Page 13 of 55 FirstFirst ... 38910111213141516171823 ... LastLast
Results 181 to 195 of 825

Thread: DIY LTCC or similar system for LT1s

  1. #181
    Fuel Injected!
    Join Date
    Oct 2013
    Posts
    1,022
    spfautsch

    At this point, I definitely would bet against switching to the next coil on each low-res crank signal working correctly. I too highly doubt that the PCM could command 46* advance at any rpm and still restrain the coil charging time to be between two low-res signals. I'd bet that the PCM turns the coil back on to fire the next cylinder coil before the low-res pulse identifying the cylinder that just fired even appears.

    This doesn't make it that much more difficult to use 8 coils. Do it by using the opti signals to determine which coil to connect next and the EST to switch to that coil. In that way, you connect the next coil required immediately after the previous coil has fired. This doesn't give the ideal dwell time, but would work. However, this eliminates the main advantages of using 8 coils. A proper 8 coil system will overlap the operation of the 8 coils to maintain the ideal dwell time for the coils at every rpm. That's too big an advantage to give up.

    I too would bet against the LTCC module simply multiplexing EST line. It is advertised as having more spark energy which likely means it does maintain the ideal dwell time for each coil. This means it also does some much more complicated calculations to be able to switch 2 coils on at once. It doesn't need to do any timing calculations, but rather just use the EST and rpm along with the desired dwell time of the coil to calculate the time when the next coil should turn on. The EST can still be used to trigger the correct coil firing. In this way, even if there was some error in the dwell timing calculations as the engine operating conditions vary, it wouldn't affect the actual engine timing. Or in other words, the calculations don't have to be as exact as doing timing would require for it to still work just fine.

    I recall posting something about having a full plan before going off half-cocked on a solution fairly early in this thread. It seems to have taken another 100+ posts to start coming up with a plan that could work but I think there is a decent understanding of what is required to properly do a coil near plug system now. It's much more complicated then just multiplexing the coil driving wire.
    Last edited by lionelhutz; 11-11-2017 at 07:22 PM.

  2. #182
    Fuel Injected! spfautsch's Avatar
    Join Date
    Apr 2015
    Location
    Montgomery City, MO
    Age
    52
    Posts
    883
    Quote Originally Posted by lionelhutz View Post
    I'd bet that the PCM turns the coil back on to fire the next cylinder coil before the low-res pulse identifying the cylinder that just fired even appears.
    I would imagine it depends heavily on RPM and commanded spark advance. My assumption is that beyond some pre-determined RPM (for the sake of discussion let's pretend it's 5000) it simply switches the ICM back on immediately after it switched off as a "best effort" type methodology. But without someone dumping the TPU ROM or a bunch more waveform captures it's all conjecture.

    Quote Originally Posted by lionelhutz View Post
    ... This doesn't give the ideal dwell time, but would work. However, this eliminates the main advantages of using 8 coils. A proper 8 coil system will overlap the operation of the 8 coils to maintain the ideal dwell time for the coils at every rpm. That's too big an advantage to give up.
    This is my sticking point. If it's worth doing it's worth doing correctly. To use your metaphor, there's no point building a half cocked setup here. By my reckoning there are two types of LT-1 enthusiasts - the folks who want to preserve the stock setup as cheaply as possible and the modders who want to squeeze every ounce of power out. If you're the former go buy an aftermarket opti, a new water pump, a new optical sensor, suck it up and drive for another 200,000 miles. It's a damned reliable ignition system and will make great power well into the 6500 RPM neighborhood, until your water pump s**ts the bed.

    Quote Originally Posted by lionelhutz View Post
    I too would bet against the LTCC module simply multiplexing EST line. ... It doesn't need to do any timing calculations, but rather just use the EST and rpm along with the desired dwell time of the coil to calculate the time when the next coil should turn on.
    I doubt it calculates advance but simply keeps a "running" variable derived from comparing when the EST line switches off to high res pulses since low res falling edge. Maybe it's more complicated than that, but I doubt it. But ultimately, to know exactly when to start dwell the controller must know about commanded spark advance. Else it's just a "close enough" thing, which gets us back into the realm of "half-cocked".

