1985 GL1200 Limited ECU Replacement/Upgrade - Part 3 - Road Trials

Road trial yesterday went well. Adjusted several VE table cells to bring engine operation into line with the 14.7 air-fuel mixture. Hear the exhaust "burble" every now and then, looking into this. I have read that some OEMs have calibrated the engine tune in and around the cruising range so that the AFR is at or close to stoichiometric ratio of 14.7. I have done the same and the engine is operating well with this philosophy.

I looked into the exhaust "burble" (could be from the engine cylinder when the exhaust valve is opened) in respect to another engine tune parameter - RPMs per second (RPM/s). This parameter indicates a change in RPM/s that is either positive or negative, or spot on. I'm thinking that this would indicate a richer fuelling condition when positive and leaner fuelling condition when negative, and may be related to the "burble" I am hearing. This "burble" could be slight misfires when the RPM/s dips into the negative zone. I checked the RPM/s data in MegaLogViewer (MLV) and there were a few RPM/kPa cells in th negative zone. Added a bit more fuel in these cell areas. For cells that were too positive (not many), reduced in these cells. Thinking I should aim for a slightly positive RPM/s in all cells.

Understanding how the ECU engine tune applies these table cell values is necessary. Speeduino or any other ECU, interpolates operating values based on the table cell value in question, in conjunction with the table cell values that surround the table cell in question. This picture is an example of the VE (fuel) table cell values:


The cell value circled is 47 at an engine load of 84 kPa and 2000 RPM. The cell values bordering are used in interpolating the required fuel percentage. The average of the nine cell values in question is a VE of 45 - two points lower than in the table - leaner fuel mixture. Not only do you have to consider the cell value in question, but what is around it. In this case if the aggregate value of 45 is lean, some adjustment may be required. Another strategy as the engine tune progresses is to adjust the periphery cell(s) to minimize the change to the cell value in question, may be better than adjusting the cell value in question. The cell values outside the 9 cell values circled can also affect the cell value in question. Balancing act.

As an aside, I am noticing that the engine "pulls" better at the lower RPM ranges in a higher gear. Thinking to what I have read on other forums is that the original engine tune ignition and fuelling profiles were a significant compromise. It was quite easy to "lug" the engine with the original OEM CFI engine tune. The fix for this was to operate the engine above the 2500/2800 engine RPM - getting into the power range.

This may be because the engine tuning parameters I am using are different - good possibility. Tuning strategy may be similar to that from 40 years ago, but the toys (tools - software and such) are better, a lot more is available for us DIY backyard mechanics. Being able to adjust the various operating tables, and EFI component settings such as coils and injectors can't hurt. Let's not forget the "Oracle", the on-line resources.

Great learning experience. Cheers

Have the bike on the bike stand, like not having to work on the garage floor. The new Speeduino ECU clone arrived, do believe I mentioned this, a Spark Gap X4 from openlogicefi. Chose this unit instead of the WMTronics Oscelot purely from a budget perspective, both good aftermarket Speeduino ECUs.

Nice unit, well put together. It's the size of two decks of cards, go figure:


Starting the conversion from the Speeduino ECU, have to cut out the original CFI enclosure connector and install the Spark Gap X4 wiring harness connector. This ECU comes complete with case and everything needed to connect up to the OEM wiring harness:

The internals are good:

Getting back to the thread, Have to remove the OEM ECU enclosure connector and install the Spark Gap X4 wiring harness connector. A prevalent thought when it comes to older vintage GWs, is to replace the timing belts. For an EFI conversion/modification the slogan should be: "Psst, oh and by the way, GET THE WIRING CORRECT!!!". Nothing spoils your good efforts than having a wiring glitch.

Will finish the wiring harness connector today. Starting to think about another wiring upgrade, probably have too many wires than needed, not just for the CFI system, but overall.

Recommendation for an ECU change on an '85/'86 1200 GW FI model, get an aftermarket ECU that has all the features needed and more. This unit has a dedicated fuel pump connection, the other ECU required a pin reassignment in the tuning software. A dedicated tach reading output, the other ECU required wiring changes. An installed barometric sensor, the other ECU required wiring changes. Need to keep it as easy as possible.

The Speeduino interface board v0.4.4 that I started with is still available - inexpensive, but you will have to purchase the board components separately and install the components yourself. Did not come with a case. More options available today than 5 years ago when I bought the Speeduino interface board.

Getting at the new ECU connector. Takes some thought to figure out how to use the provided wiring harness connector:

Have to take the back off, the side that the wires go through, then feed the wire through, add pin connector and insert in the connector. Wire size greater than 18 gauge doesn't work well.

Ordered some 18 gauge wire to do some wiring this winter. Buying wire that is striped:

Buying two different colour batches because all too often we use the same colours again and again. It becomes hard to differentiate what goes where even if you have a good wiring diagram. It will be different wire marking than the OEM. Have wanted to do this for a long time.

Short update on "garage" progress. Have the wiring harness connector installed, will do one more confirmation of wiring connection. Have decided to work on the wiring changes needed.

The original ECU wiring harness is gone, the CFI system has been modified (for the better I think), and I use some of the original wiring harness connectors and wiring for other component connections. Having mentioned this, I will be changing the original wiring to suit the new configuration, no coming back from this precipice, and tidying up what is there. I will be modifying the wiring schematic in KiCad to incorporate any changes being done.

Thought about wire size while caging the ECU wiring harness connector. the original Speeduino ECU interface board uses 20/22 gauge wire between the interface board and OEM wiring harness where the wire gauge is a minimum 18 gauge. The new ECU uses 18 gauge wire directly from the OEM wiring harness. 16 gauge or larger wire would not fit in the new connector.

Two "power pigs", fuel pump and rad fan, use 18 gauge wire. I have been using 16 gauge and even 14 gauge wire depending on the requirement. I am convinced this is overkill, and takes up more space in an already space constrained environment. There are a few circuits that may require 14/16 gauge wire, but not many.

Having mentioned the above, will use 18 gauge wire for most if not all circuits.

Took a picture of the two Speeduino ECUs. I bought the one on the left in 2019. It came complete with the add-on chip for tach, fuel pump, and other connections, but no enclosure or wiring harness to connect to the OEM wiring harness, or microprocessor (Arduino Mega 2560 - bought separately). The new Speeduino ECU on the right, less expensive model with more features included such as dedicated fuel pump, rad fan, tach connections. A barometric sensor is installed on the board as is the Arduino Mega Pro microprocessor - very clean look, and it came with a case/enclosure. The new ECU cost less than the one I bought in 2019:


Cheers