help with C.O.P. being run from the ecu.
Thread Starter
Honda-Tech Member
Joined: Sep 2005
Posts: 282
Likes: 0
From: fountain inn, sc. where they still rock banjos on porches, usa
help with C.O.P. being run from the ecu.
hi i post this on pgmfi.org and i'm really need help with this. i do alot of reading here and alot of other forums. so i've been do some research using C.O.P. found this thread.
http://forum.pgmfi.org/viewtop...+pack
so i feel that this will be a contunious of that last one. but i come with a bunch of info but need help in decoding it cause i feel i have all the info but i can't seem to decphier it. so here it is:
I really like the CBR coils(CBR600RR) and the fact you are using a farely unknown ems. Does the vems system use the factory harness?
Yea VEMS is connected to factory harness. Although I've made my own wiring for engine position sensors (crank and cam) and injectors/coils since I've DPFI harness in my car. Also WBO2 has its own harness (yes it has realtime closed-loop WBO2 control even at WOT, could even use AutoTune to make fuel tables simply based on given AFR table).
VEMS for sure is fairly unknown in US, since the mainboards are made in and sold from Hungary. VEMS is originally a fork of MegaSquirt (originally called megasquirtavr), based on Atmel's ATMega128 MCU. Its basicly an "open" "DIY" project very similar like Linux is in computer world (btw Linux originated from Finland).
Look at the vems.hu website. Its extremely powerful and versatile system with almost any imagineable function. It only lacks diagnostics and datalogging is limited only to the sensor inputs, but it can write them into SD card. It also used separate LCD screen and can be tuned with simply PS/2 keyboard. Has eight coil outputs (IGBT powerstages, connects coils directly to ECM), eight injector outputs, two WBO2 controllers, two EGT channels and LOTs of more. For a cost of 1/5 or less of Apexi/AEM unit with lesser features.
I might do a separate VEMS introduction thread, and if someone is seriously interested to getting one I can provide both the hardware and some support.
found some info.
http://causeforalarm.thecarthi....html
but my question is how did you elimate the crank sensor? with a Hall-Effect sensors?
well a COP DIY would be really difficult, since its hugely EMS-related stuff. And that link is mostly BS. You do not need any AEM products and especially no CDI. Most ppl don't even understand what CDI is and how it really works. CDI stands for Capasitive Discharge Ignition. It has switch mode power supply similar to car amplifiers and computer PSUs to raise car's 12V voltage to around 400V. This high voltage is stored into capasitor, and then suddenly released thru the ignition coil. Stock ignition systems are known as inductive systems.
You gotta understand this; CDI- and inductive ignition coils are VERY different in construction and operation. Inductive systems flow current thru the coil and restore the energy into coil's magnetic field. When current is suddenly cut, the magnetic fiel collapses and because of induction, the secondary voltage raises because current flow doesn't want to change.
In CDI system, the energy is stored into the capasitor in so-called CDI-box or amplifier, NOT in ignition coil. Capasitor's voltage is suddenly conducted into the coil that creates very fast current change and that is converted into high voltage.
So, those ignition coils are very different. Where they also differ is the turn/transform ratio. While typical inductive coil has a ratio of around 400:1, CDI-coils have usually around 100:1.
They also differ in performance. While CDI has very high spark energy, the spark duration is VERY short (one tenth to one hundredth) compared to inductive systems. Inductive systems could have spark duration of several milliseconds, while CDI durations are in microseconds. This is where MSD (Multiple Spark Discharge) originated; they rapid-fire several short-duration sparks in series. Both systems have pros and cons.
Its needless to say that id you use wrong coil with wrong type system, you get VERY bad performance, if it works at all. Those CBR coils are all inductive systems, so DO NOT USE THEM with CDI-system! Since they're inductive and my VEMS ECU has eight IGBT-powerstage outputs, I can connect those coils directly into my EMS.
AEM has their pencil coils made especially for their CDI-setups.
Yes I have two Hall-effect sensor for monitoring the engine position. One is located at the crank pulley, and another one is at the camgear(you can see the red tubular sensor there). This is for maximum accuracy. Honda's original setup is very poor since its located in the dizzy, at the end of camshaft. Why its so poor then?
First issue is the timing belt. Belt itself causes sloppiness of several degrees into timing. It can easily be +/- 5 degrees! Another factor is the camshaft itself. It flexes, twists and springs all around. This adds again the inaccuracy of few degrees. This problem only gets multiplied with stiffer valvesprings, hotter cams and higher revs. Thats why its so important to have accurate engine position sensor setup in a performance engine. To be most precise, the crank position sensor should be located at the flywheel end, since there is where you get the output from engine and where the EMS timing should be referred. Look at the D17 good-bad thread in NA section for more info related to this.