    Quote Originally Posted by lionelhutz View Post
    The EST can still be used to trigger the correct coil firing.
    It must, else the controller would need to duplicate a bunch of spark advance functionality that the PCM handles just fine. To do that it would need to know about MAP, TPS, ECT, knock, (need I go on?).

    Quote Originally Posted by lionelhutz View Post
    It seems to have taken another 100+ posts to start coming up with a plan that could work but I think there is a decent understanding of what is required to properly do a coil near plug system now. It's much more complicated then just multiplexing the coil driving wire.
    Yes it is.

    I think I have a picture in my mind of how to do this on a single arduino "core". Hopefully I'll have a chance to build out a sketch that will demonstrate the concept before one of my 7 other projects get in the way.

  3. #183
    Fuel Injected! vilefly's Avatar
    Join Date
    Sep 2017
    Age
    53
    Posts
    217
    Well, spfautsch, ya got me. Mostly stock preservationist here. I already have my spare distributor and such standing by. But I would like to give myself alternatives as well as get back on the electronics design horse. This project accomplishes both. Not to mention someone is bound to build on that work, and come up with something better. That's just how it goes.

    Perhaps next week I can get those waveforms you need, as well as limp mode behavior when the Hi-res signal is lost. I hope my workload lightens a bit so I can do it. I am still dedicated towards collecting all data possible.

  4. #184
    Fuel Injected! spfautsch's Avatar
    Join Date
    Apr 2015
    Location
    Montgomery City, MO
    Age
    52
    Posts
    883
    Quote Originally Posted by vilefly View Post
    But I would like to give myself alternatives as well as get back on the electronics design horse.
    It sounds like you have a bit of a tinkerer thing going on and I can certainly relate to that and have nothing but respect for you. For me, it's catching up on the hot rodding I needed to do in my 20s but got married and had kids instead and was broke for the next 24 years. I remember reading the stories on the LT1 in the early 90s (which were always a footnote compared to the ZR-1 / LT5 stuff) and thinking that's the smallblock I'd want to build if I ever had the means.

    Quote Originally Posted by vilefly View Post
    Perhaps next week I can get those waveforms you need,
    Don't worry about me - I have enough to work with here. The only thing that would contribute slightly is seeing dwell at (above your rev limiter) and I wouldn't ask you to stress your engine that hard for something I've already determined isn't going to be "enough".

    Quote Originally Posted by vilefly View Post
    as well as limp mode behavior when the Hi-res signal is lost.
    I'm not too worried about this at the moment. Honestly, my take on "limp mode" is that if it doesn't start after 3-4 seconds of cranking something's broke - call your significant other to come get you so you can tow it back to the shop. This isn't a grocery getter. If I can conquer step 1 completely I might consider time based sequencing and the possibility of limp mode, but it's not something I'm going to lose sleep over.

  5. #185
    Fuel Injected!
    Join Date
    Oct 2013
    Posts
    1,022
    Quote Originally Posted by spfautsch View Post
    I would imagine it depends heavily on RPM and commanded spark advance. My assumption is that beyond some pre-determined RPM (for the sake of discussion let's pretend it's 5000) it simply switches the ICM back on immediately after it switched off as a "best effort" type methodology. But without someone dumping the TPU ROM or a bunch more waveform captures it's all conjecture.
    Why can't anyone read a whole paragraph?

    Quote Originally Posted by lionelhutz View Post
    At this point, I definitely would bet against switching to the next coil on each low-res crank signal working correctly. I too highly doubt that the PCM could command 46* advance at any rpm and still restrain the coil charging time to be between two low-res signals. I'd bet that the PCM turns the coil back on to fire the next cylinder coil before the low-res pulse identifying the cylinder that just fired even appears.
    As in the PCM must at some point turn the coil on again almost immediately after firing it to always be able to produce 46* of advance. When it does it exactly doesn't matter, the fact is that it's EXTREMELY likely that it does it.

