1985 GL1200 Limited ECU Replacement/Upgrade

Help Support ClassicGoldWings:

Rednaxs60

Well-known member
CGW Supporter
Joined
Dec 25, 2015
Messages
2,448
Reaction score
38
Location
Victoria, BC,
First Post - have mentioned fellows from the NGW forum, but have found most of my information over there, even seen posts from younger versions of Joe and Dan. It's going to be an interesting quest. Have sent an email to DIY regarding the Megasquirt unit.

Still have a fuel/oil smell and some blue smoke on start. Investigated this. Pressurized the fuel system and did a borescope inspection of the backside of the intake valves. All four cylinders dry. Checked numbers 2 and 4 cylinders after sitting for a night on the side stand. Number 4 cylinder dry, number 2 cylinder some oil:
IMG_1693.JPG
Not a big issue, but annoying. This will be the main item I will take care of this coming winter, rings and remove cylinder glaze. When I had the heads off, you could see the original cross hatch marking, but the cylinders were quite shiny. This work should finish off the mechanical work this bike needs.

Now to the subject of this thread. I have been learning about the CFI system since I bought the bike. I researched it before I bought it and some items came to the fore such as TPS and crank angle senor (Ns) replacement. Other areas that have been looked at and I have replaced are the PB sensors (MAP) with more modern Suzuki MAP sensors - better technology than the older Honda OEM Pb sensors.

The TPS is a linear rheostat so just about anything works. There is a catch with this in that the QA for these TPS components is sketchy at best. New TPS components can be problematic, found this out the hard way, right from the start. A faulty TPS can make the engine have "hard" misfires that feel like the engine is going to stop. Found a web site that had oscilloscope graphs showing good and faulty TPS components that were new. The faulty components had a lot of chatter in the mid range voltage that would send a faulty signal to the ECU and the ECU would react accordingly. Heat also affects these components adversely as well.

The Gr/Gl sensors - commonly called PG sensors - located on the back end of the right cylinder head, are not available but the folks over on the CX500-CX650 turbo forum have found the LX579 Sensors used in Dodge products are a suitable replacement component.

I have replaced the Ns sensor as well. My bike had the 1500 mod where the engine case had to be modded to allow the 1500 sensor to fit. I replaced this with a set of PG sensors from an '85 Aspencade that fit in the timing belt area. I have both installed but only use one. The second one will be a spare in case the one I am using fails. In having the second sensor installed won't have to get into the timing belt area, just have to switch the wiring under the shelter.

The primary components that make up the computerized fuel injection (CFI) system are:

Air temp sensor (T1)
Manifold pressure sensors (PBr/PBl) qty 2 - one for each cylinder bank
Throttle position sensor (TPS)
Coolant temp sensor (Tw)
Crankshaft angle sensor (Ns)
Cam shaft angle sensor (Gr/Gl)
Fuel pump relay
CFI main relay
Fuel shut-off sensor
ECU

I have been reviewing the literature about the Megasquirt unit and it is quite comprehensive, and there is a lot of it. I also read/reviewed the various forum threads regarding EFI projects.

Socrace, thesteve, irishcarbomb, toehead, FirstYearDeek, and taunusrainer (lots of other threads but can't mention everyone) threads have been very informative.

So far I have yet to find a thread on just an '85/'86 GW fuel injected bike ECU upgrade/replacement. My quest will be to, hopefully, achieve this without too much modification/component change out. I am intrigued; however, with socrace's use of Ford COP, and the various spark igniter units mentioned in the Megasquirt literature. The wasted spark coil in the Megasquirt literature is also interesting. Socrace also mentions that the spark igniter units used in the '85/'86 FI systems can be used.

I read on one of the threads where it was mentioned why one should go with EFI and the benefits of such. For myself, I prefer FI over carbs. When EFI came mainstream in the auto industry, my Father being a mechanic, thought he'd won the loto. Cold winter mornings, boosting and getting flooded cars going dropped to a minimum. Carb and EFI have issues. I believe socrace mentioned that dong an EFI project is like owning an older vintage bike, not for the faint of heart, and nothing is inexpensive.

The dual PB (MAP) sensors used by Honda has not been explained, and questioned in a few threads. I have read where a fellow did "marry" the two into one, use a splitter to connect back into the system and everything worked fine. I think this is completely plausible. To upgrade to the Megasquirt unit, this is probably what will have to happen - not a big issue.

Socrace has mentioned that the Honda OEM TPS can be used, but any linear TPS can be also. Have used a TPS for an early model Civic as a replacement for the Honda OEM unit, worked fine once I got a good one.

