[Rednaxs60]

@JHolland - thanks for the reply. The second reason for this project is to further and renew my knowledge of electronics and get into testing components - struggle with this as well, lots to learn at this old age.

Yesterday's road trial was primarily engine timing, no VE changes. Adjusted COP dwell a point or two. Engine tune is coming into focus. Looked at the tables and looked at the 3D maps. The ignition table is looking good, but the VE table is rustic to say the least. Looked at the data log and it confirms where there is VE table tuning to do. Smoothed the table (expect further changes) to get rid of some of the rich feeling swings. Will be testing in the near future. I modified the VE table with the thought that as the engine RPM goes from a low RPM to a higher RPM, the VE table cell values should increase, and when the engine load goes from a low to high load, the VE Table cell values should increase (adjusted the cell values manually - table smoothing still not quite understood).

Here's a pic of the tables in 3D view:

VE-Spark Tables 11 Mar 2024.jpg (82.81 KiB) Viewed 1821 times

Noticed that the TPS signal is not what I would expect. For example, engine RPM 2800 or so, no TPS signal. Will be investigating.

[Rednaxs60]

Continuing with the road trials. Have upgraded the ignition timing (spark) table and working on adjusting it. Ve Table is a work in progress. The interesting aspect of tuning the engine is that I can feel and hear when there is a small misfire. I have noted that when this occurs, it is generally a too rich VE setting. Pull over, adjust the specific cell or just a cell or two around the offending VE cell, back on the road again, try to replicate the scenario and check - same for timing.

The reason for this post is I am seeing an anomaly in MLV at the 3000/3100 RPM range and engine load of 70 kPa. Engine operation is good, probably needs more tweaking. These pictures are from the road trial 16 Mar 2024. This is picture of the MLV trace that appears not too bad:

15 Mar 2024 - Normal.jpg (66.29 KiB) Viewed 1726 times

Further review and this happens, RPM stays the same and everything else goes south:

15 Mar 2024 - Wierd.jpg (76.13 KiB) Viewed 1726 times

This is the VE table to match the going south:

MLV VE Table.jpg (89.69 KiB) Viewed 1726 times

Still have more tuning to do, but this weirdness is puzzling. This repeats itself on a random basis during the same road trial and from other road trials. Only happens at the 3000/3100 RPM range and 70 kPa. Has not happened at any other combination of RPM/engine load that I have noticed. Any thoughts on this issue? I will be chewing on this as it does not make any sense and considering I'm only adjusting the spark/VE tables.

I timed one incident and it went on for some 6 seconds before correcting itself, and heading north to the 70 kPa engine load range. The MLV graph settings are not what I use, put the data log on MLV this morning just to get the pictures. Question - MLV graph setting "MAT" - is this the IAT reading, think so just want to confirm.

Tune and data log:

Archive.zip

(3.34 MiB) Downloaded 93 times

[Rednaxs60]

Road trials are progressing well. VE table has been adjusted to what I think is a good representation of what the end product will be. The spark (ignition timing) table is being adjusted as well - spent the afternoon adjusting the spark table, a point at a time. Realize the AFR table cannot be populated until the VE and spark tables are well defined. After the AFR table is defined, can enable the O2 sensor. Fuel consumption appears to be getting much better.

Take a data log for each road trial, but changes are very quick to do, feel most of the issues with timing and VE soon after change(s) are made. Need to do a road trial and just drive, leave settings/adjustments alone to see what has actually been achieved. Tomorrow may be a good day for this road trial.

View the plugs after each road trial, quick and dirty way to assess what is happening. The plugs seem to indicate that the engine tune may be richer than needed, timing indication is pretty good.

Have a three week hiatus coming up, going to be helping sister do some home renovations, a break is always good. Will have a clean data log with no changes for viewing.

Not ready to post a tune and data log of the road trials where changes are made, better to post a clean tune and respective data log.

When reviewing the data logs, look for consistency in the traces, specifically that each trace is linear, not spiking or bouncing around. Indicates to me that the engine tune is coming together.

Cheers

[Rednaxs60]

Spent today and yesterday just riding and logging. Wanted "clean" data logs to compare and analyze. RF is at 6.2 ms. TS calculated 6.4, but I was using a lower RF in the high 5.8 ms range. Want to get to the 6.4 RF calculated value, should be able to. Will be looking at these logs over the next three weeks when I'm at my sister's doing some home renovations. Plugs are being looked at after each trial. Provide some good information.

Here are the tunes and data logs for 1 Apr 2024, and 2 Apr 2024:

Archive_1_Apr_2024.zip

(2.07 MiB) Downloaded 55 times

Archive_2_Apr_2024.zip

(2.07 MiB) Downloaded 73 times

[Rednaxs60]

Lots happening and getting in the way of my road trials. 5 weeks in Ontario helping my sister and her hubby get their house ready for sale. Heading back for the month of July for the final work period. One of those years.

Corresponding with another fellow on how best to keep these older GW FI models on the road as well as the earlier '82/'83 cx500/650 turbo FI models.

I have had time to reflect on what I have done and what is necessary to be done. I have mentioned on several occasions that the 1200 GW FI Models have all the required FI components installed, but the main issue is failure of the OEM ECU - no replacements available except on the used market. Hence the ECU replacement project.

It is necessary to realize that there is always a requirement for some change because of a project like this.

Reflecting back on what I have done and need to do is that there is only two changes that needed to be done. There is an install of a WBO2 sensor, and changing the VR camshaft sensors to a single Hall Effect sensor. The new ECU will not use two camshaft sensor signals. It is advantageous to change the crankshaft VR sensor to a Hall Effect sensor so that the signal is digital, not necessary but nice to do. Changing the crankshaft VR sensor to a Hall Effect sensor allowed me to use the original crankshaft VR sensor wiring to connect the new WBO2 install to the new ECU without adding/modifying the OEM wiring harness.

