GS1000G mods- big bore and fuel injection

[GS1000G Shopper]

Here are a couple pics of the mechanical advance unit, before and after modification to lock out the advance:

 

[ArttuH]

I have made a 2-part trigger wheel setup that shows as being machinable and will bolt on. I looked into the press-fit, but adding to the rear of the wheel makes it difficult if not impossible to machine.

Part 1 is a spacer that takes the place of the advance assembly. It has a notch in the bottom so it fits in to the crank opening. The top has 2x 8-32 holes in it so the wheel can be bolted on.

Part 2 is the 12-1 trigger wheel. I added 4 sets of holes so it can be mounted in various positions. Coupled with an adjustable Hall sensor, this would allow for a great deal of timing variation.

Looks good except one thing. Based on my experience 12-1 wheel may not work reliably on bike engines. The problem is sync up during cranking. If crank speed changes too much between two teeth controller firmware may get confused and not detect the missing tooth properly. This problem is pronounced with bike engines since they usually have quite light crank and high compression compared to car engines. 12-1 seems to be on the edge with this issue. It may work on some engines but give trouble on others.

 

[GS1000G Shopper]

Using E Machine shop, the 36-1 wheels are not possible in such a small diameter. Maybe a 24-1 like you have. I've been looking for a decent Hall sensor that will fit under the cover, and have yet to find one shorter than 1" or so. This really limits wheel size. A second concern I found when reading Honeywell technical papers on Hall sensors was that some need fairly wide/deep teeth to work.

My latest idea is to take a 1.5" or 2.0" diameter thick flat washer with a template (easily done on E Machine shop) and cut out the teeth area, and then notch the center opening so it will press onto the single tooth shaft. This would allow retention of the stock mechanical advance unit, once advance had been eliminated.

One thing I found when pulling the TBs off was that I did not fully tighten the ground lug for the relay powering the coil. This may have been an issue causing weak spark.

My fuel rail mod has now dried, so I can reinstall it and check for leaks, and then see if the spark is better with the relay ground lug tight.

 

[posplayr]

While I was looking at the advance unit today, I decided to see how the timing marks related to the piston movement. I puled the #4 plug, which is on the same cycle as #1, and rotated the crank with a long screwdriver in the #4 plug hole. When it was at about TDC, I checked the timing mark and saw I was close to the "T" mark for 1-4. I backed it up roughly 7 degrees and saw I was close to the "F" mark.

Now for the odd part- when I looked at the pickups, I thought 1-4 was the rear one. When I had the timing mark at the 1-4 "F", the tooth was pointing at the front pickup. If this is the case, then I've wired my ignition backwards. Fortunately, since I used a Weatherpack connector for this, it will be a simple matter to reverse the pickup wires at the connector so the proper wire goes to the Microsquirt.

I also drew a 24-1 wheel using the gear template on E Machine shop. It's 2" in diameter, and has a .785" center hole with a .08 x .17 cutout for the tooth. My first version had multiple slots 30 degrees apart. The nice thing is the cost estimate and analysis. No unmachineable areas of concern, and about $65 shipped.

I'll do the math and figure out where the slot for the tooth needs to be so 1-4 firing is properly aligned. I think there is some adjustability in the MS software for this.

At this point I don't remember how I determined it was the rear pickup that fired 1-4 on compression. I'm assuming I did it right but who knows. I vaguely remember checking it out on my bike rather than just assuming arbitrarily. Did you check that you were on a compression stroke at TDC?

http://www.thegsresources.com/_forum/showthread.php?t=168652&highlight=firing+order

 

[posplayr]

Aren't 1 & 4 on the same wasted spark cycle- when #1 is at TDC on compression, #4 is on TDC for exhaust, and vice-versa? The ignition will fire at the same time, hence the "wasted spark" on the exhaust cycle. I have your document here that has the graph with cycles and firing, this appears to be correct. I'm still working on this as far as the slot for the tooth.

I also was reading MS documentation and I see I needed a 24-2 wheel. Updated design is below with one slot for the tooth. Prices out @ $70 shipped.

As I said I don't remember, but you would need to pull #1 plug and crank until it blows your finger off. Then move to the TDC (on that same compression stroke) and see which pickup the rotor tab is closest to. That is the pickup firing the 1-4 coils.

I remember being a little surprised at the result of concluding it fired on the back and not front. But because the engine is going clockwise (looking at the right hand side), the back is before the front (with rotations through positive angle) so it kinda made sense, but it is kind of arbitrary as well because of the lost spark ambiguity.

 

[coombehouse]

When the T 1-4 mark on the base plate lines up with the timing mark the timing should align with the pickup at the 9 o clock position. From the picture of your welded up assembly you have assembled it 180 degrees out I think. Nick

 

[posplayr]

When the T 1-4 mark on the base plate lines up with the timing mark the timing should align with the pickup at the 9 o clock position. From the picture of your welded up assembly you have assembled it 180 degrees out I think. Nick

That makes since and why I marked the #1 compression 1-4 coils as being fired by the back (i.e. 9 o'clock) pickup.

