engines running high boost can benefit greatly, atmo engines, no, unless you have really high CRs and not so high octane fuel.

power1.jpg
 

blane.c

Well-Known Member
HBA Supporter
Joined
Jun 27, 2015
Messages
4,492
Location
capital district NY
I am interested in more information about this NXP system and Mouser, has anyone committed to experimenting with this one?

The other thing I am wondering is can we get a complete system from Jbiplane for $200 because it is beginning to sound pretty economical? The only thing is sending engine to Russia for tuning is impractical unless there is magic shipping I am unaware of so having a tuning system to tune the system is important?

In fact because of the plethora of stubbornness and independence on this forum there is no standard to tune to so a inexpensive universal way of tuning to individual needs/wants is going to be key to any kind of success.
 

Hot Wings

Grumpy Cynic
HBA Supporter
Log Member
Joined
Nov 14, 2009
Messages
7,489
Location
Rocky Mountains
User guide for the the reference board design. Shows pretty much exactly how development works on these type boards.

https://www.nxp.com/docs/en/user-guide/KT912_P812ECUUG.pdf
Found this chip - which is still available. Long lead time but reasonably priced. The guide should still be a good reference?

https://www.nxp.com/support/developer-resources/evaluation-and-development-boards/analog-toolbox/evaluation-kit-33814-2-cylinder-small-engine-control:KIT33814AEEVBE

No idle speed control built in?

Has some nice built in features, all of which can be duplicated in a generic MCU with software except the VR input? One thing lacking for aircraft use is provision for dual ignition, but that could be supplied by a completely separate system. I don't see much advantage to using this compared to what is offered by the OEM. In fact it might be a reasonable bet that this is what they are using?

Fiddled around with my STM32 boards for far too long today. I just can't get a boot loader that works through the USB port to make it user friendly. My clone boards may be the problem...... By using the FTDI adapter the Arduino IDE will upload C++ sketches. Without being able to load via the USB will make it harder to do field updates.

A quick audit of the STM32 pins available indicates that there are enough to do a 4 cylinder engine with dual ignition and stepper idle speed control Along with 4 channel CHT and 2 channel EGT if using external O2 and VR hardware.
 

rv6ejguy

Well-Known Member
Joined
Jun 26, 2012
Messages
4,078
Location
Calgary, Alberta, Canada
Any decent EFI system can be tuned with a wideband on unleaded fuel with a prop attached, mounted to a test stand. Tune to 12.5 AFR for best power. 5 minutes to tune an SDS at WOT using the PC data logging.

https://www.youtube.com/watch?v=FS4jvLlXZhE One run up to WOT, review the AFRs on the PC. Change the RPM fuel values at each 100 rpm break point. Done.

Works the same on a 400hp supercharged 540: https://www.youtube.com/watch?v=jHp2poqf_Bs
 

Hot Wings

Grumpy Cynic
HBA Supporter
Log Member
Joined
Nov 14, 2009
Messages
7,489
Location
Rocky Mountains

Vigilant1

Well-Known Member
Lifetime Supporter
Joined
Jan 24, 2011
Messages
5,754
Location
US
Any decent EFI system can be tuned with a wideband on unleaded fuel with a prop attached, mounted to a test stand. Tune to 12.5 AFR for best power. 5 minutes to tune an SDS at WOT using the PC data logging.
If Speeduino can interface with this software we should be able to do the same. Good well sorted documentation is going to be worth as much as the hardware and software.
Wouldn't use of TunerStudio introduce another "uncontrollable external dependency" for this EFI project (i.e. another third-party we hope remains afloat and supporting their software, because we depend on it)? It seems like a lot of eye-candy for what we need, and maybe an extra layer of of complexity.
 

rv6ejguy

Well-Known Member
Joined
Jun 26, 2012
Messages
4,078
Location
Calgary, Alberta, Canada
Wouldn't use of TunerStudio introduce another "uncontrollable external dependency" for this EFI project (i.e. another third-party we hope remains afloat and supporting their software, because we depend on it)? It seems like a lot of eye-candy for what we need, and maybe an extra layer of of complexity.
I wouldn't depend on 3rd party tuning software. Since you're writing code anyway, make your own interface so you are not at the whim of someone else and any changes they make to complicate your life. I agree, you don't need this stuff for the intended target of your system.
 

rmeyers

Active Member
Joined
Dec 20, 2013
Messages
32
Location
Denver, CO
The MC33814 is an engine control chip, not a microprocessor. For those that aren't in to this kind of stuff it means that the MC33814 can't run software, it's basically a driver chip.

Warning-techno babble to follow
It will need to be paired with an MCU through SPI and what looks like a minimum of 11 GPIO lines. It does have some cool features. It appears to have a built in power supply/vreg and can supply 5v to the MCU with VBAT input. It also can supply highly accurate and protected 5v for sensors. The only sensor it has is O2. If you want to use the 5V supply for sensors you will, or at least should, use it to supply the MCU with it's input power so the grounds match. Otherwise you may end up chasing flaky sensors forever. It probably communicates the O2 status to the MCU through the SPI line. Injection and timing events must be supplied by the MCU.
techno babble off

Overall it looks like a nice driver for small engines. However it does mean having a board made.
 

