1985 GL1200 Limited ECU Replacement/Upgrade - Part 2

[Rednaxs60]

Have some time to get the new sensors installed and switch back to the 8 tooth trigger wheel. Broken main spring on garage door, won't go up. New door on order. Think that the door and parts/pieces should last longer than 25 or so years.

Can raise the door with the bike lift when I get the changeover done.

 

[Rednaxs60]

Today was template day. I have made the brackets and installed the cam and crank Hall effect sensors on the test engine, looks good. I made an engine timing pointer and installed, as well as marking the OEM trigger wheel with what I hope are good degree marks, really only need #1 TDC position.

This is the engine timing pointer:

This is the crank Hall effect sensor installed in test engine:

This is the cam Hall effect sensor:

This is the camshaft sensor covers without spacer. This is a better fit:

These Hall effect sensors are good up to 125 degree C. Will be transferring these to the GW this weekend.

Going to install the 8 tooth OEM crankshaft trigger wheel and check timing with this configuration. The issue at the start was the camshaft VR sensor and the signal from it to the ECU.

Using a cam sensor also provides the ECU provides the ECU with a better engine cycle input. With just a crank signal, the ECU estimates the engine cycle over 360 degrees; however, using a second cam input signal provides the ECU with a 720 degree (full cycle) input. As has been mentioned on a different forum, the second cam input signal can be used for full sequential fuelling, or COP installs, and the more sensor inputs to the ECU can be beneficial for tuning.

Will tackle the wiring connections as well. Have an idea what I will do.

If this works out as I hope it will, should be able to have the engine started early this coming week for tuning.

 

[pidjones]

I'd be interested in seeing the Hall probe scope traces. Years ago (the '90s), a researcher in our lab was having timimg issues with a pixel selector for histogramed data. The sweep speed on his CRT was causing timing issues. I pointed out that he was using 741 op-amps, which don't have a real fast slew rate. I loaned him an AD509 and he was in business! Lesson learned - make sure all of your circuitry is fast enough.

 

[Rednaxs60]

Here are the traces:

Cam trigger wheel at 3100 RPM:

Crank OEM 8 tooth trigger wheel - 1100 RPM - with hall sensor:

Crank OEM 8 tooth trigger wheel - 3100 RPM - with hall sensor:


Think there is a bit of "flyback". There is a trigger filter in the tuning software that should compensate for this. The wiring for the bench testing was not that good, but expect the wiring to be better on the bike.

 

[Rednaxs60]

Initially thought there was an issue with the OEM trigger wheel so went with the 36-1 crank trigger wheel and no cam signal. Learned quite a bit in the past year, and with research and a better understanding of the FI system components, realize it was the cam VR sensor. A new VR sensor might have fit the bill, but EFI system conversions are using Hall Effect sensors to get away from using VR conditioning boards.

VR conditioning boards can be inverting or non-inverting. Have to know the specs for the VR conditioning you are using, this is a question asked and answered regularly on the Speeduino forum.

Will see how it works this coming week.

This is the PDF spec sheet:

 

[Rednaxs60]

Wiring update. Still using the OEM wiring harness to connect all the components. Keeps extra wiring to a minimum. Have a wiring project I will do for the accessories I have added and will use the spare wiring harness wires for the new harness. Like to have the various colour coded wires instead of the same colours for multiple uses.

Will be using the original PB sensor connections for the crank and cam signals. These connections have a ground, power and signal wire. Will use the VR crank sensor wiring for the O2 sensor connections, has a ground and signal wire.

If all goes well, will make permanent.

Have to drain the engine coolant to change the crank trigger wheel and new sensor, and give me better access for the timing add-ons.

 

[Rednaxs60]

Progress with the new sensors. Have made brackets and installed the Hall effect sensors on the engine. Use the OEM 8 tooth crank trigger wheel.

Can use the camshaft cover without a spacer. New timing pointer installed, engine timed quite nicely at 188 degrees. Started and stayed idling, little rough, without throttle assist. Forgot I had the rad off so only operated the engine for a few minutes. Time to put it back together.

Engine spun up quite nicely, no timing kickback, getting better at this. As soon was the air was out of the system, idled not too bad. Checked for sync losses, and none showing.

The crank pulleys are held on with the original bolt, but will be changed for the alternator setup.

Have an issue with operating in the garage. Hard to raise the door, new one on order, main spring broke. Will work on this.

Here are pics of the install.

Camshaft hall sensor installed:

Crankshaft Hall effect sensor installed:

OEM 8 tooth crank trigger wheel and timing pointer installed:

Using the PB sensor connections for the Hall sensors because these connections have 5VDC, ground and signal wires. The O2 sensor has been connected through the now redundant crank VR sensor connection.