 

[GS1000G Shopper]

AHA! That makes sense, I realize the insert can be flipped around.

 

[coombehouse]

AHA! That makes sense, I realize the insert can be flipped around. It's locked in place now, so I guess I get to make a 1-off wheel.

Here's a graphic I made up using the MS 36-1 example.

A eureka moment I think. It can only get easier now (hopefully)
Your new 24-2 wheel should work fine with the existing pickup but I guess it will need re positioning. Try mounting the pickup on a small plate which in turn can be fixed to the baseplate via slotted holes as you will most likely have to mess with the distance between the the teeth & the pickup. Too small a gap seems to be just as bad as too large. Use the logger function in Tunerstudio to see when you get a good tach pulse.
Carefull you don't break the magnet that sits underneath the pickup - easy to do as I found out twice!!
Nick

 

[coombehouse]

I think I've got it!

I was able to trim the weight arms & free & flip the toothed part of the advance assembly.

I made a 24-2 wheel with a second slot 180 degrees from the first one so I could run a straight line through my E Machine drawing, and also added a TDC line after offseting the slot by 17 degrees. I could then rotate it on screen 180 degrees (or any increment such as 17) and check the alignment. It looks like this should work. The best news is I can make 5 of them for under $100. I deleted the second slot and ended up with the drawing below.

Now I need to find a small Hall pickup.

The picture looks OK except for the positioning of the slot relative to the first tooth. The ms works by counting the teeth after gap & the software allow you to tell the number of degrees to tdc. From this it then works out advance etc in conjunction with the ignition table. My understanding is that it would be better to have it count some more teeth before it reaches the tdc tooth. I do stand to be corrected though but I have read the manual over & over again to try & get my head around it. It worked for me.
Ref the hall sensor, I don't understand you reluctance to use the OE vr sensor you already have. Most of the hall sensors seem to be the wrong shape & again the ms manual seems to favour a vr sensor in this application. Besides you already have 2 sensors to play with!
Nick

 

[GS1000G Shopper]

What should I look for as far as clearance for a press-fit? The shaft is like .581" according to my digital caliper.

 

[coombehouse]

If you look at the picture that I posted much earlier in the thread you will see that a 2.5inch diameter trigger wheel will work with the standard Suzuki sensor as long as it is moved to around the 6 to 7 o clock position & further outboard. There is more room here. Your 2 inch wheel will make things even easier.
If you are going to the press fit option then really the wheel needs to sit against a face to ensure it is mounted concentically. The wheel will only be around 4-5mm thick so it will be pretty tricky to mount without being creative. If you could machine the shaft so it had a shoulder it be much simpler. Something to think about.
For a press fit a hole which is 0.001 to 0.004" less than the shaft will give a good fit as long as you heat the trigger to about 270F & leave the shaft in the freezer overnight.
Just a final tip, have a look at www.trigger-wheels.com to see what the tooth profile should look like. Most OE wheels use this tooth profile too so it must be what is required.
Nick

 

[coombehouse]

The correct tooth profile is not splitting hairs. If you want a good job you have to do it right. It would be pointless to make a trigger that doesn't provide a good tach pulse because of a bad tooth profile but hey its your choice. As for costs they should be less to produce a squared off profile as its easier to machine conventionally. If its water jet or laser cut you can have anything you like. But like I said its your call, I can only pass on information that I was given when I was making a trigger which now works perfectly. It is quite possible that yours will too but I like to go with simple proven solutions.

 

[posplayr]

this is the 2.5" one?
http://trigger-wheels.com/store/contents/en-uk/d5.html[/url]

 

[GS1000G Shopper]

This will be the linked trigger wheel post from the table of contents so I can delete the older posts. Mega/Microsquirt is listed as working with a 4 cylinder 4 stroke engine with 12-1, 24-2, and 36-1 trigger wheels. In prior discussions, I learned that a 12-1 wheel may not be suitable for a bike engine. I also learned that the tips of the wheel should match the size of the pickup. The GS pickups are .081" thick. I wanted to make this as easy as is possible, so using eMachineShop software I designed a 24-2 wheel with .080" tips that would fit into the same space and use the stock VR pickups. A larger diameter wheel could be used, but it would require relocating the pickup.

GS VR Pickups

Caution: If you perform this modification, you cannot use your stock ignition again without replacing the signal generator/pickups.

The trigger wheel also only requires one pickup, the GS has two. I removed the front (2-3) pickup. Since the 80-81 GS1000 used three wires here (a common ground), you need to clip & pull the wires to the front pickup. I had already ran new shielded cable to the pickups due to trying the dual trigger setup. On later model GS1100 models, apparently a 4 wire setup was used. The point here is to eliminate any possible source of electrical interference. You do not want a second pickup generating an AC wave that doesn't go anywhere other than next to wires carrying the signal you want. The best thing to do would be to run new shielded two conductor cable to the rear pickup, but I don't see why the OEM wire would not work here. The front pickup is held in place by two screws, undo them, clip the wires as close to where they come out of the cable assembly, and then remove the pickup and magnet from the base plate. I don't recall exactly how they ran, so you may need to insulate where you clip the common ground wire that goes to the front pickup.