Now you can see why the new AEM EPM is very bad setup. It removes dizzy and allows you to run COP, but thats it. I'd personally never run such of **** in my own engines.
One thing more about the COP setup; You don't necessarily even need standalone EMS if you've decent electronics skills. It wouldn't be hard at all to convert stock Honda OBD1 EMS to have cranktrigger and run COP. Yes it only has single coil output, but it has sequential timed injection. It'd be relatively easy to design and build electronic multiplexer with simple logic arrays to spread the ignition signal into four independent powerstages. Coil dwell (charge time) could be controlled from ECU, or have simple electronic for that in the multiplexer/powerstage unit. From my experience, stock Honda coil works fine with 3ms dwell, as does my CBR929 coils. I found instructions for some CBR600F4i standalone ECU, and it also told to set dwell time to 3ms with stock CBR600 coils.
Although it would be tempting to use CBR coils with a obd1 system it would be difficult to create this "multiplexer". Most systems utilizing the factory ecu do not have the capability to adjust dwell so these "powerstages" would be needed. Are there mulitplexer/powerstage systems available in Finland?
Such "systems" are not available anywhere AFAIK, but like I said it'd be very simple and straightforward to design and build for anyone who has decent electronic (both analog and digital, hi- and low-power) skills.
Here is the basic layout what I've in my head:
First there is the multiplexer part. You take the ignitor signal from the original ECU. Then, take 4-channel MUX/switch chip and route the ignitor signal into every input of the chip. Then you need to figure out the way how to enable to correct switch channel (which cylinder fire at the time). I've two solutions for this; You could take the enable-signals from injector outputs. Of course you'd need to figure out which coil to fire when some injector is spraying. This would need a flip-flops to keep ignition channels enabled for long enough time.
Another way to determine the correct coil in firing order is to utilize the trigger signals. You could use cranktrigger signal (OBD2 engines have such in the oilpump!) feed into counter, and after it reaches the set values it'd enable the multiplexer channels in determined order. Then you've camtrigger signal, one pulse per cam revolution (stock MPFI dizzys have this, and its easy to implement separately like I did). This would be used to reset the counter.
Now you'd just need to implement the four power stages for the coils and how to determine the correct dwell time. Like I previously said, the dwell time of stock coil and CBR coil is pretty close, so you could let the ECU determine it. Since we're using analog switches they pass the ECU signal with correct dwell already so only a simple high-power switches would be needed as coil ignitors/powerstages. For power stages, you could use simple IGBTs (insulated gate bipolar transistor) like I've in VEMS. One (but very complex and expensive) way could be something like utilizing four Honda ignitor modules. If the ignitor module itself adjusts the coil dwell, this could be very well working solution (although bulky and expensive). There is separate chips made to control ignition coil drivers and dwell time, so those could also be used, but might be tricky and expensive to get.
If I had OBD1-Honda, it wouldn't be hard at all to make such system. I'd also convert the EMS to use separate sensor setup positioned at the crank and cam like I have currently to get completely rid of dizzy. Like I said, I'd personally want to stay away from inaccurate AEM EPM-style solution, although it'd make the COP-system very easy to implement.
About my wiring harness, there is nothing special. ECU is connected to stock harness, just changed the big connector in it. Crank/cam sensors are wired with shielded microphone cable, and each injector/coil has own wire going into ECU with common +12V supply from stock harness.
EDIT: BTW what if you use generic aftermarket high-power ignition coil like MSD Blaster, Accel etc with Honda? Do you just wire it into stock ignitor in the dizzy just replacing/bypassing the stock coil and hope it works? Or is the ignitor changed too, or can you adjust the coil dwell time with Crome/Uberdata/Hondata/Neptune/Whatever?
so there is where my info ends or may i say all the help ended. so can someone help me figure this out cause i know all my answeres are with in thses post but i'm having a difficult time seeing this clearly. also so the person who did this is an automotive engineer in findland.
thank you
remo.
anyinfo on how to start this, in the sense of a list of hardwear and softwear that is capible of this please list so i can start the research. it will be very helpful.
http://forum.pgmfi.org/viewtop...+pack
so i feel that this will be a contunious of that last one. but i come with a bunch of info but need help in decoding it cause i feel i have all the info but i can't seem to decphier it. so here it is:
I really like the CBR coils(CBR600RR) and the fact you are using a farely unknown ems. Does the vems system use the factory harness?
Yea VEMS is connected to factory harness. Although I've made my own wiring for engine position sensors (crank and cam) and injectors/coils since I've DPFI harness in my car. Also WBO2 has its own harness (yes it has realtime closed-loop WBO2 control even at WOT, could even use AutoTune to make fuel tables simply based on given AFR table).
VEMS for sure is fairly unknown in US, since the mainboards are made in and sold from Hungary. VEMS is originally a fork of MegaSquirt (originally called megasquirtavr), based on Atmel's ATMega128 MCU. Its basicly an "open" "DIY" project very similar like Linux is in computer world (btw Linux originated from Finland).
Look at the vems.hu website. Its extremely powerful and versatile system with almost any imagineable function. It only lacks diagnostics and datalogging is limited only to the sensor inputs, but it can write them into SD card. It also used separate LCD screen and can be tuned with simply PS/2 keyboard. Has eight coil outputs (IGBT powerstages, connects coils directly to ECM), eight injector outputs, two WBO2 controllers, two EGT channels and LOTs of more. For a cost of 1/5 or less of Apexi/AEM unit with lesser features.
I might do a separate VEMS introduction thread, and if someone is seriously interested to getting one I can provide both the hardware and some support.
found some info.
http://causeforalarm.thecarthi....html
but my question is how did you elimate the crank sensor? with a Hall-Effect sensors?
well a COP DIY would be really difficult, since its hugely EMS-related stuff. And that link is mostly BS. You do not need any AEM products and especially no CDI. Most ppl don't even understand what CDI is and how it really works. CDI stands for Capasitive Discharge Ignition. It has switch mode power supply similar to car amplifiers and computer PSUs to raise car's 12V voltage to around 400V. This high voltage is stored into capasitor, and then suddenly released thru the ignition coil. Stock ignition systems are known as inductive systems.
You gotta understand this; CDI- and inductive ignition coils are VERY different in construction and operation. Inductive systems flow current thru the coil and restore the energy into coil's magnetic field. When current is suddenly cut, the magnetic fiel collapses and because of induction, the secondary voltage raises because current flow doesn't want to change.
In CDI system, the energy is stored into the capasitor in so-called CDI-box or amplifier, NOT in ignition coil. Capasitor's voltage is suddenly conducted into the coil that creates very fast current change and that is converted into high voltage.
So, those ignition coils are very different. Where they also differ is the turn/transform ratio. While typical inductive coil has a ratio of around 400:1, CDI-coils have usually around 100:1.
They also differ in performance. While CDI has very high spark energy, the spark duration is VERY short (one tenth to one hundredth) compared to inductive systems. Inductive systems could have spark duration of several milliseconds, while CDI durations are in microseconds. This is where MSD (Multiple Spark Discharge) originated; they rapid-fire several short-duration sparks in series. Both systems have pros and cons.
Its needless to say that id you use wrong coil with wrong type system, you get VERY bad performance, if it works at all. Those CBR coils are all inductive systems, so DO NOT USE THEM with CDI-system! Since they're inductive and my VEMS ECU has eight IGBT-powerstage outputs, I can connect those coils directly into my EMS.
AEM has their pencil coils made especially for their CDI-setups.
Yes I have two Hall-effect sensor for monitoring the engine position. One is located at the crank pulley, and another one is at the camgear(you can see the red tubular sensor there). This is for maximum accuracy. Honda's original setup is very poor since its located in the dizzy, at the end of camshaft. Why its so poor then?
First issue is the timing belt. Belt itself causes sloppiness of several degrees into timing. It can easily be +/- 5 degrees! Another factor is the camshaft itself. It flexes, twists and springs all around. This adds again the inaccuracy of few degrees. This problem only gets multiplied with stiffer valvesprings, hotter cams and higher revs. Thats why its so important to have accurate engine position sensor setup in a performance engine. To be most precise, the crank position sensor should be located at the flywheel end, since there is where you get the output from engine and where the EMS timing should be referred. Look at the D17 good-bad thread in NA section for more info related to this.
Now you can see why the new AEM EPM is very bad setup. It removes dizzy and allows you to run COP, but thats it. I'd personally never run such of **** in my own engines.
