1985 GL1200 Limited ECU Replacement/Upgrade

[Rat]

Don’t get me wrong .... it is fascinating ....

Gord(old and slow)Jones

[Rednaxs60]

Don’t get me wrong .... it is fascinating ....

Gord(old and slow)Jones

No offence taken. I enjoy the banter. I do have a carb bike but I'm leaving it alone, only enough brain power for one bike refurb. Cheers

[Rednaxs60]

Found this on Toehead's Sulfur Yellow 76: Megasquirted with TBI thread, a DSO138 oscilloscope. Have always wanted something like this for my tool box so I ordered an assembled one off Amazon.ca for $35.00 CDN delivered.

It is a DSO 138 Oscilloscope that will provide additional information for troubleshooting and determining what a sensor is putting out. Here is a picture of the unit:

IMG_1706.JPG

It has a 2.5 inch screen (on the diagonal). I purchased a power plug for it:

IMG_1707.JPG

This plug can be configured for several different voltages. The oscilloscope needs 9 VDC and the unit on the 9 VDC setting puts out 9.2 VDC - acceptable.

I wanted to test the scope so I tried it on some TPS units I had available. TPS units are notorious for lack of good QA at the factories and even when new can be faulty. This is a test shot:

IMG_1708.JPG

I wanted to see the voltage as I moved the TPS rheostat. I have done this with a voltmeter, but the scope is more precise and gives a better indication of the signal quality that a multimeter cannot duplicate; however, you use what you have.

The signal that is shown for this particular TPS unit was not bad. As I moved the rheostat, there were a couple of bumps going up to 5 VDC, but not as bad as a couple of other TPS units I have on hand. I was looking for the same voltage form through the entire range of the TPS. I had found a web site that showed a different wave form, rising from base to max voltage, but I could not duplicate it. Time and practice will help as well.

I'm looking forward to using this scope on the sensors on the 1200 to determine what each is doing, and making a case for it.

Cheers

[Rednaxs60]

Have been looking into the differences between the MS units, as well as perusing the various threads to determine what will be the most economical way ahead, and which on could be tweaked for specific upgrades. The one upgrade I think would be beneficial would be coil-on-plug (COP) in place of the wasted spark system. This has been done on an 1100 with a Microsquirt unit by Socrace. The bank-to-bank injection will be kept. I have also been perusing the various web sites and videos regarding this and the Microsquirt appears to be quite a powerhouse considering cost and features. It definitely has several advantages over the OEM ECU, but lets remember that the engineering and what Honda did back then is remarkable.

I have also read that the OEM injectors are oversize for the application. This is an interesting detail because the MS units allow you to tune the fuel injection and timing for better engine operation, and from what I can gather, compensate for this component.

The one question I have and have not been able to find the answer to is how the injector flow rate in cc (ml) per minute translates into what is required. The flow rate of the injectors for my bike is 65 ml/min (cc/min) at 40 PSI and this converts into approximately 6.2 lbs/hr. So with four of these, the flow rate in lbs/hr is 24.8 at 40 PSI. Don't know if I am correct, but it's the best I can come up with. Still searching for more info on this.

Thought I'd give a quick update.

[Rednaxs60]

Just when You thought it was safe to go back into the water. Post over on the CX500-CX650 turbo forum because of similarities in the CFI system. Was introduced to the Speeduino system. Have done an initial look and there are a similar EMS. Reminds me of the DOS wars back in the '80s. Microsoft versus Digital Research DOS. DRDOS was a much better and a more fully functioned DOS, but Microsoft did a better PR and selling job, hence the demise of DRDOS. However, in this case, both Speeduino and Megasquirt products are evolving so it's anyone's guess how this is going to go. Will be going over the Speeduino literature and the forums regarding this. Uses TunerStudio as well. Always good to have options.

[socrace]

I think you'll find the Microsquirt and Speeduino boxes will both cost about the same; about $400 fully assembled with cables, etc. The main disadvantage with the usquirt is that active development ended around 5 years ago, with the code and hardware being buggy for some things, although was able to find workarounds to get most of what I wanted.
The main disadvantage with the speeduino it that it hasn't been around very long (was not even around when I bought my usquirt). It does seem to be under active development, which is good, but their github has some basic sounding issues that are still cropping up. The speeduino in kit form looks to be about half the cost, so that is tempting. Just my random thoughts.
Bob D

[Rednaxs60]

Bob - was hoping you'd chime in. Have been doing a read into Speeduino and intend to do a comparison between MS. The version I'm looking at has the V0.4 board. It looks good on the surface. Have found that you can get this EMS out of the US, assembled in case from WTMtronics for $150.00 USD. Must include the exchange as well.