Hot Wings

Grumpy Cynic
HBA Supporter
Log Member
Joined
Nov 14, 2009
Messages
7,489
Location
Rocky Mountains
Wouldn't use of TunerStudio introduce another "uncontrollable external dependency" for this EFI project (i.e. another third-party we hope remains afloat and supporting their software, because we depend on it)? It seems like a lot of eye-candy for what we need, and maybe an extra layer of of complexity.
Other than making our software is compatible don't think of it as a required item. It's more of a convenience than anything else. As RV6ejguy noted the map can be generated using an external data gathering system, the data cleaned up and formatted in something like Excel and the resulting map file hand loaded.

The MC33814 is an engine control chip, not a microprocessor.
Missed that little detail. "MCU Interface: 16-bit SPI and parallel interface with 5.0 V IO capability" :emb: I shouldn't make assumptions based on a quick scan. Now I know why I didn't notice any clock speeds.......
 

Vigilant1

Well-Known Member
Lifetime Supporter
Joined
Jan 24, 2011
Messages
5,754
Location
US
Looking at the options so far, it seems to me that the Speeduino project/firmware may be the most attractive foundation for this OpenAero EFI idea:
1) The hardware needed (Arduino) is going to be available forever (or as close as anything in electronics can get...like the immortal 555 chip). It is hugely popular and made by several companies. And it is cheap (the board: $15-$30)
2) The Speeduino program seems healthy. There's a fair amount of activity on their forum (not nearly as much as HBA, but still there's a lot of info being exchanged), and the development team is active. As the OpenAeroEFI program advances, it will be >very< handy to be able to ask Arduino-as-EFI questions to folks who have already climbed the learning curve. Speeduino has options to support "odd-fire" engines like these industrial V-twins, so I'd bet there are folks on that board who are already running their industrial engines with a Speeduino setup. Their knowledge could be very useful, especially in the early stages.
3) The Speeduino program is mature. They've already identified and addressed the myriad of issues that are going to crop up with hardware/software compatibility, driver compatibility with sensors and injectors, etc, etc. Why go through that again? These are the kinds of niggly, frustrating problems that can significantly derail a project like OpenAero EFI that has only very limited resources for doing development.
4) Speeduino can also control ignition. I don't see that as a giant advantage (the "magnetron" poor-man's magneto on these industrial engines is very reliable) and I probably wouldn't make use of it. But there may be some folks who want the hotter spark and more precise timing that an EI can provide and will want to drill for a second plug or even use the EI as their sole ignition.
5) The Speeduino code is public and vieable. There's no mystery about what is going on inside the box.
6) While the Speeduino effort/community will be very useful to OpenAeroEFI, even if it goes belly-up it won't doom the effort here. I imagine we'd just continue using the last, best version of the Speeduino executable file and soldier on. Maybe the same holds true for TunerStudio.

One thing I'm not sure about is the ability of the Speeduino firmware to run open loop with maps. It runs open loop at idle and it has an open loop boost mode. Also, folks are using Speeduino for 2-strokes (no O2 sensor), and they are using other sensors to make it operate, this is an area in which a few users are continuing development. But running without the O2 sensor has not been a high priority of the Speeduino community--it might be the kind of niche functionality that an OpenAeroEFI effort could dig into. Here's a page on GitHub with info on how people can submit a proposed change in the Speeduino code. I suppose if we went crazy developing a dedicated open-loop option set for Speeduino, this is where that could be submitted for inclusion in the larger Speeduino effort. As an example of the process, this page shows some of the back-and-forth from a user who was developing sequential EFI.
This does point to two possible negatives of Speeduino:
a) We would be dependent on others to approve our code and include it in future Speeduino updates. If their interests diverge from ours (now or in the future), this could be a problem. OpenAero might put a high priority on things that the Speeduino dev team just doesn't want to be bothered with (running open loop all the time, altitude compensation across a wide range, very robust modes to keep running when sensors fail, etc) We wouldn't "own" or control what happens to our Speeduino changes in the future versions of Speeduino. On a positive note, another set of eyes during the review process might catch problems we overlooked.
b) Impact of future Speeduino changes. It is possible that a new bit of code, totally unrelated to anything we care about (e.g. sequential EFI, turning on an electric fan, etc) and approved by the Speeduino development team will make the software less stable, incompatible with OpenAero EFI written subroutines, etc. Yes, it shouldn't happen, but . . .