Will eventually use a baro sensor, and have decided on where to connect it. Can use the original camshaft sensor connection.

 

[Rednaxs60]

Pidjones: Have the bike back together and looked at the composite log for the cam and crank signals. Pleased with the results, and no sync losses:

Don't think it can get any cleaner. Top being cam signal, bottom being crank signal. Feel like I'm back where I started about a year ago. Only difference being I have a better understanding now than then.

Now to take on the initial tuning and look into the MAP reading. Has to be something I'm missing.

 

[pidjones]

Nice, clean signals. What are you using for throttle position? Seems that was a weakness on 80's GMs.

 

[Rednaxs60]

TPS is the OEM TPS. Have found some on the web that could work besides the go to TPS from an early model Honda Civic/Prelude:

Checked the rotation of the throttle assembly yesterday and it has a rotation of approximately 90 degrees. The sensors in the picture can be anywhere from 90 degrees rotation to 360. Since I am using a 5 volt signal should get one with a mechanical stroke as close to a 90 degree rotation as possible to maximize the voltage signal.

Would have to modify the turning arm to fit the throttle linkage. Don't think it necessary to have a spring on the TPS arm.

Have searched for info on setting the TPS to indicate "CLOSED" and "OPEN" throttle plates. Apparently Honda uses a voltage from the TPS of approximately 0.47 VDC to indicate close throttle position.

May order, fit and try one of these. Not too expensive, around $30.00 CDN to the door. Here's an eBay web site for the 90 degree: Rotary Hall Angle Non-contact 0-90 Degree Angular Torque Rotation Sensor 0-5V | eBay

 

[pidjones]

Apparently Honda uses a voltage from the TPS of approximately 0.47 VDC to indicate close throttle position.

Yes, as I remember.
Spring loading may be needed to remove any backlash.

 

[Rednaxs60]

Have ordered one of those TPS Hall sensors.

 

[Rednaxs60]

Sequential Fuel Injection (SFI) - secondary rotational input required. This is generally the cam shaft rotation keeping the crank rotation as primary. Have mentioned that the second rotational input is for sequential fuel injection, coil on plug (COP) application, and for better engine cycle resolution. Without this second rotational input, the ECU determines the firing of the engine for one crankshaft rotation the starts again. Having the second rotation sensor allows the ECU to determine engine firing over a complete cycle of 720 degrees.

Regarding SFI, the benefit is mostly at idle and low powers for emissions and fuel economy. SFI attempts to inject fuel into the cylinder when the intake valve just opens and possibly until the intake valve closes, engine tuning dictates this.

In a sequential system, the injectors are fired individually and timed with the cylinders. The ECU controls the pulse with of each injector so that the fuel injector is only flowing fuel to coincide with the opening/closing of the cylinder intake valve.

The version of Tuner Studio that I have has a table for up to four cylinders under Sequential Fuel Trim – Injector Cyl 1-4 Trim.

Semi-sequential and batch firing are the same, and follow the engine cylinder timing.

Reader's Digest version. More to learn, understand and apply.

 

[joedrum]

Hmmm good topic here ...on my FI SET UP FROM THE 80S ON MY VETTS ...THEY OPERATE ON BATCH FIRE TO THE MOTOR ON THE 350 CI GEN 1 MOTOR ...AT FIRST I THOUGHT THIS TO BE BAD SET UP AND LIKE YOI SAID USES MORE FUEL AT IDLE AND LOW RPM ...BUT O ALSO BOUGHT A AFTER MARKET FIRST PERFORMACE INTAKE WHERE THE LONG CURVED INTAKE RUNNERS ARE CONNECTED ..4 SETS OF 2 RUNNERS EACH ..THIS ALLOWED TO CUT A SLOT BETWEEN THE THE RUNNERS ALL THE WAY INSIDE THE RUNNERS WITH A DENTAL DRILL VERY SMALL SLOT ...PLUS OPEN UP THE THE GASKET BETWEEN RUNNERS AT THE HEAD ...THIS WAS A GREAT JOEDRUM TYPE MOD NEVER DONE BEFORE ..BY ANYONE ...THE PROBLEM WITH C4 VETTS WITH TUNE PORT INJECTION WAS THEY WOULD SHUT DOWN AFTER ABOUT 4500 RPM...AS THERE SUPER CHARGER LIKE VELOSITY STACK RAM AIR SET WOULD GO THE OTHER WAY AND CLOSE OFF MORE AIR ..BUT NOW ON MY VETTS THE SLOT IN THE RUNNERS AND THE OPEN INKE GASKETS AT HEAD AND BATCH FIRE ...TURED ALL THIS INTO A SHARED TWO RUNNER SET UP THAT FEEDS AIR WAY PAST THE 4500RPM OF STOCK VETTS AND IVE GOT MY VETTS RUNNING LIKE THE HOOCH BIKE ...POWER STARTS JUST OFF IDLE TO RED LINE EASY ...THE THIN SLOT MOD I MADE KEEPS ALL THE ADVANTAGES WITH LONG TUBE RUNNERS LOW RPM HUGE TORQUE THAT HAS NO CEILING RPM ...PROBABLY THE. BEST MUSCLE SET UP SPORTS CAR IVE EVER CREATED ..JUST LOVE IT ..THE HOOCH POWER LIVES ON IN A 4 WHEEL SET UP....