What is "github"? Tried to find various acronyms on the web but some are quite illusive.

As long as I can revert back to the OEM configuration either/or should be a good project. I have noticed in my perusing that the Megasquirt world has much more information available. I will be compiling the literature for the Speeduino EMS. More information is not always better.

I have my homework cut out for me. Thanks again for the input.

More to learn and then publish.

[socrace]

Github is a general internet meeting place for software developers, and is also the official host for Speeduino developers, their page is at https://github.com/noisymime/speeduino
There's a number of things about the stock gl1200 efi system that is just plain "very unique". Like the use of 2 map sensors, 2 cam sensors and the reed valves feeding the cold idle air system. Never seen anything like it anywhere else. From what I can tell, the 2 cam sensors are designed to trigger the 2 ignition coils alternately based on polarity of vr signals; very unusual. I doubt that speeduino (or any other ecu) can support these features as is, and I know the usquirt can't.
But, not really that hard to change these things out to something more conventional, if you want to use a more conventional (and tuneable) ecu.

[Rednaxs60]

Thanks. Going to visit the github site and do a bit of a read in.

The 1200 FI system is unique. It is a variant of the earlier CX500/CX650 turbo design. Why the two MAP sensors don't know. What I am going to do is route the vacuum hoses so that the right and left banks are married and output the combined vacuum to both sensors. I will be putting a vacuum gauge on the hoses as well to see how the vacuum numbers are. I had read on one of the forums where a fellow married the two cylinder banks, used one MAP sensor but split the output signal so that the ECU was still seeing two signals.

I have also contemplated that Honda engineers back in the day did a cut and paste of the circuitry regarding the MAP and cam sensors. Waste spark being easy off the crank sensor, and each bank being driven by a cam sensor and MAP sensor. Something to contemplate at 0300 when the best ideas get thought of.

Matt from DIYAutoTune mentioned that only one cam sensor would be needed to successfully operate the MS unit(s). I will be taking the front end off and check the position of the crank sensor and which cam sensor relates to it. I'm thinking it will be the Gr sensor (outboard one). I can leave the other installed just have it disconnected at the connector under the shelter.

The coils are actually fired by the ECU through igniters. You had mentioned in one of your posts that the 5 wire OEM igniters did not work with your setup but that the 4 wire igniters for the 1200 FI models would. I have looked at the HEI 7 pin igniters, and both MS and Speeduino can use these.

The MS units like to have an active IAC circuit, where as the Speeduino can apparently work with what is installed.

I am looking for info about using the 8 tooth crank trigger wheel and a single cam sensor with the Speeduino.

The Microsquirt has 2 ignition and 2 injector outputs, but as you know, there are a couple of other outputs that can be utilized for 2 more injector outputs. The Speeduino unit has 4 injector and 4 ignition outputs. Another difference that I have found is the Microsquirt/Megasquirt units can easily be configured for an ITB mode, but the Speeduino does not at this time - not that I have found yet. Most other functions seem to be similar.

Conventional and tuneable is good. Want reliability and drive ability. The newer ECUs use a 16X16 map whereas my vintage Honda CFI system apparently uses 4X4 and maybe 8X8 maps.

Cheers

[Rednaxs60]

Update required. Have done considerable review of the Megasquirt units and compared the functions to the OEM CFI system. Since I started this research, I was introduced to the Speeduino project. I have done a read in to this project, and the functionality of this Speeduino/Arduino system is quite good.

I will be comparing the two systems; however, I have to say the Speeduino project is intriguing, get to “tinker” along the way.

More to follow. Cheers

[Rednaxs60]

Another update in my quest. Have been perusing the literature and videos regarding the Speeduino project as it would apply to my requirement. The more I delve into the Speeduino project, the more impressed I am with it. I have costed the Speeduino against the Microsquirt/Megasquirt units and the Speeduino product has more functionality and ability for future expansion/modifications for less cost. In this regard, I am going to direct my efforts towards the Speeduino unit.