The '85 LTD has an 8 tooth trigger wheel for the crankshaft sensor (Ns). Read about using a 36-1 but would have to research this. The GL1500 has a 12-1 trigger wheel that would probably fit in place of the '85 LTD Ns 8 tooth trigger wheel. Possible alternative.

The Gr/Gl (PG) sensors on the backend of the right cylinder head was mentioned in the Megasquirt manual. Single tooth that will apparently work, but a change to a 12 tooth trigger in place of is recommended.

There is discussion regarding the OEM injectors for the '85 LTD. I had mine cleaned and the flow results came in at 65 ml/min after cleaning. The numbers before cleaning were 64 ml/min. Pressure setting for injector test(s) was 40 PSI. The injector resistance reading was 2.6 ohms for each. The original Honda OEM injectors are quite robust. These injectors are being used in several projects and I have been taking notes regarding the injector impact on tuning the EFI system.

The fuel pump requirement is for 630 cc/min minimum. Static pressure is 34-38 PSI and dynamic 28-34. My system came in at 38 PSI static, and 36 PSI dynamic. When I was searching for the reason I had a fuel/oil smell on start, I pressurized the fuel system and let it sit. Pressure dropped 12 PSI in one hour. Think this is a bit excessive so I am looking for a suitable fuel pressure regulating valve that will fit where the original is.

The idle air control system is a passive system that works quite well for its design. The idle air control (IAC) valve is activated by a bi-metallic element (courtesy of socrace). The IAC valve is wired in parallel with the fuel pump - fuel pump on power to IAC valve. It is never fully closed even when cold and not used for a while. The water system hoses connected to the IAC valve are to prevent freeze up in colder climates, other than this requirement, no need to have a water heated valve. The reed valves work quite well and are operated by cylinder vacuum. I replaced the IAC system hoses, but when I installed the air chamber one of the hoses came off the right side reed valve (had to remove air chamber to repair). These reed valves suck a lot of air and the noise is quite loud. If you did not know better, you would think there was a metal on metal issue, I did, But when I investigated using a stethoscope (Long screwdriver) there was no internal metal noises. The Megasquirt literature I have read indicates that this passive IAC system is acceptable.

Here are twp pics of the IAC valve internals. This is when cold:
IAC VV air opening when cold.jpg
This when bi-metallic strip heated opening IAC valve:
IAC VV air opening when warm.jpg
There will always be IAC system air going into the engine. The IAC system is not just for operation when cold, but provides air when the throttle is shut, cutting off air to the cylinder(s). A small amount of air bypasses the throttle plate(s). Here is a picture of the IAC system:
Air Chamber Underside 2.jpg
You can see the reed valves and where the output from the reed valve(s) is on the cylinder side of the throttle plate(s).

I mentioned that a hose came off the right side reed valve when I out the air chamber back in. The hose size is approximately 1/2 inch or so. This is a picture of what I found:
IAC System.jpg
The lower arrow is where the IAC valve gets its air supply.

I think I've come to the end of my introduction into the Megasquirt world, and the quest for a suitable ECU upgrade/replacement. More to follow.

Cheers
 

Rednaxs60

Well-known member
CGW Supporter
Joined
Dec 25, 2015
Messages
2,448
Reaction score
38
Location
Victoria, BC,
Had some home chores to do, but this gives me time to reflect on all the information I have been reading.

My thoughts have been turning towards the ignition requirements. The '85 LTD ECU uses the crank angle sensor (Ns), PB sensors (MAP), TPS (0th), Gr/Gl sensors (PG), and coolant temp (Tw) for ignition timing. The PB sensors are good up to approximately 3000 RPM, after which these components are not used in the equation. I mention this because in testing these components on the bench, these max out for voltage at approximately 13 in Hg.

Coolant temp is monitored, and while used for ignition timing, I surmise that it is also part of the fast idle aspect of the '85 LTD. The same can be said for the air temp (T1) sensor - used for fast idle, but once the engine warms up, and you can watch the correlation between the engine temp and the idle speed - engine warms up, RPM drops.

The T1 sensor detects air temp and sends signals to the ECU. The service manual supplement for the '85 LTD does not go further into a description for this sensor. Has to impact on the timing especially on start.

I have compared the schematics, the '85 LTD against the Megasquirt-2 V3.0/V3.57 wiring and these are quite similar. It will take me a few reviews to digest the similarities and how these can be adapted. As I mentioned in my previous post, I want this to be an upgrade/replacement with not too many alterations if possible.