I changed the spark units (coil drivers) to a single Bosch 211 coil driver as it has a proven track record with the new ECU, but not necessary. Had the engine operating with the original spark units. When I changed to sequential ignition needed the extra connections of the Bosch 211 coil driver.

Sequential fuel and ignition was not needed, wasted spark and paired injectors will work just as well.

Will be reverting to paired injectors to confirm that the reason the dash RPM indication is not registering is because I changed from paired injection to sequential.

It is necessary to let the OEM CFI operation go and concentrate on the new install from a first principle aspect. It is not always possible to duplicate the OEM FI system with the new.

Thought I'd update this thread. Cheers

[Rednaxs60]

Back from helping my sister and her husband do a 30 year house makeover. Spent 9 weeks at the house in Ontario. Won't be doing something like this for a very long time.

Perusing all my posts regarding this project, lots of good information I had not noticed before. Reading the Speeduino manual and Wiki as well. Dasq had mentioned that I had the LC and MC circuit understanding not quite right. He was spot on. Understand these now after much reading.

Looking into the tach, and it appears the dash digital tach has never worked since the start. Most tachs apparently use a 12 VDC, but would think a digital tach uses a 5 VDC signal. Searching for info on this and other forums, and on the internet, have found quite a bit. As has been mentioned, hard to duplicate the original OEM system, have to find new way ahead with the new ECU, system components, and engine tune/operation.

There is a video on YouTube regarding the v0.3 and v0.4 bards: https://www.youtube.com/watch?v=-9FqR0KGgEc

Have the idle timing set to 10 degrees, and using a timing spread from 10 degrees at idle to 45 degrees out at full power. As mentioned VE/ timing repeat, monitor effect on. AFR.

@PSIG mentioned way back, engine operating MAP appears to 60 kPa and above, will focus on this area. Did some garage tuning, VE and spark, at a few constant RPMs. At 2500 RPM adjusted VE and timing and got the MAP reading in the lower 50 kPa region.

The FP does not work from the ECU, have it hardwired with a switch. Not very convenient, but works. Changed relay, no joy. Will continue to investigate.

A short update, going to keep soldiering on. Cheers

[Rednaxs60]

Continue to review my threads and realize how much I missed along the way. Hindsight is always 20/20. @PSIG queried a while back, the requirement for two camshaft sensors on the Honda GL1200 Gold Wing CFI system, and how these as installed could be used with the Speeduino ECU. Interesting question.

These sensors were used in conjunction with the crankshaft sensor to accurately control engine ignition timing and fuel injection. The camshaft sensors provided the CFI ECU with the engine phase and an indication when a cylinder is at TDC. The camshaft GR (right) sensor controls fuel injectors 1/3, camshaft sensor GL (left) controls fuel injectors 2/4. The 8 tooth crankshaft trigger wheel sends 8 pulses to the CFI ECU. When the CFI ECU receives a TDC signal from one of the camshaft sensors, the CFI ECU triggers the respective pair of fuel injectors on the 4th pulse from the crankshaft trigger wheel after the TDC signal is received. Interesting concept compared to how new aftermarket ECUs work.

I have other thoughts on the Honda CFI system, but an issue I find very interesting is how versatile the Speeduino Project ECUs are. Reading the various forum threads is proof of this. This open source project is only limited by a person's imagination, skill set and want to do something.

Cheers

[PSIG]

@PSIG queried a while back, the requirement for two camshaft sensors on the Honda GL1200 Gold Wing CFI system, and how these as installed could be used with the Speeduino ECU.

Yes, and for clarity, how they function and their timing was not to how Honda used it, but what Speeduino has to work with. Big difference. As Speeduino can time the cylinders at any point in the cycle, we just need to know when the components triggered what signal, in order to configure Speeduino. E.g., 1 or 2 wheels, number of teeth, any missing, signals at TDC or other angle, evenly spaced at 180, 720 or 360°, Hall or VR, etc... what we have to work with, stock.

In most cases, we only need to know functional info of components (e.g., signals & their timing, active or passive coils, etc), and while the original operation, wiring diagrams, etc, may help us determine that, how Honda used it is otherwise mostly irrelevant. I hope that makes sense for projects pursuing Speeduino setup and function with any components.

[Rednaxs60]

Thanks for the reply. Not saying that the Honda way should be used, just mentioning that Honda did some different things. Have gotten way from. the Honda CFI mindset, doesn't work with Speeduino. Might have been easier starting from scratch.

[Rednaxs60]

Have been progressing road trials. Using Tune Analyze Live to help bring the engine tune into focus. It has helped, not the end game. Reviewing my threads again, always something I missed or scanned quickly. The early days are fun reading, life's too short to be too serious.

January 2023 first starts after installing Speeduino, tach worked. The engine settings were paired injectors, 4 squirts, and wasted spark. Coil dwell at 3.5 ms, spark duration 0.5 ms. Have been reading up on recommended ways to get a tach signal working, a relay install, connect a fuel injector in parallel. Won't be reverting any time soon to the OEM configuration.

Looked at the settings of the tune from those days, definitely not very good. VE (fuel) and spark tables well off the mark. Injector open time 1.5 ms, now at 0.8. Port versus throttle body fuel injection, really did not know what these meant, do now.

Have to admit I was naive and "wet behind the ears". There, I said it.

Have been thinking about engine performance and fuel economy. Highway cruising, solo or two up 45 to 50 IMPG at approximately 100 KPH. This would indicate to me that Honda operated the engine with a lean air-fuel mixture, maybe not. Will work on the engine tune to get this type of fuel economy with the Speeduino installed.

Cheers.