GS1000G mods- big bore and fuel injection

[ArttuH]

I have been doing some reading on what is called Individual Throttle Body mode or ITB. It appears most Megasquirt setups use speed density (SD), while others use Alpha-N (A-N). ITB is supposed to be a hybrid mode from the two that can be beneficial for a setup like mine. The process involves road testing with an operable SD or A-N tune, so I am not at that point just yet, but I have set up the Microsquirt per the below link to output on pin 7 (fast idle) to a temporary lamp at the handlebars when the engine is at 90 kpa. This is supposed to be similar to a dyno run, and the recorded data can then be analyzed and used for ITB mode. I've ordered a serial to Bluetooth adapter so I can log data on my phone via the Tuner Studio android app called Shadow Logger.

I have never tried the ITB mode as I mainly play with boosted engines and the ITB mode doesn't support them. But even plain Alpha-N works very well on this kind engines. And you can combine MAP value on it by selecting "multiply MAP" option which basically scales the injection amount by MAP value. Speed density usually don't work too well with individual throttle bodies. Since the A-N is much simpler to tune I would start with it and consider moving to the ITB mode once the engine is running well and all the bugs are ironed out.

[GS1000G Shopper]

The timing tab arrived yesterday and I installed it this afternoon. I set the bike at #1 TDC using the original timing mark, then installed the new plate and trigger wheel. I set the tab about where it needed to be and used a sharpie to mark the plate. Alignment off the bike was kind of tricky, I used a socket that had the same OD as the ID of the aluminum plate, mounted the socket in a vice, and used some foam tape wrapped around the trigger wheel shaft to center it in the socket. I then marked the plate and I drilled a couple of 1/8" holes near the ends of the tab and plate and used two small screws with nuts and lockwashers to mount it to the aluminum plate.

The end result was really close- all it took was a slight turn of the plate once installed and I had the marked tooth aligned on 0. Note that the center of each tooth aligns with the 0, 15, 30, and 45 degree marks, which verifies the timing marks. Each tooth is 15 degrees since 360 degrees / 24 teeth= 15 degrees per teeth.

[GS1000G Shopper]

Everything is back together and now I'm working through the intricacies of Microsquirt tuning. The high idle remains, it was about 2500 when I started it, and then when I checked the timing with a light, I saw I was 5 degrees off. I changed my tooth #1 angle and this gave me correct timing, but it also gave me another 500+ RPM at idle.

I'm reading and trying to see what is causing the high idle. I thought I had misaligned one of the old throttle bodies, but it was replaced. I don't see any vacuum leaks, all hoses are in place and nothing is dry or cracked. It looks like I'll have to tweak the VE table.

My Bluetooth adapter arrived so I can ride & log now once I get past the high idle.

[JTGS850GL]

Are you planning on using a WB A/F meter to do the tuning? If not, you might want to consider one. I've tuned a few programmable fuel injection systems (Haltech and Megaquirt) in the past and wouldn't even consider doing it without one. Don't bother with anything then a Wide Band meter. The narrow band meters are worthless and only are accurate at the switch point around stoichiometric. Anything leaner or richer is just a wild guess since the transfer curve has very high slopes on either side.

Here's a link to a decent and relatively inexpensive WB O2 meter:

http://wbo2.com/

If you have decent soldering skills then you can order the DIY kit and save some bucks.

[ArttuH]

Everything is back together and now I'm working through the intricacies of Microsquirt tuning. The high idle remains, it was about 2500 when I started it, and then when I checked the timing with a light, I saw I was 5 degrees off. I changed my tooth #1 angle and this gave me correct timing, but it also gave me another 500+ RPM at idle.

I'm reading and trying to see what is causing the high idle. I thought I had misaligned one of the old throttle bodies, but it was replaced. I don't see any vacuum leaks, all hoses are in place and nothing is dry or cracked. It looks like I'll have to tweak the VE table.

That high idle means the engine must get too much air from somewhere. Have you already adjusted the throttle stop (idle adjustment) to minimum? As you noticed the ignition timing affects to idle rpm quite strongly. The VE table has only secondary effect to the idle rpm. You want to tune the VE table so that you get good AFR and steady running at the idle. Actual rpm adjustment is then done by idle screw. Of course if your VE table is badly off it will drop rpm too but if your idle screw is correct then you can't get the rpm from correct idle to 3000rpm range by VE table.

[GS1000G Shopper]

Yes, I have had a WB O2 sensor from day 1. It's a Bosch sensor with a BPSX controller. It is linked on page 1. UPDATE, using a Spartan2 unit instead.

I tested my MAP sensor and found it was bad. I think this is the root of my idle problem. I managed to break an injector connector and nick an injector o-ring, so I'm on parts hold for the connector. The fuel leak from the injector ruined the AMP connector seal, so I had to replace it as well.

I'll update once it is back together.

[ruili]

just a question on the ignition coils
instead of using a coil driver and ignition coil wouldn't it be cheaper and better to just use lsx coils? hotter spark, and built-in drivers. or would they use up too much power?

[GS1000G Shopper]

instead of using a coil driver and ignition coil wouldn't it be cheaper and better to just use lsx coils? hotter spark, and built-in drivers. or would they use up too much power?

I didn't check the power draw for the LSX coils, but did look them up for this reply.
The specs I found for an MSD LSx coil were:
Resistance: .57
Inductance: 5.8

At 3 ms of dwell, each coil would draw 5.37 amps. The Neon super coil at the same dwell draws about 4.5 amps. The Megamanual page below says the LSx coils like 5.6~5.8 ms of dwell for optimum performance. That's about 9 amps each using the above specs.