The GW ECU and the Megasquirt appear to operate the same in that each unit controls the ground of the component to be controlled such as the coils, injectors.

The Megasquirt schematic has a main and fuel pump relay similar to the '85 LTD. The wiring is much the same because of this.

I have mentioned that the PB sensors can probably be consolidated into one. To prove this I think I will make a splitter so that I can operate the bike off one PB sensor. This would prove a theory I have, and this understanding would be beneficial when upgrading the OEM ECU.

I have to confirm that a passive IAC system can still be used with an upgrade/replacement ECU - more reading to do.

The inputs to the ECU that allows the ECU to control the engine operation are next on my list of things to think about.

Printed out the Megasquirt MS3 Gold Box V1.2 Hardware manual this afternoon. It appears to be similar to the Megasquirt MS2V3.0 Hardware Guide. More information to digest.

Cheers
 

Rednaxs60

Well-known member
CGW Supporter
Joined
Dec 25, 2015
Messages
2,448
Reaction score
38
Location
Victoria, BC,
Thanks Joe - it will be a good project. In contact with DIYautotune regarding the application - very helpful folks. Matt at DIYautotune mentioned that if I want to stay with bank to bank fueling I could probably use the Microsquirt, or upgrade to the MS3 to do more such as sequential fuel control. Matt also mentions that the cam angle sensors - Gr/Gl sensors (PG) on the rear of the of the right cylinder head can be reduced to one sensor/signal for use with the Megasquirt. This would be beneficial in that there would be a spare cam shaft angle sensor available should one fail. He also mentioned that the 8 tooth trigger wheel for the crank angle sensor (Ns) will work as well. I think I should stay as close as possible to original build, but have the ability to tune the system for new components should the old ones have to be changed out. Thinking the Megasquirt MS2 with the latest boards/software would be best.

I have been looking on the internet to determine what a used ECU would cost and have found one and it is close to $300.00 CDN. A new MS2is approximately $500.00 CDN and can be tuned, great possibilities and with new technology.

I have been reading on line where there is an ITB (independent throttle body ?) map that uses an Alpha-n and speed density values. Speed density initially uses the MAP, this would be the PB sensors of the '85 LTD FI system. It has been questioned why there are two PB sensors on the '85 LTD FI system, but that is what Honda did. I have read where this could be changed so that only one PB sensor is used, but for the OEM installation you would still have to split the signal and connect to the CFI system as if you were using two PB sensors.

The speed density mode uses MAP, TPS, temperature and RPM to control the engine. This can be converted so that all that is being used is RPM, temperature and TPS, making it the Alpha-N mode. The Alpha-N mode uses TPS and RPM to calculate the fuel amount to be injected instead of MAP and RPM in the speed density mode. This is how the '85 LTD FI system operates. The CFI system uses the PB (MAP) signal up to around 3000 RPM, after which it discounts this signal - believe it is maxed out - and uses TPS and RPM as the main signals for engine operation.

The MS unit needs a TPS signal of "0" with throttle closed. Going to check the TPS calibration to determine how close the TPS is to "0" after calibration.

Have learned a little more, but still a lot of homework to do. Cheers
 

Rednaxs60

Well-known member
CGW Supporter
Joined
Dec 25, 2015
Messages
2,448
Reaction score
38
Location
Victoria, BC,
Came to another issue last night. Lots of info to digest. I have the travel computer on my bike, great piece of kit, still haven't figured out or used all the functions. It's a legal driving while distracted device as it is incorporated as part of the bike. I'm certain it uses a lot of the signals from the sensors to provide the info it does. Went onto the DIYAutoTune website and found its article on wiring the MS in parallel with the OEM ECU so that I can keep the travel computer functionality, but use the MS for engine operation. The article does mention that there are signals that the MS does not like to share so will be looking into what is needed.

Have read socrace's thread on coil on plugs (COP) on the NGW forum, and how he achieved the upgrade. Also found a YouTube video where a car was upgraded to COP, car had a wasted spark system as well. The car video just split the signal to the original coil into two and connected the new COP units. Socrace's is a bit more involved, but there is always more than one way to achieve the aim. I like the COP concept and application. It has benefits that the OEM wasted spark setup does not, and it is less expensive. Would like to see how socrace married the COP to the OEM plug boot. This upgrade would be beneficial even if no ECU upgrade/replacement is done.

