Forum made EFI?

Discussion in 'Aircraft Design / Aerodynamics / New Technology' started by blane.c, Dec 27, 2018.

Help Support HomeBuiltAirplanes Forum by donating:

  1. Dec 30, 2018 #41

    pictsidhe

    pictsidhe

    pictsidhe

    Well-Known Member

    Joined:
    Jul 15, 2014
    Messages:
    7,351
    Likes Received:
    2,118
    Location:
    North Carolina
    V twins are uneven fire. If you can use lost spark, which I'm dubious about, you could mask off 6 cylinders of a V8 ignition pattern. Otherwise, you need 2 channels, ehich may be a hardware issue.
     
  2. Dec 30, 2018 #42

    Hot Wings

    Hot Wings

    Hot Wings

    Well-Known Member HBA Supporter

    Joined:
    Nov 14, 2009
    Messages:
    6,639
    Likes Received:
    2,529
    Location:
    Rocky Mountains
    Are you remembering/thinking about the SU type carbs that added a stepper motor to the bottom to move the jet in response to the O2 signal? We - could - do the same with most industrial type carburetors. Instead of moving the jet we would have to move a needle in an enlarged jet.

    Kind of depends on how we define uneven? I haven't looked into the source code much yet but it appears that the Speeduino uses one timing register per cylinder, thus the need to use the Mega for anything more than one cylinder. (All of the other ATmega based Arduinos only have one 16 bit timing register?)

    We might be able to just slave the second cylinder off the first with a simple delay calculated from the RPM. At 6000rpm with a 36 tooth wheel one of the 8 bit registers might be enough to keep track of the rpm - provided it isn't a missing tooth type. If this were true then the same 16 bit register could be used for the second cylinder delay, with adjustments for software overhead. We could then use any Arduino board/chip.
     
  3. Dec 30, 2018 #43

    blane.c

    blane.c

    blane.c

    Well-Known Member HBA Supporter

    Joined:
    Jun 27, 2015
    Messages:
    3,807
    Likes Received:
    678
    Location:
    capital district NY
    Why not an electric gate, with crank triggering once each 360*, and 90* V engine firing 270*+450* the gate allows path to the 270* cylinder odd revs and to the 450* cylinder even revs?
     
  4. Dec 30, 2018 #44

    Hot Wings

    Hot Wings

    Hot Wings

    Well-Known Member HBA Supporter

    Joined:
    Nov 14, 2009
    Messages:
    6,639
    Likes Received:
    2,529
    Location:
    Rocky Mountains
    I'm not following what you are suggesting? This would be a whole lot easier, and faster, with a couple of beers and a napkin between us. :beer:
     
  5. Dec 30, 2018 #45

    blane.c

    blane.c

    blane.c

    Well-Known Member HBA Supporter

    Joined:
    Jun 27, 2015
    Messages:
    3,807
    Likes Received:
    678
    Location:
    capital district NY
    Maybe I should call it a shuttle. First trip past trigger signal goes #1 path and trips shuttle to second position, second trip past trigger and signal goes #2 path + time equaling 90* difference and trips shuttle back to first position.
     
  6. Dec 30, 2018 #46

    blane.c

    blane.c

    blane.c

    Well-Known Member HBA Supporter

    Joined:
    Jun 27, 2015
    Messages:
    3,807
    Likes Received:
    678
    Location:
    capital district NY
    Every other trigger signal would be slaved to the calculations of the first position plus the 90* difference. That is still 30 adjustments per second at 3600RPM.

    Corrected to 30
     
  7. Dec 30, 2018 #47

    blane.c

    blane.c

    blane.c

    Well-Known Member HBA Supporter

    Joined:
    Jun 27, 2015
    Messages:
    3,807
    Likes Received:
    678
    Location:
    capital district NY
    I love beer, and have flunked PC sensitivity repeatedly.
     
    BJC likes this.
  8. Dec 30, 2018 #48

    AJLiberatore

    AJLiberatore

    AJLiberatore

    Well-Known Member

    Joined:
    Nov 27, 2011
    Messages:
    108
    Likes Received:
    22
    Location:
    Canton MI
    I can't discuss it without permission, but such a system was in the works per a random mtg almost a yr ago. Not small engines, however think popular engines & their airframes in the Exp Category. I will possibly see the gent in early March, if it hasn't been dropped or they are about to unveil it, I will discuss it then.
     
  9. Dec 30, 2018 #49

    blane.c

    blane.c

    blane.c

    Well-Known Member HBA Supporter

    Joined:
    Jun 27, 2015
    Messages:
    3,807
    Likes Received:
    678
    Location:
    capital district NY
    Thinking about it some more, you could drop the data down two paths simultaneously at the first passing of the trigger and have one path blocked for execution. The blocked path would add 90* and be opened at the second passing of the trigger.
     
  10. Dec 30, 2018 #50

    pictsidhe

    pictsidhe

    pictsidhe

    Well-Known Member

    Joined:
    Jul 15, 2014
    Messages:
    7,351
    Likes Received:
    2,118
    Location:
    North Carolina
    I haven't come across the SUs I know Bosch was playing with pierburgs, though. I'm thinking of an adjustable air bleed to vary air and therefore fuel pressure in the float chamber. If it is closed loop with a pressure sensor, a simple pwm solenoid actuator would be fine. Could even have a simple hysteresis control if it the pressure changed slowly compared to the solenoid.
    Blane, you are thinking of flip flops.
     
  11. Dec 30, 2018 #51

    Hot Wings

    Hot Wings

    Hot Wings

    Well-Known Member HBA Supporter

    Joined:
    Nov 14, 2009
    Messages:
    6,639
    Likes Received:
    2,529
    Location:
    Rocky Mountains
    I think that is what I was meaning by 'slaving' the second (cloned) event from the first. It would just be done virtually, in software. Using 2 physical triggers would need a dual tooth wheel and 2 sensors? Need another napkin. I just spilled a bit of beer...............

    I have found one potential problem with the Speeduino software for aircraft use. Easy to modify but probably needs to be done. The way the rev limiter is implemented could cause EGT problems during long descents and damage once it comes out of the hard limit mode. A big ball of fire and popping out the exhaust might make for good theatrics on the street but I don't know if I'd want to fly the same way.
     
  12. Dec 30, 2018 #52

    Hot Wings

    Hot Wings

    Hot Wings

    Well-Known Member HBA Supporter

    Joined:
    Nov 14, 2009
    Messages:
    6,639
    Likes Received:
    2,529
    Location:
    Rocky Mountains
    Should be adequate for an aircraft at steady state cruise. May not even need the pressure sensor with a wide band O2? A servo, rather than a solenoid, might be easier electronically? Same PWM control. There seems to be quite a bit of resistance to closed loop due the presumed need to be able to use 100LL. I wonder how much of a real world problem that is for potential users? My local airport has no mogas. We have to haul our own but it's not that big a problem.
     
  13. Dec 30, 2018 #53

    rv6ejguy

    rv6ejguy

    rv6ejguy

    Well-Known Member

    Joined:
    Jun 26, 2012
    Messages:
    3,749
    Likes Received:
    2,828
    Location:
    Calgary, Alberta, Canada
    We see Widebands last between 2 and 300 hours on 100LL. Not sure why the variation.
     
  14. Dec 30, 2018 #54

    blane.c

    blane.c

    blane.c

    Well-Known Member HBA Supporter

    Joined:
    Jun 27, 2015
    Messages:
    3,807
    Likes Received:
    678
    Location:
    capital district NY
    I got this O2 info from the nano forum.

    APSX D1 Digital Wideband WBO Controller Module (Board)p://www.wide-band.com/product-p/apsx_d1diy.htm
     
  15. Dec 30, 2018 #55

    blane.c

    blane.c

    blane.c

    Well-Known Member HBA Supporter

    Joined:
    Jun 27, 2015
    Messages:
    3,807
    Likes Received:
    678
    Location:
    capital district NY
  16. Dec 30, 2018 #56

    blane.c

    blane.c

    blane.c

    Well-Known Member HBA Supporter

    Joined:
    Jun 27, 2015
    Messages:
    3,807
    Likes Received:
    678
    Location:
    capital district NY
    The way I understand this unless your EFI comes pre-tuned for your engine, you need a tuner. This APSX site is the least expensive I have seen yet.
     
  17. Dec 30, 2018 #57

    blane.c

    blane.c

    blane.c

    Well-Known Member HBA Supporter

    Joined:
    Jun 27, 2015
    Messages:
    3,807
    Likes Received:
    678
    Location:
    capital district NY
    The tiny holes are susceptible to being plugged with carbon deposits?
     
  18. Dec 30, 2018 #58

    Hot Wings

    Hot Wings

    Hot Wings

    Well-Known Member HBA Supporter

    Joined:
    Nov 14, 2009
    Messages:
    6,639
    Likes Received:
    2,529
    Location:
    Rocky Mountains
    Does the difference correlate with LSU 4.2 vs the LSU 4.9?
     
  19. Dec 30, 2018 #59

    Hot Wings

    Hot Wings

    Hot Wings

    Well-Known Member HBA Supporter

    Joined:
    Nov 14, 2009
    Messages:
    6,639
    Likes Received:
    2,529
    Location:
    Rocky Mountains
    The price has come down since I bought one from them. The only draw backs are that you need to build your own harness, it only works with the LSU 4.2, and - it's been a while - I think it requires open air calibration?

    Next time I'll use the Bosch chip. <$3 but you have to be able to solder SOIC.
     
  20. Dec 30, 2018 #60

    rv6ejguy

    rv6ejguy

    rv6ejguy

    Well-Known Member

    Joined:
    Jun 26, 2012
    Messages:
    3,749
    Likes Received:
    2,828
    Location:
    Calgary, Alberta, Canada
    Bosch has discussion papers on the 4.9 sensors saying that lifespan is reduced on leaded fuels substantially. Nothing to do with carbon. The TEL is conductive. We see average lives of around 30-50 hours on 100LL. At $65 a pop, you don't want to be replacing these to "save money" by installing a closed loop EFI system over a carb.

    We also see early failures due to thermal shock/ condensation concerns with aftermarket WB controllers.
     

Share This Page

Group Builder
arrow_white