One thing more about the COP setup; You don't necessarily even need standalone EMS if you've decent electronics skills. It wouldn't be hard at all to convert stock Honda OBD1 EMS to have cranktrigger and run COP. Yes it only has single coil output, but it has sequential timed injection. It'd be relatively easy to design and build electronic multiplexer with simple logic arrays to spread the ignition signal into four independent powerstages. Coil dwell (charge time) could be controlled from ECU, or have simple electronic for that in the multiplexer/powerstage unit. From my experience, stock Honda coil works fine with 3ms dwell, as does my CBR929 coils. I found instructions for some CBR600F4i standalone ECU, and it also told to set dwell time to 3ms with stock CBR600 coils.
Although it would be tempting to use CBR coils with a obd1 system it would be difficult to create this "multiplexer". Most systems utilizing the factory ecu do not have the capability to adjust dwell so these "powerstages" would be needed. Are there mulitplexer/powerstage systems available in Finland?
Such "systems" are not available anywhere AFAIK, but like I said it'd be very simple and straightforward to design and build for anyone who has decent electronic (both analog and digital, hi- and low-power) skills.
Here is the basic layout what I've in my head:
First there is the multiplexer part. You take the ignitor signal from the original ECU. Then, take 4-channel MUX/switch chip and route the ignitor signal into every input of the chip. Then you need to figure out the way how to enable to correct switch channel (which cylinder fire at the time). I've two solutions for this; You could take the enable-signals from injector outputs. Of course you'd need to figure out which coil to fire when some injector is spraying. This would need a flip-flops to keep ignition channels enabled for long enough time.
Another way to determine the correct coil in firing order is to utilize the trigger signals. You could use cranktrigger signal (OBD2 engines have such in the oilpump!) feed into counter, and after it reaches the set values it'd enable the multiplexer channels in determined order. Then you've camtrigger signal, one pulse per cam revolution (stock MPFI dizzys have this, and its easy to implement separately like I did). This would be used to reset the counter.
Now you'd just need to implement the four power stages for the coils and how to determine the correct dwell time. Like I previously said, the dwell time of stock coil and CBR coil is pretty close, so you could let the ECU determine it. Since we're using analog switches they pass the ECU signal with correct dwell already so only a simple high-power switches would be needed as coil ignitors/powerstages. For power stages, you could use simple IGBTs (insulated gate bipolar transistor) like I've in VEMS. One (but very complex and expensive) way could be something like utilizing four Honda ignitor modules. If the ignitor module itself adjusts the coil dwell, this could be very well working solution (although bulky and expensive). There is separate chips made to control ignition coil drivers and dwell time, so those could also be used, but might be tricky and expensive to get.
If I had OBD1-Honda, it wouldn't be hard at all to make such system. I'd also convert the EMS to use separate sensor setup positioned at the crank and cam like I have currently to get completely rid of dizzy. Like I said, I'd personally want to stay away from inaccurate AEM EPM-style solution, although it'd make the COP-system very easy to implement.
About my wiring harness, there is nothing special. ECU is connected to stock harness, just changed the big connector in it. Crank/cam sensors are wired with shielded microphone cable, and each injector/coil has own wire going into ECU with common +12V supply from stock harness.
EDIT: BTW what if you use generic aftermarket high-power ignition coil like MSD Blaster, Accel etc with Honda? Do you just wire it into stock ignitor in the dizzy just replacing/bypassing the stock coil and hope it works? Or is the ignitor changed too, or can you adjust the coil dwell time with Crome/Uberdata/Hondata/Neptune/Whatever?
so there is where my info ends or may i say all the help ended. so can someone help me figure this out cause i know all my answeres are with in thses post but i'm having a difficult time seeing this clearly. also so the person who did this is an automotive engineer in findland.
thank you
remo.
anyinfo on how to start this, in the sense of a list of hardwear and softwear that is capible of this please list so i can start the research. it will be very helpful.
Joined: Oct 2006
Posts: 0
Likes: 0
From: Usolie_Siberian, Siberia, Russia
Junior Member
Joined: Jun 2008
Posts: 141
Likes: 0
From: Covington, GA, USA
I'm not on PGMFI.org, but if you are replacing the ECU with a standalone anyway, there are a number of ECUs out there that can use the stock sensors in the distributor to trigger coil on plug.
Honda-Tech Member
Joined: Jul 2002
Posts: 812
Likes: 0
From: Peg City, 204
Thread
Thread Starter
Forum
Replies
Last Post
D3U1LT
Honda Civic / Del Sol (1992 - 2000)
9
Nov 28, 2004 02:06 PM