It was commented on in an EFI thread that the original CFI system, while functional, lacked the nuances of the newer technologies, yes I agree, but when Honda did this back in the '80s, state of the art and on the cutting edge. We now get to reap the benefits of the past 30 odd years, and can keep these older FI bikes on the road, and see what can be done with the ability to do tuning.

More to follow. Cheers
 

joedrum

Well-known member
CGW Supporter
Joined
Dec 8, 2009
Messages
19,374
Reaction score
4
Location
chuluota florida
Did this once ..can’t remember much about it ..but I’m thinking heat was the problem and of course the way high plug boots of the set up was quite out of place it seemed ...I’m thinking I rode it a few times and it took out one of the cop I was using ...but not sure ...very long time ago and unlike you I documented nothing ..I just love your complete posting of everything
 

Rednaxs60

Well-known member
CGW Supporter
Joined
Dec 25, 2015
Messages
2,448
Reaction score
38
Location
Victoria, BC,
One of the issues with using coil on plug (COP) units is that these really work well with computerized fuel injection, not so much with other forms of ignition. The COP units I'm looking at are the same as socrace used, for a Ford F-150. The resistance of these are approximately 0.7 ohms each and when in parallel the total resistance is approximately 2.85 ohms, quite similar to the 3.0 ohm coils presently used by the OEM, so the system should not see an appreciable difference - could add a small resistor just for good measure. I mention this because some of the circuits on the 1200 FI model needs a certain minimum load, such as the running lights.

I appreciate socrace's interest in this in that it frees up space under the shelter, gets rid of plug wires, and better engine performance. I mention engine performance in that it should have a hotter spark to each plug. As with everything, when something is shared, it is possible that the medium being shared will take the easiest route. Wasted spark works well with non hi performance vehicles, but not so much the other.

The firing order for the 1200 is apparently 1/3/2/4. Don't know if sequential would benefit the system, but something to consider. An issue that came up to be resolved is maintaining the ancillary items such as the travel computer working. I have perused the schematics and the travel computer uses a signal from 1 and 3 injectors for certain calculations. Going to sequential firing would probably be more complicated wiring, don't know, just a thought.

Socrace documented his COP mod in this thread: https://ngwclub.com/forum/viewtopic.php ... on#p534261

Very well detailed and makes good sense. 8 of these COP units is approximately $60.00 CDN with shipping. Connectors another $5.00. One aftermarket coil for the 1200 is $70.00. This type of mod should work with the 1200 carb models, don't know about the 1000, 1100.

I liked where socrace posted that he adjusted his EFI system on his 1100 to operate quite rich and the OEM wasted spark system did not take kindly to it, couldn't get past 6000 RPM. Put in the COP units and no issue with the engine. Definitely a mod worth investigating.

Cheers
 

Rednaxs60

Well-known member
CGW Supporter
Joined
Dec 25, 2015
Messages
2,448
Reaction score
38
Location
Victoria, BC,
Starting to read into the Megasquirt unit. Lots of information to digest, but quite informative. Been through the "Injectors and Fuel System" literature. Have a handle on this one, but will be reading again.

Question regarding the Megasquirt unit. Ready made or DIY? Did soldering on circuit boards many moons ago, but would be extremely rusty at this point. Worth the headache, time and resources as a DIY project, or just bite the bullet and get a ready made? Tending towards the ready made simply because I will be doing an upgrade/replacement of the ECU not a complete build.

Starting in on the "Tuning Your Megasquirt". This section mentions that having an operating system is extremely beneficial. Mine is working well so this is a system upgrade/replacement project. I will also have the benefit of having the system tuned so that the input to and outputs from the Megasquirt should be easier. This section also indicates that you do not need an O2 sensor for the system to operate, page 6.

Have looked at the MAP requirement. The '85 LTD has two PB (MAP) sensors. I have commented on these before. I looked at the service manual supplement and there is a troubleshooting guide for these sensors with the required specs for the PB sensors. Have to correlate these values to what is needed by the Megasquirt unit for programming.

Did a comparison between the OEM PB sensors and the Suzuki MAP sensors. Info regarding is:

Brand: Denso.
Type number: 100798-5630. Suzuki ordering number: 15620-35F00.
Suzuki description: Sensor Boost, IAP (Inlet Air Pressure) sensor.

The reason for this is the fellows on the CX500-CX650 turbo forum needed replacement units for their turbo bikes and the Suzuki sensors were found to be acceptable alternatives. One of the fellows took an OEM PB sensor apart and found the manufacturing technique for joining the wires in the OEM PB sensor was less than adequate by today's standard(s). I have been using these in place of the OEM PB sensors for the past two years and all is well. Here is the CX500-CX650 turbo thread: https://cx500forum.com/forum/cx500-cx65 ... ement.html

I did a bench test comparison between the two sensors as follows:

The power source for this was 3 - AA batteries in series giving me 4.8 VDC. I hooked up a test circuit with multimeter. Here are my readings, read in 3 columns - inches of Hg/Suzuki sensor - VDC/Honda OEM PB sensor - VDC:

0 in Hg/3.45/3.63
5 in Hg/3.1/3.14
10 in Hg/2.65/2.59
15 in Hg/2.18/2.0
20 in Hg/1.73/1.43
25 in Hg/1.26/0.88

The table below is from the supplement:

Service manual supplement Test Vacuum specs for the PB (MAP) sensors, read in 5 columns; mm HG/in HG; kPa; kg/cm2; psi; output voltage:

100-3.94/13.73/0.14/1.93/2.85-3.50V
300-11.81/40.21/0.41/5.83/2.16-2.45V
500-19.68/66.69/0.68/9.67/1.21-1.40V

Have some readings to do regarding the kPa requirement for the Megasquirt.

Need to further look at the TPS signal and correlate it to what is needed for the Megasquirt. Need to correlate the calibration procedure in the supplement to the Megasquirt requirements.

More to follow. Cheers
 

Ansimp

Well-known member
CGW Supporter
Joined
Jan 27, 2013
Messages
9,837
Reaction score
4
Location
Brisbane Australia
In my opinion if you want to use a megasquirt I would use later model injectors and design a completely new efi system.
 

Rednaxs60

Well-known member
CGW Supporter
Joined
Dec 25, 2015
Messages
2,448
Reaction score
38
Location
Victoria, BC,
Tony - Thanks for the input. I bought a book on EFI a while back and intend to read it again to determine if I can get more insight into EFI.

I've looked at a lot of GW EFI projects and most have been done quite similar to and with the '85/'86 CFI components. I have considered upgrading the various components to modern day components, but not looking at changing the design.

I think the best way ahead with the Megasquirt ECU is to have modern components. Injectors are either high/low impedance. COP units instead of the OEM coils and plug wires. May look at a change in the spark units, but don't think it's required. The Megasquirt unit has pulse width modulation that can accommodate low impedance injectors. The MAP sensor comes as part of the Megasquirt unit. TPS is a linear rheostat, not a lot to be done with this. Changing out the trigger wheels for the crank and cam sensors may result in better tuning. The passive IAC system could be changed into an active system.

Once the Megasquirt is installed and the CFI system can be tuned, will have a better idea what might be beneficial to change/upgrade. Also need to keep the functionality of the travel computer and such.

I'm still learning about the EFI system and how everything interacts. Having a Megasquirt unit installed and being able to tune the system will be beneficial at the onset. I'm looking forward to seeing how the maps configure. Have never been this involved in any vehicle before.

I've mentioned in other threads that this bike is my retirement project to keep me busy and challenge the mind. I'm also a believer that it is better to do a little, see the results and determine what is the next step.

Don't know what the end result will be but the journey is going to be great.

Cheers
 

Ansimp

Well-known member
CGW Supporter
Joined
Jan 27, 2013
Messages
9,837
Reaction score
4
Location
Brisbane Australia
Your main problem Ernest is the Gl1200 is a very antiquated system and you would be much better off using 30+ years technologies. :good:
 

joedrum

Well-known member
CGW Supporter
Joined
Dec 8, 2009
Messages
19,374
Reaction score
4
Location
chuluota florida
Well FI seems to be as elusive as home bred carburetor modding ...in both cases it hard to swallow how many fail at getting the oldwing to perform right ...with either old tech or newest tech around in FI ...strange that it’s this way ...I try all the time to help others in SCC set up but it seems most just fade away and not realize any success ...Ernest I’m thinking you may get there and answer some ? On the FI projects ...
 

julimike54

Well-known member
CGW Supporter
Joined
Feb 1, 2015
Messages
2,362
Reaction score
6
Location
Euless, TX
Curious on the old versus new. Some point in the past you had mentioned you had used a scope to find the failure of a component. Wouldn't that need to be done to see how 'clean' the signals are for all of the old components, versus what a new parts signal might be? I know from my past work we were always fighting stray 'noise' on inputs to the equipment CPU. Sometimes it was the quality of the parts and other times it was design issues. Good luck on the project :popcorn:
 

