[ATOMonkey]

Good work steve! Keep the updates coming.

[91talon]

Indeed. You get many kudos from me for blazing a path not many have been brave enough to take, let alone see to completion.

[TheSteve]

Thanks for the kind words guys, I keep getting to the point of wanting to give up but every time I do I have another "breakthrough" and find a way to keep moving forward with it. Next time I play with it I'll try resyncing the throttles. That would be awesome if thats the only thing causing it to not work correctly..

[TheSteve]

Well I tried to sync the throttles just a minute ago. At an ok idle, set at 1500rpm it shows ZERO vacuum on any of the cylinders. ZERO. Absolutely none. Continues showing none till about 3000rpm where the gauge starts to move. This doesn't make any ****ing sense whatsoever. And if you touch the throttle the engine races to 6000rpm instantly, and sits there "idling". Plenty of vacuum then, and theyre all synced nicely too. I don't get it. I synced them before at an idle of 1500 give or take and it worked fine. Now I have no vacuum at idle, it wants to die, and if I touch the gas even a little it revs high and wont come down till you kill it. The wideband says the AFR is perfect the entire time. I'm getting real sick of this things ****. I've got other **** to fix, and this doesn't seem to be going anywhere anytime soon.

As usual, once it gets moving it runs pretty well but again while riding at low rpm it occasionally tries to start racing against my will. I'm open to all suggestions at this point, because I'm finally starting to lose patience. Why is there no vacuum til 3000 rpm? Why just now and not when I did this last time? I checked for vacuum leaks and didnt find any.

The obvious answer to having no vacuum is theres too much air getting in; too much throttle. But if I take any throttle away by turning down the idle screw it dies now. Let it sit at its shaky idle long enough (if it doesnt die) and it shoots up to 6-7k rpm, and pulls a vacuum. Again sounds like too much air, but the air fuel ratios are spot on the entire time. No excessive smoke, no backfiring, no lean misfires.

I dont get it.

[ericej2344]

If im getting this, you switched to the 2 cyl. mode to get it to start easier. Now its not idling correctly, and your rpm is jumping way to high. Have you tried going back to the 8 cyl. mode with the HEI module. I know it may be allot of work, but I think it is worth investigating. One more question, is your bike a 600 or 750. I've got a 750 and don't know weather or not to look for 600 or 750 throttle bodies.

[TheSteve]

Sort of. The reason for the switch from 8 to 2 was because the injectors take 1ms to actually spray after recieving the pulse, so pulsing twice as often as they have to results in the injectors fluttering around uselessly and not spraying. The only time this is noticed in this case is while cranking. The same would happen if you tried to run giant injectors on this bike, theyd be open for so little time they probably wouldnt really open at all. Again I've got an idea I need to try...

I haven't yet tried the 8cyl mode again. While it does make it run, it wont start so I'm really back at square one if I try that. I could go back just for a quick sanity check though.

The idea I have now is a simpler one. Before I was running off the signal gen because the coils "couldn't" work. Now that I know that isnt true I'm on a single coil. If I can do both coils I'm getting one tach pulse per ignition event. In other words, 4 cyl mode should work. The hard part is getting both coils to talk to MS without interfering causing the other coil to fire when its not supposed to. A pair of diodes between the coils should do the trick. Each coil could then send a signal OUT thru that wire, but nothing can go IN and fire the coil accidentally.

I have a 600. I'd probably stick with the 600 TBs anyway, the 750s and 1000s are so large I think that you'll have trouble fitting them into the Kat intake boots. The GSXR throttles (even the 600s) are so much larger than the Kat carbs I don't think youll notice much going to the 750 or 1k throttles instead of the 600. Any of them are serious upgrades airflow-wise over the Katana hardware.

[ATOMonkey]

Well I tried to sync the throttles just a minute ago. At an ok idle, set at 1500rpm it shows ZERO vacuum on any of the cylinders. ZERO. Absolutely none. Continues showing none till about 3000rpm where the gauge starts to move. This doesn't make any ****ing sense whatsoever. And if you touch the throttle the engine races to 6000rpm instantly, and sits there "idling". Plenty of vacuum then, and theyre all synced nicely too. I don't get it. I synced them before at an idle of 1500 give or take and it worked fine. Now I have no vacuum at idle, it wants to die, and if I touch the gas even a little it revs high and wont come down till you kill it. The wideband says the AFR is perfect the entire time. I'm getting real sick of this things ****. I've got other **** to fix, and this doesn't seem to be going anywhere anytime soon.

As usual, once it gets moving it runs pretty well but again while riding at low rpm it occasionally tries to start racing against my will. I'm open to all suggestions at this point, because I'm finally starting to lose patience. Why is there no vacuum til 3000 rpm? Why just now and not when I did this last time? I checked for vacuum leaks and didnt find any.

The obvious answer to having no vacuum is theres too much air getting in; too much throttle. But if I take any throttle away by turning down the idle screw it dies now. Let it sit at its shaky idle long enough (if it doesnt die) and it shoots up to 6-7k rpm, and pulls a vacuum. Again sounds like too much air, but the air fuel ratios are spot on the entire time. No excessive smoke, no backfiring, no lean misfires.