    I just don't get the reason to cherry pick little bits of a response to make the exact same comment as the COMPLETE paragraph that you cherry picked has. You did it again when I was describing a method that could be used to provide 8 coils with the proper dwell at all time. Why else would I have wrote that the EST line can still be used to trigger the spark, besides the FACT THAT I WROTE the 8-coil LTCC module doesn't have to do any actual timing calculations in the previous sentence right before the line you felt you had to make a special point about basically repeating what I posted.
    Last edited by lionelhutz; 11-13-2017 at 05:31 AM.

  6. #186
    Fuel Injected!
    Join Date
    Jul 2013
    Location
    Newzealand
    Posts
    483
    The truck coils with finned heatsinks self trigger with too much dwell to stop them overheating it's the built in protection.the delco ecu I use with an external ls1 coil driver box the cranking dwell is too high and if it doesnt crank fast enough and start fast enough they cause a nasty kick back .lots of people with aftermarket ecus have the same issues too

  7. #187
    Fuel Injected! vilefly's Avatar
    Join Date
    Sep 2017
    Age
    53
    Posts
    217
    You sure are cute when you are pissy, Lionel.

    zoolander-doom-guy-party-car-.jpg

  8. #188
    Fuel Injected!
    Join Date
    Mar 2013
    Posts
    1,469
    Vilefly Thanks for the pics. They give some good hints.
    Did the dwell time input decrease gradually with rpm increase.
    If that is the case it can be calculated from the low res input or be a function of it. If that scaling can be turned off there will be always 5ms dwell available. Engine will run like crap at high rpm though but with 8 coils will be fine.
    What some of you miss is that there is always 90 degree charging time for the coil available, and that doesn`t depend on commanded spark advance.
    At 6000 rpm that will be 2.5ms between pulses.

    Can I assume that 5 volts on the est charge the coil and 0 volts fires the coil.
    I guess there is end of spark time just like end of injection and dwell is controlled by similar means.
    End of dwell time is calculated to happen at commanded spark advance degrees.

  9. #189
    Fuel Injected! vilefly's Avatar
    Join Date
    Sep 2017
    Age
    53
    Posts
    217
    so delcowizzid,
    You're saying that the finned truck coils add a certain amount of additional dwell to the trigger signal sent to it? Any specifics on that theory? We are all ears. I figure the feed back is for ignition burn time monitoring. Similar to nissan in design.....sort of.

  10. #190
    Fuel Injected! vilefly's Avatar
    Join Date
    Sep 2017
    Age
    53
    Posts
    217
    kur40,
    The dwell transition was smooth, and went down with increasing rpm.
    0v fires the coil, yup.
    Unfortunately, for the full 90 deg to be available, one would need a very fat rotor to get away it. but I have no definite intention of changing the spark advance much. So stock will still work for me. The 8 coils would allow one to reprogram a crazy-ass timing curve for sure.

  11. #191
    Fuel Injected!
    Join Date
    Oct 2013
    Posts
    1,022
    Quote Originally Posted by kur4o View Post
    What some of you miss is that there is always 90 degree charging time for the coil available, and that doesn`t depend on commanded spark advance.
    At 6000 rpm that will be 2.5ms between pulses.
    Yes, with a single spark control output, there is close to 90* of crank rotation charge time available if you start charging the coil again immediately after it was fired. It can never be a full 90* because there has to be time to turn the ICM off so it triggers the coil before turning it back on again.

    But, there is never even close to 90* of crank rotation charge time available if you start charging the coil each time a low-res signal appears. If you start charging when the low-res signal appears then at most you have (90 - commanded timing advance) degrees of crank rotation available to charge the coil.

    I agree with 2.5mS being the theoretical maximum, which means the measured minimum dwell number of 3.2mS is not possible.

    It's common to find the low vacuum high rpm area of the spark tables containing 46*. At 6000rpm and 46* of advance, the theoretical maximum coil charging time is 1.22mS if the coil charging starts when a low-res pulse is received.