Since these are individual coils, you'd be firing 2 of them with any given output from the Microsquirt. One would be a "wasted" spark so it would not have the higher current draw, but even if it were half you're looking at over 13 amps.

In one of my datalogs I made while the engine was running, I noted the volts were consistently 13.8, so the Neon super coil is not killing the charging system. The bike has the Compufire 3 phase regulator rectifier recommended on this site.

Another problem with doing this would be their size & having to use 4. There's only so much room under the frame tubes. You then have 4x as much wiring, to include heavier gauge power feed, plus there is still added wiring via two recommended capacitors.

I'd have to find the dimensions for them, apparently the LS2 coils are narrower but perhaps taller.

As for cost, I paid about $37 for the Neon super coil and another $10 for the wiring pigtail (plus shipping). A few $ could be saved by making your own connector. I used my (free) stock ignitor box with a $4 Radio Shack perfboard, and an $18 pair of BIP373 transistor kits, so this was $61 total (plus shipping). The cheapest new LS2 coil I saw on eBay was almost $20, that's over $75 for the coils, plus there is another $14 for 4 connectors with terminals, so that's $89+ (plus shipping) and then you still need to add capacitors for optimum performance.

In summary:
Cheaper? No.
Better? Not likely. More draw required for the improved output, plus added wiring and then possibly not enough room. I'd say the LSx coils would likely get the nod for an automotive project where there is more room and more amperage available.

At this point, as the saying goes, the coil isn't broken, so I'm not going to fix it.

[reliant_turbo]

do you know the specs of a stock neon coil?

stock neon coils have no issue lighting off heavily boosted neon motors.... like 650hp on up.

it may decrease your current and still run the engine just fine.

i bought a setup off of arttu as well but am working on my turbo plymouth reliant right now so the bike still waits for EFI. i have my gsxr throttle bodies and the pump controller but i still need to get a microsquirt and fuel pump.

i just realized that i have a stock neon coil and pigtail sitting on a parts neon in my yard that im slowly stealing things from lol. less stuff to buy.

Brian

[GS1000G Shopper]

No, I don't have that data. I suspect it would be fine. I like the size of the smaller Subaru quad output coil. Since I'm not building a race motor, I think any of these modern coils would be great. The Emgo coils + new caps and wires made a night & day difference on my 1980 GS1000. I may have ended up with a stock coil if I had not found the super coil on Rock Auto for $37.

I'll likely use the Subaru coil on the GSX1100 project I'm planning now. The parts bike makes an easy test mule for sizing things up.

Glad to hear you'll be undertaking this project at some point, I hope my documentation of the good, the bad, and the ugly here will help.

[reliant_turbo]

Glad to hear you'll be undertaking this project at some point, I hope my documentation of the good, the bad, and the ugly here will help.

yup. ill probably use a few tips from this thread. some stuff ill go on previous megasquirt experience.

thanks

Brian

[GS1000G Shopper]

In a project of this scope, errors will be made. I think that is apparent by this thread. I look at them as being good that I made them so others will not do so. The ongoing issue I had with the bike when it started was a high idle. This is indicative of a vacuum leak. When I looked over the throttle bodies, vacuum hoses, and boots I could not find one. I did find the MAP sensor was bad, but this was not the cause of the idle problem. Due to the rail leaking by the pressure gauge, I had to remove the TBs again. I've made a change to the fuel parts diagram and uploaded it. basically I went from a pipe thread setup that did not allow for precise alignment of the pressure gauge to a pipe/AN setting that will allow me to tighten the AN fitting and align the gauge as needed.

I had ordered new o-rings for the carb/TB boots a while back, getting two sets so I could do the GSX-G when it had some engine problems. I forgot to install them on this bike, a really dumb move on my part. I found out about this when I pulled the #1 boot and saw there was no o-ring present. #2~4 were all hard and dried up. I found the brand new o-rings still in the parts envelopes.

At this point, I'm waiting for the revised fuel rail parts to get here so I can put it back together and see if this was the cause of the high idle.

This project while frustrating at times has been a good learning experience. In my down time, I have laid out a Microsquirt setup for the GSX-G and am starting to acquire parts for it when I can find them cheap. The main problem with it was the airbox, it was likely designed by the guy that designed the V65 Sabre carburetors after he was fired by H***a. I ended up cutting it off since it cannot be removed with the engine in the bike. Having a parts bike gives me a test bed for part fitment and other measurements. Using pod filters frees up enough room for the fuel pump, Microsquirt, new fuse panel, quad coil, and the other electronics such as pump control and wideband O2 control. I've identified the wiring connectors and can make an almost plug & play harness. Once I get this bike operable 100% I will start a new thread for that project.

[reliant_turbo]

i wasnt knocking your trial and error stuff... ill just be using chrysler and gm sensors for megasquirt cause ive used them before with a megasquirt chrysler turbo project. thats what i meant. :-)

[GS1000G Shopper]

I didn't think you were- no problem. I wanted to post an update since I think I have found the main idle problem.

[Thumper88]

Special thanks goes to:
GS1000G Shopper for his detailed trials and tribulations, which gave me a head start on making decisions on how to do my conversion.
ArttuH for his excellent trigger setup, which is the perfect bolt on trigger setup for our application--and saved me time from developing my own, well worth the money!
coombehouse for his base tune, it was a good starting point to develop my own tune off of (and continue to develop).

Rode my bike for 40 miles without issue (VE auto tune running), starts on the first crank, and responds well enough to pull a wheelie.