For this setup to work, the advance must be locked on the stock unit. I did so by welding the arms to the base.

Once this has been tested successfully, I will sell the wheels if anyone wants to use this setup.

Here are the 2.0" and 1.15" wheels next to each other:

UPDATE
I figured out how to make an affordable .80 square tooth small wheel. The OD of the circle is key. I was able to downsize the 2.3" one to 2.0". I also added a third 1.15" wheel that has the larger center. It would require the OEM tooth to be ground down.


Application in Tuner Studio for Microsquirt (MS2 Extra version 3.24)

The MS2 Extra page explaining this setup is here.

The software allows for several variations. Under Basic Setup- Tach input/Ignition settings:
Spark mode: Toothed wheel.

Under Basic setup- Trigger Wheel Settings:
Trigger wheel arrangement: Single wheel with missing tooth
Trigger Wheel Teeth: 24
Missing Teeth: 2
Tooth#1 Angle (deg BTDC): 107
Wheel speed: Crank wheel

Here is how I calculated the 107 degrees based upon using the wheel on the right in the above drawing:

The above page says to count the gaps and multiply by the degree per teeth, in this case 30 degrees since it is a 24 tooth wheel. Doing so gives us 105 degrees, but it is not exactly right since the tooth does not line up perfectly with the pickup.

When the OEM tooth is even with the 1-4 pickup, the #1 piston is 17 degrees BTDC. Since the 24-2 wheel mounts in this same position and the #1 tooth is straight up, the gear needed to rotate 17 degrees to show where it would be at TDC. After I did so using the machine shop software, I drew lines intersecting the pickup (horizontal), the vertical, and the #1 tooth (diagonal). I then started rotating the verticle line 1 degree at a time until it aligned with the #1 tooth. I moved it 17 degrees. 90+17= 107.

Another easier way of looking at this is to realize the #1 tooth is 90 degrees away from the pickup when the notch is even with the pickup. Adding the GS 17 degrees BTDC to that gives you 107 degrees.

Here's a graphic of how I verified the 107 degrees:


Here is a graphic with the above details with screencaps from Tuner Studio:

 

[GS1000G Shopper]

I've been in a holding pattern waiting for o-rings for the injectors. The ones I measured were slightly too large, so I've re-ordered some 7.8mm x 1.9mm ones and updated the thread where I linked them.

I noticed a slight oil leak at the front of the cylinders between the head and the cylinders and re-torqued the head today. I hope that will stop the leak.

Yesterday I ordered a pair of the small 24-2 trigger wheels. It should take about 11 days to get them. I'll post pics and results once they arrive.

UPDATE
I pulled the front (2-3) pickup and wiring from the shielded cable, plugged the Weatherpcak connector holes where it had been, and ran a ground wire to the cable shield.

 

[GS1000G Shopper]

The throttle bodies have been reinstalled. The fuel pump is now acting weird, it ran and then slowed down, the gauge was slow to come up, and then it pegged, so I unplugged the fuel pump and cranked the bike relieve pressure. No fuel leaks so far.

I reinstalled the advance unit the right way and set the MS software as "basic trigger" with 17 degrees. Got a few sputters and a backfire, so timing is closer. Also tried 0, 10, and 25 degrees with no results. The trigger wheel is due here by the end of the month.

 

[AJ]

Reading about the leaks you had with your fittings... Loctite PST. This stuff will seal threaded parts - even loosely threaded parts - and will stand up very well to gasoline and high pressure oils. You can also disassemble the parts, unlike epoxies (which usually don't handle gasoline too well).

 

[Steve]

Quite some time ago, you expressed some concern about your wiring size and temperature.

The red harness wire that goes to the ignition switch is too small in my opinion- it was warm when I was holding it. It looks like 18 gauge or so. I'd like to replace it with some high quality automotive wire of 16~14 gauge or so.

I just wanted to point out that if the RED wire is getting warm, then the ORANGE wire that goes from the ignition switch back to the fusebox will also get warm, as it is carrying the same current. If you upgrade your red wire, be sure to also upgrade the orange one.

A better solution, though, would be to add yet ANOTHER relay. Use the fusebox end of the orange wire to trigger the relay, have the output of the relay feed the fusebox.
Simple, somewhat elegant, and removes that large load from the contacts in the ignition switch.

.

 

[posplayr]

Quite some time ago, you expressed some concern about your wiring size and temperature.

I just wanted to point out that if the RED wire is getting warm, then the ORANGE wire that goes from the ignition switch back to the fusebox will also get warm, as it is carrying the same current. If you upgrade your red wire, be sure to also upgrade the orange one.

A better solution, though, would be to add yet ANOTHER relay. Use the fusebox end of the orange wire to trigger the relay, have the output of the relay feed the fusebox.
Simple, somewhat elegant, and removes that large load from the contacts in the ignition switch.

.

To complete this mod he should definitely get one of those easternbeaver electronic fuseboxes.

http://www.thegsresources.com/_forum/showpost.php?p=1637803&postcount=9