I am feeling more confident in my understanding of the installed OEM components and how these interrelate. I will be making up test pigtails to plug into the various sensors to determine what the output/input signals are as installed. Should be of benefit down the road.

The Speeduino unit uses Tuner Studio the same way that the Megasquirt units do. As Socrace has mentioned, the Speeduino project is still being actively developed. Both products are touted as learning platforms, and as such, the onus is on the person using the product.

The Speeduino unit will account for the passive OEM IAC system. My research has indicated that it will do a good job of interfacing with the existing OEM wiring system so I will lose no existing functionality.

I have found a good YouTube video regarding using trigger wheels with the Speeduino (probably applies to the MS as well). The fellow doing the video makes a good case for doing a bench test of the new system before install. On my hit list.

Found another video with pearls of wisdom. This fellows thoughts are that one must READ, and learn everything possible - within reason I would think, about your existing system, or what the system components will be, the new system software and hardware and how you will connect the two. To me this is the preparation stage, and I have mentioned that the better the preparation phase, the better chance for success.

It was mentioned that I look into the injector" dead time". Apparently this is a useful number that has an impact on idle, low RPM and warm up. Another mystery to be solved.

I have re-read the CFI self-diagnostic system in the OEM Service Manual Supplement for my bike. Honda did a great job on this ECU programming.

For the MAP (PBR/PBL) sensors, if one fails the other operates the system as per normal. If both fail, ECU is programmed to use Ne-Oth map for both ignition and injection timing. Engine will operate normally with only one MAP sensor.

The cam (Gr/Gl) sensors are similar in that if one fails theECU uses the signal from the other sensor to operate the system as per normal. If both cam sensors fail, fuel injection and timing are stopped, time for a tow truck.

The crank (Ns) sensor fails, engine is dead.

The TPS sensor fails, Ne-Pb map is used to control fuel injection and ignition timing. If MAP (PB) sensors also fail, fuel discharge and ignition timing are both fixed - get home scenario.

Fuel injectors, if both fail - open circuit/shorted signal wire, fuel injection and ignition timing are both stopped. If a single injector has a short circuit, engine will operate, but an error code will be generated.

Interesting how Honda has the ECU configured to use an alternate signal for normal engine operation. The crank (Ns) sensor is a single install, but it can be installed either above or below the crank trigger and the engine operates as per normal. The upper and lower mounting position is exactly 180 degrees apart.

Honda did a good job of setting up the CFI system. Pleasantly surprised and get a better understanding of the CFI system each time I read the OEM manuals (for whatever reason).

More to follow. Cheers

[Rednaxs60]

It was mentioned over on the Speeduino forum that I may have to operate a parallel EMS system in order to keep the travel computer (ETC) and LCD dash functioning. This is a good point and has made me delve more into the interrelationship between the ECU/LCD dash/ETC.

There must be a comparator circuit in the OEM ECU that checks for a signal to the various CFI components. If all is well the "FUEL SYSTEM" (check engine) light goes out - no more power to it, or no ground. The system allows for starting of the engine before this check is completed, done this many times. This indicates to me that this "comparator" circuit is not critical to the engine operation, but only indicates to the operator the state of the CFI system.

I am checking the wiring schematic and the OEM supplement to determine the relationships between the three units. The ETC does mention the "FUEL SYSTEM" light when troubleshooting a couple of issues.

The CFI system components that are being monitored are the MAP (PB) sensors, TPS, ECU, IAT, CLT, Injectors, crank (Ns), and cam (Gr/Gl) sensors. 11 components in all.

Been a while since I did any programming, but seems to me these are if/then type statements. With the newer technologies such as the Arduino, a small comparator circuit may be able to be put together to take care of this issue. It could also be as easy (nothing is easy or free) as a signal to the light from the new ECU.

If a comparator circuit was put together, and found to emulate the OEM ECU circuit, a small LED circuit could be added. This could be a numerical indication instead of the flashing lights that are presently used. Be a good diagnostic of the system should something go astray.

Looking at the dash/ECU/ETC schematics. The dash "FUEL SYSTEM" indicator light is connected to ECU-14 directly. It appears that the ECU grounds this indicator when it is lit and removes the ground when all is well. Removing the OEM ECU should only result in this light never coming on.