I dont get it.

Mine is doing the exact same thing right now after I messed around with the idle screw. I suspect there is either a vacuum leak, or something goofy with the linkage. I haven't checked the vacuum on the carbs, because I don't have a sync tool either.

Vacuum in a throttle body is generated by the incoming air accelerating through the venturi (choke point). That is why there is ALWAYS vacuum no matter what the throttle position. Especially in a CV carb.

I suspect the EFI throttle bodies have a similar venturi shape to them.

If the air is not accelerating through the carb/throttle, then it must be coming from somewhere else. Once you get up in the RPM range, you overcome your leak.

Now, I just don't know where the heck this leak is coming from, since all I did was turn the idle knob.

Is there any kind of idle air bypass on these throttles? Do you have a manual or any section views of the throttles?

Fatten up the AFR at idle a couple of points and see what that does.

[TheSteve]

Its not really much of a venturi shape, its almost a straight cylinder. It tapers a tiny bit at the end though, but only by I think 3mm. Theres no idle valve on this. GSXR OEM has a fast idle cam that sort of holds the throttles open moreso that normal. I removed it before installing them on the bike. I'm about to go outside and mess with it for a little bit, I'll let you know how it goes. I'll also doublecheck for vac leaks.

[Theeun_Holy]

you are hardcore dude!

I just read the whole thread from beginning to end, couldn't peel my eyes away!
...and since I'm not supposed be browsing the internet here at work all day...
I just scored myself a verbal warning from the boss man, woops

very cool stuff, will be waiting eagerly for updates.

[TheSteve]

Hehe glad you enjoyed it. I got the wiring changes done today but had to leave for work immediately afterward. Didnt even have a chance to test it out... We'll see what happens tomorrow.

[TheSteve]

I figured that since the kat can run off one coil as a tach input, it would also run off two for a true 4 cyl mode. As mentioned above, simply T'ing off the coil wires will cause both coils to fire at the same time. This is bad, and you dont want it to happen for obvious reasons. So I grabbed some diodes and drew up a quick schematic:



The two diodes prevent power from flowing back into the coils, but it allows a signal out. The downfall of this is that it causes no voltage to flow into the MS. So you need to wire in a 12 volt "pull up" using a 1kohm resistor. This allows a tiny amout of current to bring the line up to 12 volts when a coil is not firing. When the coil fires, it quickly drains that line faster than the resistor can refill it. So the MS sees the signal go 12-0-12-0-12-0 etc, where with no pullup it would be nothing but zero volts ever. I might have been able to reverse the diodes so the line fills with power when the coil isnt firing, but depending on the rate of drain the MS has on the tach line it may not have worked. My diagram looks very spread out, but all of the wires you need are on the 6 pin plug on the CDI box. Power is orange-white, coil one is white, coil 2 is black-yellow. So to do this I didnt even add wires, just bridged a few by soldering diodes and resistors in.

And it worked! Bike now idles smoothly down to a previously unseen 800 RPM! Tach signal is smooth and accurate. Bike revs freely and comes down to rest at idle with no hiccups.

I still have NO idea why 2 cyl mode doesnt work. It worked for the turbo guy in the link above. It ran my bike too, but wouldnt idle right. I don't get it, but I'm not wasting anymore time worrying about it. It works now, in the correct 4 cyl mode. Now I need an oil temp sensor and this thing should be ready for DD duty.

EDIT: Wow that picture came up small. If anyone really needs to see it in full detail I'll resize it.

[jwilliams]

Hello,

To TheSteve: so I have a fuel pump, I have set of TBs from the GSXR 600, I have access to any of the GM sensors (dad works as a parts guy at a GM dealership). I was thinking about doing this conversion anyhow so the closer you get the more I want to try it. Just need to get a MS and LC1 and some time.

So I was originally thinking of using the MicroSquirt due to it's size etc but $400 is fair amount. So the next best thing is a kit for a 2.2 + PCB3 for around $260.

Questions:
1. How difficult of an assembly job is the MS? I have some experience and have done my own electronics and have a pretty good soldering iron.
2. About how long did it take to assemble? I think you mention this in the other thread but while I am typing I am lazy and didn't look it up
3. How difficult is it to find a good place on the Kat for the larger MegaSquirt vs the MicroSquirt?

I hope to at some point also use it for ignition as well so I figured I wold like the newest PCB etc.

Thoughts?

Thanks, Jeff

[TheSteve]

I think I need to update that first post to reflect everything I've learned since then. To start, I found the MS 2.2 to be very easy to build. Theres not too much stuff on the board and with the exception of the clock circuit for the CPU, everything is spaced out pretty well. That one area is a little packed though. The 3.0 board I did NOT enjoy building. This was probably mostly due to my soldering iron having a dull tip at the time. The 3.0 parts are very close together, but still doable if you take your time and have a good iron. I've built 3 of the 2.2 boards now, I think it takes me about 2 1/2 to 3 hours to complete one. The 3.0 board I've only done once and I think it took about 4 1/2 with a short break in there. I did get each done in a single night though, so its not too bad. You can always stop and pick up where you left off later.