    Even though it can't be the real minimum dwell, the measured minimum dwell of 3.2mS above 4000rpm does indicate that the coil charging does not start when a low-res pulse is received but rather right after the coil is fired.
    Last edited by lionelhutz; 11-14-2017 at 08:07 AM. Reason: did the calculation in bold wrong.

  12. #192
    Fuel Injected!
    Join Date
    Jul 2013
    Location
    Newzealand
    Posts
    483
    Quote Originally Posted by vilefly View Post
    so delcowizzid,
    You're saying that the finned truck coils add a certain amount of additional dwell to the trigger signal sent to it? Any specifics on that theory? We are all ears. I figure the feed back is for ignition burn time monitoring. Similar to nissan in design.....sort of.
    No just if you dwell them for to long they will cut the 5v signal themselves and fire a spark no matter the crank postion it's a nasty kick back when they fire on the compression stroke well before tdc lol

  13. #193
    Fuel Injected! spfautsch's Avatar
    Join Date
    Apr 2015
    Location
    Montgomery City, MO
    Age
    52
    Posts
    883
    I was wondering what happened when the dwell limiting kicked in. The info I've seen states they all have dwell limiting but the LS1 and the non-truck LS2 coils must employ a different limiting mechanism like clamping charge current.

    That's bad with a capital B.

  14. #194
    Fuel Injected!
    Join Date
    Oct 2013
    Posts
    1,022
    Sounds like a broken starter and/or block waiting to happen.

  15. #195
    Fuel Injected!
    Join Date
    Oct 2013
    Posts
    1,022
    Quote Originally Posted by vilefly View Post
    Unfortunately, for the full 90 deg to be available, one would need a very fat rotor to get away it. but I have no definite intention of changing the spark advance much. So stock will still work for me. The 8 coils would allow one to reprogram a crazy-ass timing curve for sure.
    I don't think you're understanding the maximum 90* of dwell number. This is a physical limit based on the basic engine layout of the LT1 and the fact that a single coil is being used in the ignition system. It has nothing to do with the timing advance being used. As long as the LT1 PCM is kept in control it also doesn't change even when switching to 8 coils.

    On this engine, a cylinder fires every 90* of crank rotation. At most, a single coil ignition system can have the coil turned-on to magnetically charge it for the whole time between each ignition event. This will end up being a little less than 90* of crank rotation, but lets call it 90* just for simplicity.

    At 6000 rpm, 90* of crank rotation is 2.5mS. That is the theoretical maximum possible dwell at 6000 rpm. Of course, some of that time will be lost firing the ignition and turning the coil back on so in practice the dwell time will be less. What this means is that the 3.2mS you measured can't be the minimum possible dwell of the LT1 ignition system. It is physically impossible to have that much dwell at any RPM above 4687 rpm and still fire every cylinder on it's compression stroke.

    All the above assumes you turn the coil back on to begin charging it again right after it fires, not at the next TDC event where you see the next low-res signal. Turning it back on at the next TDC event or leading edge of the next low-res signal will significantly reduce the dwell time.

Similar Threads

  1. Which TBI system is better?
    By KeyAir in forum GM EFI Systems
    Replies: 41
    Last Post: 05-13-2019, 09:39 PM
  2. Hard start 93 LT1 with LTCC Ignition Mod
    By beestoys in forum GM EFI Systems
    Replies: 0
    Last Post: 05-18-2015, 08:58 AM
  3. ABS system?
    By K1500ss4x4 in forum Gear Heads
    Replies: 3
    Last Post: 02-06-2014, 06:21 AM
  4. Vortec EGR System?
    By EagleMark in forum OBDII Tuning
    Replies: 40
    Last Post: 06-02-2013, 10:07 PM
  5. Quicker way to do Spark Hook test on the street for LT1s and others?
    By sherlock9c1 in forum Fuel Injection Writeups Articles and How to New and Old
    Replies: 15
    Last Post: 03-03-2013, 01:52 AM

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •