News

Carburettor syncronizing "tool".

Discussion in '365 GT4 2+2/400/412' started by hoverland, Dec 20, 2011.

  1. Steve Magnusson

    Steve Magnusson Two Time F1 World Champ
    Lifetime Rossa

    Jan 11, 2001
    20,313
    Texas
    Full Name:
    Steve Magnusson
    He's saying that the measuring tool is very sensitive (and adding some more damping to the output signal display might make determining the average value easier) -- not you...
     
  2. To remove this ad click here.

  3. Crowndog

    Crowndog F1 Veteran
    Rossa Subscribed

    Jul 16, 2011
    7,038
    Fairfield,Pa
    Full Name:
    Robert
    +1 I was indeed talking about the metering sensitivity, sorry I should have been a little clearer. Very cool device, kudos.
     
  4. vespasian

    vespasian Karting

    Nov 28, 2011
    55
    Germany
    Full Name:
    Alex
    Have to beg your pardon! Totally misunderstood your post! Quite embarrassed right now...
     
  5. Crowndog

    Crowndog F1 Veteran
    Rossa Subscribed

    Jul 16, 2011
    7,038
    Fairfield,Pa
    Full Name:
    Robert
    I understand how public forums can lead to misunderstanding. No harm. It was a very short terse response that I made and I was wrong, simple. I am just sitting here a little bored on a very slow day at work and just zipping through posts putting my $.02 in here and there. I will be a little more careful in the future.

    If I may suggest to OP, one simple way to slow down the response is to either sample less often (not as accurate) or average the samples taken over a given time period and display the results. Nyquist sampling theorem states that we need to sample at twice the highest frequency of the range of frequencys we want to look at. So one could use this as a guide to get the resolution you want without giving up accuracy. I would imagine then that 25msec averaging time would give good resolution and still get a smooth moving graph that responds smoothly while making adjustments. Just my $.02 again.
     
  6. hoverland

    hoverland Formula Junior

    Aug 14, 2010
    265
    Norway
    Full Name:
    Harald Ø
    Thanks for valuable feedback, I tried different values in sampling rate and averaging but need to look into averaging logic to get the "oil dampened motion" look right.
    A contributing factor may, as posted earlier be that PSI resolution changes with value, and I might have aimed too high -it may not be neccessary with 0,05PSI accuracy when diff between inlets are becoming very small. (The brilliant idea was to avoid need for fiddling with the computer/software while tuning).
    Will try a software version with fixed range 0-15PSI and see how it looks. It may not be neccessary with adjustable range at all.

    Still -even if there is potential for huge improvement -the tool as is, made it really easy (even for me) to finetune carbs and linkage a whole lot better than i managed with the STE Flowmeter.
    Now idle rpm hardly drops at all when gear selector set to drive (Yeah it's a auto ;)

    Harald
     
  7. To remove this ad click here.

  8. Crowndog

    Crowndog F1 Veteran
    Rossa Subscribed

    Jul 16, 2011
    7,038
    Fairfield,Pa
    Full Name:
    Robert
    #31 Crowndog, Feb 27, 2012
    Last edited: Feb 27, 2012
    This is way a way cool device, certainly could have used it thinking back on some multicarbed bikes and cars in the past. Best of luck.

    I was just thinking back to my old days programming in machine code. We would set up a register to keep the Average then add to it the next sample and divide by the number of samples so far (incrementing register). We would time the loop and then set a counter to run the loop long enough to meet our goal. In this case 25 msec. At the end of the count the register was set back to the next sample and run again. I don't know what language you are in but that was the principle unless my old memory is up to tricks again. We may have used a timer but back then we needed speed and this was faster.
     
  9. hoverland

    hoverland Formula Junior

    Aug 14, 2010
    265
    Norway
    Full Name:
    Harald Ø
    Thanks, basically this is the way it's written now :)
     
  10. hoverland

    hoverland Formula Junior

    Aug 14, 2010
    265
    Norway
    Full Name:
    Harald Ø
    #33 hoverland, Feb 28, 2012
    Last edited by a moderator: Sep 7, 2017
    My softwareguru have updated software, and I will make new test tonight.