Rednaxs60

Well-known member
CGW Supporter
Joined
Dec 25, 2015
Messages
2,448
Reaction score
38
Location
Victoria, BC,
Tony - "Your main problem Ernest is the Gl1200 is a very antiquated system and you would be much better off using 30+ years technologies." Hear you on this but I tend to disagree a bit. It's the component circuitry that I am inclined to look at. The MOSFET technology now being used in shunt/series RR units, and that units with the MOSFET technology are now the preferred, stator life extension/reduction in possible problems. The Honda OEM PB (MAP) sensor versus the newer Suzuki MAP sensor - the internal circuitry and how it is constructed is far superior to the older Honda PB sensors. The Megasquirt manufacturer and community has a preference for the GM high energy ignition (HEI) spark units. Expect the manufacturing to be better (perspective) and how it is controlled is different that could possibly result in better tuning of the system. The technology is old, parts are scarce, but having a component/part failure that has the bike sidelined is a good possibility. I'm looking into this upgrade/replacement as part of my keep the bike on the road program, and to frustrate my friends who always buy new.

Mike - the component I mentioned in another thread was the throttle position sensor (TPS). Had to replace a new aftermarket unit on the '85 LTD that I brought back from Ontario last year. Did some looking on the web and found a site that detailed oscilloscope graphs and how new does not always mean good. The oscilloscope graphs showed that a new TPS can have chatter in the output signal and this chatter causes all sorts of engine operation issues - BTDT. The article also mentioned that QA for these units is less than stellar.

Appreciate the input because it keeps me grounded, and focused on the way ahead. Too easy with this type of project to fall down the proverbial rabbit hole and get off track. I've given myself a 2 year time frame for this upgrade/replacement project so that I can address as many issues up front, get baseline data, and of course ride and tour - life's too short not to.

Cheers
 

Rednaxs60

Well-known member
CGW Supporter
Joined
Dec 25, 2015
Messages
2,448
Reaction score
38
Location
Victoria, BC,
Have a book on fuel injection, picked it up a couple of years back when I bought my '85 LTD. Got it out and reviewed it this afternoon - Motorcycle Fuel Injection Handbook. Lots of good information; however, the author, Adam Wade is not correct in his knowledge of Honda's early venture. He mentions and details information regarding the FI system on the CX500-CX650 turbo bikes, but does not mention anything about the '85/'85 FI Goldwings. He mentions that after completing the CX500-CX650 turbo bikes, Honda went home and fuel injection would not appear on another Honda-badged motorcycle until 1998 - not quite correct (page 108 of the book).

Have been looking into FI for some time, and now have the bike where it would be a sin not to look more seriously into the CFI system, and look at what is needed for longevity.

The aspect of trigger wheels and what these achieve is another area of FI that I'm not quite certain of. The increase in trigger wheel teeth gives finer tuning aspects. Need to determine what is needed to change the trigger wheels on the '85, the crank and cam wheels.

Going to be interesting going forward.

Cheers
 

joedrum

Well-known member
CGW Supporter
Joined
Dec 8, 2009
Messages
19,374
Reaction score
4
Location
chuluota florida
Yes trigger wheels are great ..on the c5 ignition 90 different things can be program per one revolution ..one of the reasons it’s the best ignition ever made for oldwings
 

Rednaxs60

Well-known member
CGW Supporter
Joined
Dec 25, 2015
Messages
2,448
Reaction score
38
Location
Victoria, BC,
[url=https://www.classicgoldwings.com/forum/viewtopic.php?p=209273#p209273:g7m7nek4 said:
Ansimp » April 13th, 2019, 4:25 pm[/url]":g7m7nek4]
In my opinion if you want to use a megasquirt I would use later model injectors and design a completely new efi system.

Have been reading the literature for the Megasquirt units, browsing the web for info on trigger wheels and exactly how these work and correlate the cam/crank signals.

The bones of the EFI system are there with the exception of the O2 sensor, and this can be worked around if needed. The injectors are fine and can be used with the OEM resistor pack, or the pulse width modulation (PWM) that Megasquirt is programmed with. Using high impedance injectors would probably be a good change. The TPS sensor can be reused, as can the spark igniters; however, the spark igniters can be changed for the 7/8 pin GM HEI igniters that apparently work well with the Megasquirt. COP units would also be beneficial, but there is a space limitation that I have to assess. The CFI covers don't allow for a lot of flexibility for injector orientation. Using COP units would eventually have me look at sequential injector/ignition.