The full size megasquirt fits perfectly between the frame rails under the passenger seat. You can't fit the plastic "hood" over the connector, but nothing is squeezed. Its a perfect fit, you just need to make a way for it to mount to the bike. I welded a little tray under there, but with all the extra bolt holes back there I'm sure you'll find something of use. The microsquirt is made for motorcycles and small watercraft so it'll fit pretty much anywhere that isnt too hot (keep both away from the engine). The added benefit of the micro is its actually watertight, which comes in handy if you think the bike will see a lot of water.

Now on to the MS choice between the full size models. In my first post, I say I bought a MS1 2.2 with no plans of ignition control. I then said that due to the Kat engine having a VR sensor the 2.2 won't work without the HEI module. The MS site says no CDI coils for tach triggering, but I think the name "CDI" is a mismoner on the Katanas. Theyre standard coils and you CAN trigger off of them. So my outlook has changed. You don't need VR decoding, the 2.2 board is fine for this. You don't need all the fancy idle controls and the numerous inputs and outputs of the 3.0. I'm reversing my stance and will now recommend people use the 2.2 PCB on the Katana.

For the CPU, it still depends on what you want to do. The MS2 is a nice unit and can do some cool stuff. It also has more accurate fuel control and native spark timing. The MS1 with the firmware I'm running can also easily accomplish spark control even on the dual coil Kat. Basically, the MS2 has all the spark outputs on the chip where the MS1Extra remaps the function of the LEDs to control spark. So instead of the LED blinking during an injection event, the LED is wired into the coil and is told to flash when the coil should fire. Very cool "backdoor" to doing it, and it works just the same as the MS2 in practice. Bottom line is the CPU is your call. Both will handle this engine just fine. I'd do the MS1 either way in this case, but will definitely throw the MS2 a bone in that the increased accuracy would be very nice to have sometimes. Theres a feature comparison table somewhere out there but its hiding from me right now.

Personally, now that I figured it all out the MS1 2.2 seems the way to go unless you want the smaller, watertight microsquirt. The 3.0 board is pretty much unnecessary, and the MS2 processor is very nice but not the only way to do this. If I ever convert mine to control ignition my current setup will handle it with only two wires coming from the LEDs going to a pair of transistors controlling the coils.

Congrats on starting the conversion! Yours is bound to go smoother than mine and I look forward to seeing how ours differ in the end. Good luck! Let me know when you're ready to add spark in (DONT do it from the start, too much initial troubleshooting) and I'll show you how to wire it. You'll keep the ignitor box to provide the tach, but you'll be cutting the coils from the ignitors control.

And for anyone else wanting to do this, I'll edit and update the first couple posts to clear out the obsolete information like the HEI and 8 cylinder mode. Pretty much any mention of the VR wheel is now completely obsolete.

[ATOMonkey]

Amazing job Steve! Glad to see you got it sorted out.

[TheSteve]

Todays goal is to find a suitable oil temp sensor and mount it to the engine somehow. I need some feedback on this one. What if I take another chevy sensor with the 1/4" snout and make it a press fit into the cooling fins on the motor? Obviously I'd have to drill the fins a little bit, but would this give me an accurate enough temperature or will the fins sap the heat away too quickly? I should clarify first:

The coolant temp sensor has almost NOTHING to do with how the engine runs. The CTS does two things: Warmup enrichment (more fuel when the motor is cold for better drivability) and Cranking pulsewidths. For whatever reason, the Katana seems immune to the warmup problems many engines have. Once mine starts, its completely fine to ride, and needs zero warmup enrichment. So for my purposes, the CTS only has to get the engine temp so it knows what to inject while cranking. For example, on a fuel injected GM engine, at 160 degrees the cranking pulses equal a stoichiometric mixture, 14.7AFR. At -40F I believe the target is a 1.5:1 AFR. Yes, exteremly rich.

The temp sensor doesnt have to be accurate, but it MUST be consistent if that makes sense. I don't care if 80degrees actually is 80, but if its say 93F, it has to be 93F everytime the sensor says 80. The consistency is key to easy starting. If the sensor is not consistent (the sensor cools faster than the engine) then I'll be cranking until the hot crankcase oil circulates and warms the sensor to the correct number.

My ideas, please add your own or give me your input:

1. Drill small area in cooling fins to fit sensor up against the block, secure with JB weld.
2. Try to remove the oil pressure switch, fit a T in there, and have the temp sensor right next to the pressure switch. Not sure if theres room.
3. Get a skinny temperature probe and put it in the hot side of the oil cooler. Might take a few cranks to get hot oil in there though.
4. Take the sensor and secure it just to the top of the valve cover. Dont know if this would work at all, if its windy outside the wind would for sure take away all the heat between the engine and tank.
5. Drill and tap the oil pan. I really dont want to risk destroying it. What about JB welding the sensor to the side of the pan, if it fits?

Leaning towards option 1 right now, I'm going out to the garage to see if its feasible. Let me know your ideas everyone.