    For fittings in inlets and testing of prototype I shortened some cheap std 6mm brake-bleed fittings. Actually when fitted -they dont look too massive even on my odd-colored valve covers, when covers returned to original black they will be even less obvious to spot.
    With the rubber caps -the testtubes/hookup are done in seconds :)

    Harald
    Image Unavailable, Please Login
    Image Unavailable, Please Login
     
  11. Crowndog

    Crowndog F1 Veteran
    Rossa Subscribed

    Jul 16, 2011
    7,038
    Fairfield,Pa
    Full Name:
    Robert
    Are your sensors able to measure positive pressures as well as negative?
     
  12. To remove this ad click here.

  13. hoverland

    hoverland Formula Junior

    Aug 14, 2010
    265
    Norway
    Full Name:
    Harald Ø
  14. Crowndog

    Crowndog F1 Veteran
    Rossa Subscribed

    Jul 16, 2011
    7,038
    Fairfield,Pa
    Full Name:
    Robert
    Looking over the data sheet that's a neat device. Depending on how crazy you want to get you might preselect devices to get matching one's who's outputs match within the range of pressures you are designing for. Or you might ask your supplier if they can obtain matched devices. Just a thought.
     
  15. hoverland

    hoverland Formula Junior

    Aug 14, 2010
    265
    Norway
    Full Name:
    Harald Ø
    #37 hoverland, Feb 28, 2012
    Last edited: Feb 28, 2012
    Gradually improving :)

    Thanks you all for positive feedback and sharing thoughts.

    Software now without the "autoranging" and that made a real difference.
    Not easy to see but scale is 0-15 PSI, so resolution is 1 PSI

    Video shows engine start, idle and then with some power applied -holding the car on the brakes.
    As you can see the timelines shows "history", and change between bars and lines are just a click.
    http://youtu.be/8E9frvrJNqw

    Even on this poor video, it's easy to see that the leftmost two blue bars representing carburettor cyl 1 & 2 needs correction :)
    This test was good, will have my softwareguru to come by and tweak the settings further.

    Anyway -there are some "wavelike" strange fluctuations on the video right after engine start, anyone have a explanation what this might be?
    They only appear after start, when engine is running steady no waves like that -any thoughts appreciated.

    Harald
    (Now planning to kill the Dell, and get the software into my iPhone)
     
  16. Crowndog

    Crowndog F1 Veteran
    Rossa Subscribed

    Jul 16, 2011
    7,038
    Fairfield,Pa
    Full Name:
    Robert
    #38 Crowndog, Feb 28, 2012
    Last edited: Feb 28, 2012
    That is looking really great. Just some thoughts. Are you pulling power from the car to run the board? As far as the artifacts, you may need to filter it better or you may be getting RFI from the car which shielding your board in the enclosure with proper precautions would eliminate. It appears that you have a working model there, congratulations.

    p.s. cylinder 1 and 2 need adjusting. LOL

    It is possible that the averaging software routine is sending results in the beginning that don't make sense untill the values settle down. See if putting a start-up delay before outputting data doesn't get rid of it.
     
  17. Ak Jim

    Ak Jim F1 Veteran
    Rossa Subscribed Owner

    Dec 23, 2007
    5,527
    North Pole AK
    I would really like to get one of these.
     
  18. Steve Magnusson

    Steve Magnusson Two Time F1 World Champ
    Lifetime Rossa

    Jan 11, 2001
    20,313
    Texas
    Full Name:
    Steve Magnusson
    #40 Steve Magnusson, Feb 28, 2012
    Last edited: Feb 28, 2012
    It's not as pronounced as at the start, but if you freeze the video at 0:54 it's got the same sort of structure where the three blue bars on the left side of the screen are down while the three blue bars on the right side are up, and the three green bars on the left side are up while the three green bars on the right side are down. Do these groups of "three" physically correspond with each carb, and does "green" represent one bank and "blue" the other bank?
     
  19. rkljr

    rkljr Formula Junior

    May 16, 2011
    720
    South of Boston, MA
    Full Name:
    Richard
    Cool stuff and very useful. Reading continuous data can be a little tricky.

    Sounds like a fun project, I may be interested in one of the boards.
     