The IAC circuit is presently a passive system, but most of the literature is for an active system for fast idle control. Going to an active system appears to work well with Megasquirt, and if so, the unit would probably fit where the OEM IAC valve placement is.

Matt at DIYAutoTune has mentioned that the 8 tooth crank trigger wheel can be used, as can one of the sensors on the cam. Would only have to disconnect one of the cam sensors (Gr/Gl) from the system. This would give me an 8/1 trigger wheel system.

The MAP sensor can be a single unit as installed, and the water/air sensors can be reused.

With real estate being at a premium under the shelter and such, would use the existing fuse panel relays.

The Megasquirt units support an ITB mode that uses the MAP sensor up to approximately 3000 RPM and then goes to a TPS-RPM mix much like what is presently being used by the OEM. This is good news as the system works well in this configuration/mode.

With all this want to keep the dash working and the travel computer.

The other interesting aspect of the Megasquirt unit(s) is the CAN Comms. Will be looking into this, but it is a secondary consideration.

The wiring would have to be looked at, but it should be a beneficial exercise. Think that Honda used a lot of wire that might not have been needed. A rework of what I have done could probably result in fewer wire runs as well. If a new wiring loom for the Megasquirt is used, I would make it so that it would be compatible with the OEM system just in case a reversal is needed. Need to find a good electrical wiring program that is easy for us older gents to use.

One issue is with the original grounding system. Lots of wires going forward mainly because the main ground points are all located at the front. Using selective ground buses would eliminate a lot of wire.

Cheers
 

Rednaxs60

Well-known member
CGW Supporter
Joined
Dec 25, 2015
Messages
2,448
Reaction score
38
Location
Victoria, BC,
More research done. Find the Megasquirt literature extremely detailed, and at this stage not necessary. The Megasquirt ECU manuals are good, as are the threads I've been reading and reading again.

I have been watching some YouTube videos regarding the Megasquirt setup, using the ITB mode and others. The fellow who is doing these, Andy Whittle, narrates and explains the video concept at my level of understanding.

Looked at some of the PDF files I have and have found the pin placement for the '85/'86 LTD/SE-i: View attachment Travel Computer-ECU Pin Schematic.pdf Since I am looking for an upgrade replacement ECU, this will be extremely beneficial. The schematics detail the connections and if newer components are integrated, should be able to use the existing wiring. Using the OEM wiring harness to connect a new/newer ECU unit should keep all other OEM components working.

I have been looking for a ready made Megasquirt unit, and fairly certain that I will be settling on the MS3 unit. It comes with an integral MAP sensor that would replace the dual OEM CFI setup. This unit is very flexible with regards to injector/coil firing. A lesser unit would be more work because the OEM CFI system does not leave you with an option of controlling just the fuel, you must do fuel and timing as a minimum.

Info on the 8 tooth crank trigger wheel, and the cam trigger wheel. From Clymers: This 8 tooth trigger wheel sends out 8 pulses per revolution. The ECU then triggers the fuel injector at the 4th pulse after a TDC pulse is received from the camshaft angle sensors. This information is used to control the ignition timing and fuel injection volume accurately. Good explanation of the correlation between the crank and cam sensors, and is a feel good indication that the 8/1 trigger wheel setup will work with the Megasquirt unit. The Megasquirt unit apparently only requires one cam shaft trigger, not the dual OEM setup.

The OEM fuel pressure requirements is for a static pressure (not operating) of 34 to 38 PSI. Operating pressure of 28 to 34 PSI. Had the injectors cleaned and flow tested at 40 PSI. Injectors flow rate was 65 ml each after cleaning. Most literature that I have read indicates that a constant fuel pressure of 40 PSI should be used, but the OEM system does not require this.

The '85/'86 FI bikes have a fast idle circuit, but the IAC system does not control it. The IAC system is a passive system that relies on manifold vacuum to activate the reed valves to draw additional air into the intake during cold start and at other times such as when the throttle plates are closed on deceleration. I have not found any information on how the OEM CFI fast idle works, but surmise that the ECU uses air/coolant temp to advance the timing at start, and uses coolant temperature to control this timing advance. I mention this because my '85 has a fast idle and as the engine gets warmer the idle will start to decrease until the engine is at operating temp and the idle is at the recommended setting. It is recommended to use a stepper motor with the Megasquirt that holds the throttle slightly open on cold start and closes as the engine comes up to operating temp.

Just a short update, but keeps me focused.