 

[Rednaxs60]

Joe - good mod(s).

Update to my sequential fuel injection (SFI) post. SFI injection tuning is to try to inject fuel just before or as the intake valve opens on the intake stroke. Will have to read more on the SFI fuel trim; however, there is another table in TS that is applicable to SFI and I think the other. modes as well, that of Injector close angles.

I was wondering what this is all about and it was clarified on the other forum. This table can be set to determine when the fuel injector(s) should cease flowing fuel to coincide with the closing of the intake valve for the cylinder in question, or to adjust the fuel injector(s) fuel flow to coincide with the operation of the intake valve. Instead of adjusting when to start the fuel injection process, you are adjusting the ending of the fuel injection process and backing it up. Interesting concept.

 

[Rednaxs60]

Interesting day for the Speeduino. Can't get the bike out of the garage so implemented a sequential fuel injection install.

Temporary wiring, and using Speeduino pins 1,2,3,5. In the Engine Constraints, kept 4 squirts per cycle and alternating. Injector layout is Sequential, Injector pairing is 1+3, 2+4. Kept wasted spark as Spark Output Mode, did try Sequential but the engine did not like it.

Had to adjust the VE table settings. Good learning experience.

 

[Rednaxs60]

Soldered the wiring for the Speeduino to interface connector between the OEM harness and Speeduino. Interesting how the wires for the 40 pin IDC connector are so small, but the wiring harness has much larger wires, but it works. This change appears to help in smoothing out the engine idle.

There is a noticeable engine RPM change when the rad fan turns on. This is my next investigative project, the why?

The injectors being setup in sequential appears to be a benefit to the install.

Purchasing the components for the small add-on baro board. Thinking that using the MPX4250 component will be good. Small footprint allowing me to put this circuit board in the enclosure with the ECU and it has been recommended on other threads.

The new resistor pack is working well, a 6 ohm resistor per injector. Will mount where the OEM resistor pack is.

The engine stumble did not show its ugly head yesterday.

The MAP reading is still around 65/70 kPa, but the VE cell values are set at 40, progress on the VE mapping. The spark table cell values are set at 15, but can go down to 13.

Found an older thread that discusses the differences between Alternating and Simultaneous injector pulse widths on the Speeduino forum. Discusses the differences regarding Speeduino and the tuning software. Don't understand everything that's mentioned but will. Sequential ignition will probably not be done, but will look into benefits and what is needed to achieve this.

Looking at the other features/functions in engine tuning software for down the road.

 

[pidjones]

There is a noticeable engine RPM change when the rad fan turns on. This is my next investigative project, the why?

I would carefully monitor the 12V. That fan is a pretty good load (as are the horns).

 

[Rednaxs60]

Thanks - I have installed an 1800 rad fan in place of the 1200. Was hoping it is less of a power pig.

 

[Rednaxs60]

Probably getting ahead of myself, but as mentioned considering sequential ignition to complement the sequential fuel injection. Doing this up front. The Speeduino can accommodate 4 ignition circuits. This was done on an 1100 EFI conversion by socrace over on the NGW forum. He used a megasquirt ECU that only had 2 dedicated ignition circuits and had to reconfigure two other circuits to suit. Socrace found that power consumption is reduced and the engine seemed to operate better. His conversion and using sequential fuel and ignition is the only EFI conversion I have found that has done this.

Socrace used ignition coils from a 2003 Ford F-150 (lots of these on eBay and inexpensive). These apparently have a dwell time of 1.5 ms and a spark discharge of 2.0 ms. This would free up space under the shelter where the coils are (not a deciding factor - maybe a better spot for the vacuum mixing block then under the seat) and finalize the ECU upgrade - not much else to do.

Have been looking at the space available and it may be tight.

As has been mentioned over the years by many, in for a penny, in for a pound.