  20. Crowndog

    Crowndog F1 Veteran
    Rossa Subscribed

    Jul 16, 2011
    7,038
    Fairfield,Pa
    Full Name:
    Robert
    #42 Crowndog, Feb 28, 2012
    Last edited by a moderator: Sep 7, 2017
    It appears that the cylinders are numbered starting left to right front row 1-6, then left to right back row 7-12. It may be that the screen refresh rate with the new data streaming in has that wavey artifact. The problem is the shutterspeed of the camera plus the shutter speed of youtube really changes what we see vs what is actually on the screen. It would be interesting to see if a slow speed analysis would reveal the firing order of the engine!
    Image Unavailable, Please Login
     
  21. Jonny Law

    Jonny Law F1 Rookie
    Silver Subscribed

    May 6, 2008
    2,925
    Over yonder.
    Full Name:
    IT Guy
    I know of a couple of people who would be interested in this. Keep us informed.
     
  22. hoverland

    hoverland Formula Junior

    Aug 14, 2010
    265
    Norway
    Full Name:
    Harald Ø
    #44 hoverland, Feb 29, 2012
    Last edited: Feb 29, 2012
    The picure of settings-pane is old. (at time picture were taken showing correct ign sequence, which is totally irrelevant in this case)

    So on actual videos the green bars represent cylinder 1-6 left to right, and the blue bars cylinder 12-7 left to right -this to get (geo)graphics right, -to visually show engine layout, making it easier to "see" where to aim your screwdriver.
    By rearranging in settings-pane you can get the right picture regardless of what side of car you are placed.
     
  23. hoverland

    hoverland Formula Junior

    Aug 14, 2010
    265
    Norway
    Full Name:
    Harald Ø
    Thanks for encouragement

    When prototype is OK, I will summarize changes needed and see if tool can be produced with a sensible pricetag.
    This project would not happen without you guys here in forum, the combined "knowledge-bank" and willingness to share is of outmost importance to people like me -who don't know it all. So guess this is my contribution so far.
    I've been to some shows and there are too many engines out there that in running pretty rough, and with the low power from our V12's they should run really smooth as they did at one time rolling out from Maranello. (70bhp pr liter was not very "racing" even in the seventies).

    At the moment this is a cool enjoyable hobby, both for me and the software-guy.
    I get to play and learn more about the beautiful engine, and have closed my eyes regarding cost so far.
    Need to open them soon :)
    Harald
     
  24. vespasian

    vespasian Karting

    Nov 28, 2011
    55
    Germany
    Full Name:
    Alex
    As the old experienced "carb guys" are getting really rare these days, this tool ensures some independance for carb addicted people like me. This tool - as stated before - is a "Must have".

    Did you already change carb settings accordingly to the results of the tool or is your engine still running "as is"?
     
  25. hoverland

    hoverland Formula Junior

    Aug 14, 2010
    265
    Norway
    Full Name:
    Harald Ø
    These two videos shows before and after 1st quicktuning.
    http://youtu.be/nNWEGW9O90g
    http://youtu.be/8E9frvrJNqw

    Still need to finetune as you can see.
    Working on ideas to enhance graphic to improve "visual accuracy" -it's a bit tricky to compare the two cylinderbanks with the current view and no common reference line.

    Harald
     
  26. vespasian

    vespasian Karting

    Nov 28, 2011
    55
    Germany
    Full Name:
    Alex
    +1

    It seems - with all the video latency etc. - that the engine runs way smoother after your 1st quicktuning. Would you confirm this?
     
  27. hoverland

    hoverland Formula Junior

    Aug 14, 2010
    265
    Norway
    Full Name:
    Harald Ø
    I may be biased and overenhusiastic if I say yes; -well, the engine was actually not too badly adjusted to start with, so most noticeable difference is throttle response from idle/lowspeed, and of course less drop in rpm when gear selected (auto).
    The throttle-response is very crisp/quick and it's strong/stable at idle.
     
  28. skl63

    skl63 Karting
    Silver Subscribed

    Dec 13, 2007
    212
    Melbourne, Australia
    Full Name:
    Scott
    Fantastic effort Harald,

    I am truly envious of people like you and your s/ware friend. To even think of a digital solution for these old cars is amazing... I think my brain goes very analog when I work on my Boxer (and I am in IT).

    Firstly, I hope the 365 Boxer has test ports (???), as I would be very interested in your tool should it ever be commercialised.

    And second, given the rumored new faster quad core iPad, what chance of porting your s/ware to the iPad and IOS? Surely a broader market and easier to deal with (although not sure how the hardware interfacing would go??)

    Anyway great work. I watch daily with interest your progress.,

    Cheers

    Scott
     

Share This Page