Semi-related note Very unsophisticated open-loop EFI: While cruising around the Speeduino forum I came across a posting regarding a very simple, crude "redneck" EFI shield for Arduino developed by a guy named Joe Harkness. This has nothing to do with Speeduino. Total cost for everything except the injector (it even includes its own fuel pump) was about $130. It is compatible with a GM injector. It was designed for use with riding mowers, runs open loop, and also drives a small LCD display (included in that $130 price). It takes no sensor inputs from the engine, so almost any calamity forward of the firewall that left the power and the injector intact would allow you to stay in the air. In lieu of a TPS it has a user-adjustable potentiometer, and it has a mixture control knob. It might make a pretty good backup to any more sophisticated, more integrated OpenAeroEFI system. For max redundancy, give it its own injector (throttle body) and a "can't fail" source of 12VDC and you'd be in business. Here's a link to the post, but I think you need to join the Speeduino forum to get there.
Edited to add: Here's a link to the original post (now archived) the developer made on MyTractorForum.com in 2010. A subsequent post provides the text of his very simple code to run the EFI and the display. It explains how he uses it on his mower (start and run, etc). It seems likely that with an EGT gauge you could probably do a good job of getting it to run well at almost any altitude. And, with it's own Arduino to use, we could get as fancy as we want (mixture maps, incorporate a MAP sensor input to reduce need for manual mixture adjustments, etc). But getting too cute might not be a great idea, especially if this is to be a backup: as-is it could keep the engine supplied with fuel and providing power even if every sensor, injector and related component of the primary EFI failed. Making it "better" or more polished by adding sensors and more complex code will make it less reliable. EFI Shield.jpg
 
Last edited:

blane.c

Well-Known Member
HBA Supporter
Joined
Jun 27, 2015
Messages
4,492
Location
capital district NY
From above;

""b) Impact of future Speeduino changes. It is possible that a new bit of code, totally unrelated to anything we care about (e.g. sequential EFI, turning on an electric fan, etc) and approved by the Speeduino development team will make the software less stable, incompatible with OpenAero EFI written subroutines, etc. Yes, it shouldn't happen, but . . .""

What we need is a Forum "Guinea Pig" to try out any improvements. Maybe someone with an industrial engine multi? Then they could try the improvement one engine at a time.
 

blane.c

Well-Known Member
HBA Supporter
Joined
Jun 27, 2015
Messages
4,492
Location
capital district NY
Semi-related note Very unsophisticated open-loop EFI: While cruising around the Speeduino forum I came across a posting regarding a very simple, crude "redneck" EFI shield for Arduino developed by a guy named Joe Harkness. This has nothing to do with Speeduino. Total cost for everything except the injector (it even includes its own fuel pump) was about $130. It is compatible with a GM injector. It was designed for use with riding mowers, runs open loop, and also drives a small LCD display (included in that $130 price). It takes no sensor inputs from the engine, so almost any calamity forward of the firewall that left the power and the injector intact would allow you to stay in the air. In lieu of a TPS it has a user-adjustable potentiometer, and it has a mixture control knob. It might make a pretty good backup to any more sophisticated, more integrated OpenAeroEFI system. For max redundancy, give it its own injector (throttle body) and a "can't fail" source of 12VDC and you'd be in business. Here's a link to the post, but I think you need to join the Speeduino forum to get there.
Edited to add: Here's a link to the original post (now archived) the developer made on MyTractorForum.com in 2010. A subsequent post provides the text of his very simple code to run the EFI and the display. It explains how he uses it on his mower (start and run, etc). It seems likely that with an EGT gauge you could probably do a good job of getting it to run well at almost any altitude. And, with it's own Arduino to use, we could get as fancy as we want (mixture maps, incorporate a MAP sensor input to reduce need for manual mixture adjustments, etc). But getting too cute might not be a great idea, especially if this is to be a backup: as-is it could keep the engine supplied with fuel and providing power even if every sensor, injector and related component of the primary EFI failed. Making it "better" or more polished by adding sensors and more complex code will make it less reliable. View attachment 77196
Couldn't one throttle body have two injectors one for normal and one for crude?
 

Vigilant1

Well-Known Member
Lifetime Supporter
Joined
Jan 24, 2011
Messages
5,754
Location
US
Couldn't one throttle body have two injectors one for normal and one for crude?
Yes, that wouldn't be a problem. But, from what little I've seen of these V-Twins, issues with the induction systems are very common, likely due to the way they fire (bang... pause(90 deg crank rotation)..bang..pause...pause...pause...pause (630 deg total crank rotation)....bang). Fuel atomization/droplet size, etc can lead to one cylinder being over lean and the other being over rich when there is a big difference in the time the induction runner is idle before the cylinder gets its gulp of mixture. Most engines are even-firing and this isn't an issue. So, for the main EFI you'd want individual injectors for each cylinder and a way to tweak them individually to even things out. For the backup, maybe a single injector in the throttle body would do...you don't care about optimizing fuel efficiency, you just need the engine to keep making reliable power.

The guy who made the Redneck EFI noted that there's are enough extra digital outputs on the Arduino board to drive a second injector.
 
Last edited:
Top