For the ETC. ETC-15 is connected to the wiring that supplies battery voltage input to the ECU, but is external of the ECU - good. ETC-5 is connected to 1/3 injector signal - external of the ECU as well - good.

I'm still thinking that there is a possibility that the system will work as normal when the ECU is changed; however, stranger things have happened.

The ETC takes care of the LCD dash digital tach. The dash tach indication can be displayed as digital and analogue, or just digital - switch is on the ETC. The ETC can switch the speed from Kms/hr to MPH - default for Canada is Kms/hr - nice feature when touring. The ETC fuel management circuits can also be changed from litres to USG.

Have also been looking into the injector "dead time", and how to get this information from the existing install. Have an AC/DC clamp meter - can read DC amps - and will look at this further.

As an aside, when Honda developed and designed the 10th anniversary fuel injected model, it forgot to include reverse. Don't know how Honda missed this considering the bevy of features that are included that newer GWs don't have to this day.

Keeping my thoughts focused.

Cheers

[Rednaxs60]

After having done a comparison of the Megasquirt family and the Speeduino project, will be going with the Speeduino. Have ordered the unit, comes in at approximately $250.00 CDN assembled and delivered. Have found in the past that committing to a project only comes with some monetary outlay. Have Tuner Studio and Mega Logger, licence bought.

I have had a good look at the OEM wiring for the ECU, travel computer (ETC) and LCD dash. The LCD dash and ECU only have the "FUEL SYSTEM" light in common - from the self-diagnostic check. The ECU grounds this light on start, and removes the ground if all is well. With the new ECU, no ground, no light.

The ETC looks to use 1/3 injector pulse for the tach signal, and from this takes care of all the other machinations it does, including the LCD dash readouts.

I have done a pinout comparison and I'm thinking I will be able to install the new ECU and have the OEM system operate as per the OEM ECU.

I have also been thinking about the offset from this project. I intend to learn how the Arduino is programmed and in doing so, think that there will be other smaller projects that can be done using this as the platform. May not be for the bike, but for other things and just for giggles maybe.

Over the next couple of weeks will be doing some work on the bike, and during this time intend to install some inspection adapters (wiring) to get readings to use to tune the system through Tuner Studio. Have the information on the MAP sensors - installed newer Suzuki sensors. Want to find the injector latency, signal from the crank and cam VR sensors. The Speeduino unit has a 2 channel VR signal conditioner installed to convert the signal to a square wave. Look at the TPS signal, IAT and CLT as well.

The Speeduino will work with the passive OEM IAC system, space is at a premium on these FI models so this is good.

Doing engine work this fall and will install an O2 sensor boss in the exhaust at this time. Should be able to not use an O2 sensor, but it is recommended. Wonder how Honda did a CFI system in the '80s without one and had a good idle - go figure. O2 sensors are mostly a byproduct of the emission requirements, and from what I have read, used for idle only.

Intend to bench test this system using an 8/1 cam/crank trigger wheel arrangement.

Have given myself a 2 year window for this project. Lots to do and learn.

Cheers

[Rednaxs60]

I just watched a video where the fellow switched from a Megasquirt unit to the Speeduino. Very good information, took notes. At the end of the video, he mentioned that he used a crank sensor with the Megasquirt, but installed a trigger wheel in the old distributor for the Speeduino. He then went and operated the engine on the Megasquirt, but at the same time had the Speeduino running in a bench test mode for lack of better vernacular, and tuned the Speeduino to match the car operation. Once he had the Speeduino set up, he did the swap.

I would expect that he used a bench model to simulate the coils and injectors, and would piggy back with the various other sensors.

The reason I mention this is that I have two crank sensors installed, exactly 180 degrees apart, use one and the second is a spare just in case so I don't have to remove the front end to replace a sensor. Since the engine will operate the same with either sensor, BTDT, I could use the second sensor as a signal to the Speeduino. There is also two cam sensors, only need to use one, and the same scenario, leave one hooked to the OEM ECU for engine operation, use the second for an input into the Speeduino for test/trial. Use the bench test model to imitate coil/injector loads.

This has merit, should not impact on the engine operation and aid in the transfer to the new ECU. Once the Speeduino is set up, use the same sensors as per the trial.

Interesting concept and would probably work with the Megasquirt as well.

Cheers

[Rat]

Trying .... valiantly ....to follow along ....

Gord