Cheers
 

Rednaxs60

Well-known member
CGW Supporter
Joined
Dec 25, 2015
Messages
2,448
Reaction score
38
Location
Victoria, BC,
Quick update. Received the book "Performance Fuel Injection Systems" by Matt Cramer and Jerry Hoffman. Have read through it as well as my other book "Motorcycle Fuel Injection Handbook" by Adam Wade. Between the two a lot of questions have been answered, but have some additional questions as well. In addition to these books, have been perusing the Megasquirt manuals and Megasquirt documentation.

I have put together spreadsheets for OEM ECU pin designations, and correlating these to the travel computer, and LCD dash. Doing this so that I can determine how the replacement may affect the overall system. Have to look into the "User Defined" pins of the Megasquirt to understand how these work.

The TPS, crank (Ns), and cam (Gr/Gl) sensors as fitted can all be used. Megasquirt documentation indicates that the two cam sensor arrangement is not supported by the Megasquirt unit - this is corroborated by Matt Cramer at DIYAutoTune as well (email correspondence), but that one can be used.

The crank sensor uses an 8 tooth wheel that can also be used, Matt Cramer has indicated that a 12 tooth would be better, but finding a 12 tooth wheel that would fit/work may be a quest onto itself.

The crank sensor is interesting as well. There are two mounting positions exactly 180 degrees apart. Either position can be used on the '85 1200 FI system with no change to any other parameter. I can vouch for this because the OEM sensor failed sometime before I bought the bike. It is the subject of a thread on Steve Saunders forum where the fix was to install a 1500 sensor in its place. The PO installed the 1500 sensor in the upper position by modifying the mounting screw bosses (removed some of the aluminum boss) so that the 1500 sensor was installed to coincide with the 8 tooth trigger wheel. I have since removed the 1500 sensor, and mounted a set of PG sensors from an '85 Aspencade. This PG sensor install has a sensor mounted in the upper and lower position (Honda used the same block for both models), but have only one of the sensors hooked into the CFI system. The second sensor is there in case the one being used fails, then its just a wiring change under the shelter, and I don't have to go into the timing belt area.

I will be doing an investigating into the engine timing to determine the position of #1 TDC and how it relates to the crank/cam position sensors. Should be able to determine the various other piston TDC positions as well.

The engine firing is 1-3-2-4. As is known the waste spark is for cylinders 1-2, and 3-4. The injectors are a bank to bank firing pattern for 1-3 and 2-4. My thoughts on this are to continue initially with the wasted spark system, but change the injectors to sequential firing. This will assist in reducing what goes out the back end.

I have been contemplating the OEM spark igniters, specifically how these work. The OEM spark igniters for the '85 LTD FI system are 4 wire units. There are three 12 volt wires and a ground. I have come to the realization that these spark igniters are a very specialized relay, much like a basic automotive relay. My understanding is that the ECU grounds the 12 VDC supplied battery voltage to ground and in doing so the spark igniter internal "coil" switch is closed, allowing power to flow through the primary winding of the coil(s). At some programmed time, the ground for the 12 VDC supplied battery voltage to the spark igniter is stopped/removed, and the internal "coil" switch is opened, sending the amassed power out through the secondary coil to the plugs.

The CFI system does not have an O2 sensor, and I have not fully looked into whether this is a show stopper at this point, but reading the Megasquirt literature indicates that this is not. The primary benefit from an O2 sensor, as I understand it, is for tuning at idle so will investigate this for install.

The '85 LTD FI system uses an ITB mode for operation. The initial mode is speed density because a MAP sensors (PBR/PBL) are used up to about 3000 RPM after which the mode reverts to an n-alpha mode (n for engine speed, and alpha for throttle angle). If this is not quite correct, then Honda has one intricate ECU in that it uses all the inputs and chooses the appropriate part of the 3D control maps to operate the engine based on an n-alpha mode. Be an interesting hack to look at the 3D control maps for the '85 LTD and '86 SE-i for someone in retirement.

A Readers Digest version of this can be found at: https://powersports.honda.com/Experienc ... c0812182ba

The OEM ECU has an integrated self diagnostic system that, if there is a fault in the electrical portion of the CFI system, the "Fuel System" dash fault light will come on and an error code or codes will be displayed on the ECU. I surmise this is the forerunner of the OBD scanners that are in use today. I have not determined how this will be integrated into a new ECU such as the Megasquirt.

I think I have my thoughts on this issue focused in the right direction. Still a lot to research to compile, and will be perusing the various threads for guidance. The Megasquirt documentation is starting to be a little less daunting as well.

More to follow.

Cheers
 
Top