MicroSquirt.com read only forum (Jan.15,2018)

Yes, a turbo'd 500cc bike. It was factory EFI but considering the age of the unit and the decreasing availability of spare parts, the cost of repair is escalating.

Here's what I'm thinking about doing:

1) Keep the bike itself as close to stock as possible, including intake tract, injectors, bi-metal idle air valve and such.

2) Re-use as many sensors as possible- tps, coolant & intake temp- it should be simple enough to input values so those are mapped correctly. The bike would retain whatever other senders are necessary for ancillary functions like the "Turbo" indicator on the dashboard, and those would not be connected to the MSq3.

3) Use 2 MAP sensors- one for intake charge and one for barometric. The bike has 4 pressure / vacuum sensors stock, and one of those is for the ignition system which is separate from the injection system, but I'm not sure any of those could be used with MSq3, so I am thinking that since MSq3 can control injection and ignition, one would only need 2 MAP- one for manifold pressure and the other for barometric (or placement after the air filter to accomodate restricions therein), basically one ahead of the turbo and one after. This should allow it to adapt to changing weather, elevation and air filter restriction more accurately than stock. The bike is capable of over 19 PSI boost so it's possible 2 different MAP sensors (one ahead of the turbo and a different one after the turbo) would be required; I'd like more input on how accurate the system is when pushed.

4) Use a wide-band O2. This should help the bike run better in low speed / low boost conditions for better mileage, and be beneficial for performance under boost as well. Questions arise about how to handle the transition to boost. Obviously we want the system to run more rich under boost but how will MSq3 handle that? I'm trying to figure this out so any help would be appreciated. Still learning....

5) Use the dual-input / dual-output for injection / ignition events. The bike has 4 position sensors: two on the crank and two on the cam, and 80 degree separation between cylinders. That will have to be compensated for, but I think Honda worked that out by putting the cam senders 160 degrees apart, and the crank senders 80 apart, so I could theoretically use the 2 crankshaft sensors as triggers, correct?

6) Use 1 GM LS2 coil per cylinder. This elliminates the issue with the odd firing angle and wasted spark, correct?

7) Knock sensing. Is that possible?

After a bit of research, and much head scratching, this is the direction I think I'm going to go: Use the crank pickups, #1 cylinder as VR1IN, #2 cylinder as VR2IN.
Sequential, Alternating, 1 coil per cylinder. Base timing is 25 BTDC, so trigger offset would be -335, correct?
Does this sound like the best option?

Overall, this looks like a good plan, although I think the ignition offset value would be 25 instead of 335. You would only need a MAP sensor after the throttle - a pre turbo sensor is not needed.

My first bike was a non-turbo CX500.

Matt Cramer wrote:Overall, this looks like a good plan, although I think the ignition offset value would be 25 instead of 335. You would only need a MAP sensor after the throttle - a pre turbo sensor is not needed.

My first bike was a non-turbo CX500.

I know the pre-throttle MAP is not required. I was thinking though that the factory EFI has it, and I can see the benefit of it. It would be plumbed in between the filter and turbo. Besides, it's only 30 dollars.

I was reading here http://www.microsquirt.info/dualspark.htm and it says:
"With the Dual Spark function enabled, the trigger offset should be set 'negative'."
... and ...
"For example, if tach1 occurs at a trigger offset of +40° BTDC, tach2 occurs -320° ATDC. For dual spark mode we always want a negative trigger offset, so you would use -320° instead of +40°."

Now that makes me think. Tach 1 would trigger at 0 for cylinder 1, and tach 2 would trigger at 0 for cylinder 2. Maybe the offset should be 0?

I saw a couple people have tried this but no reports of success yet, that I can see. I don't have the system yet, so can't try it out. This is why I'm posting here: for feedback and comments. ;)

Thanks,
Kirk

The section you are reading is for older code - use 3.780 code and follow the V3 microsquirt Quickstart quide that shows how to configure with Auto Trigger, which is much easier to follow than the older codes.

Thanks for the advice boss. ;) I didn't realize I was reading the wrong doc.

*edit to delete incorrect photos. back to the puzzle bin.

Please move this to the MSRuns "Project" forum. This mod will be done and will probably move to a full Megasquirt setup instead of the Microsquirt.

Not sure you had your post moved over to the megasquirt side yet.
Thought I would add a comment, as I started out with a large box full of borrowed parts from a CTX650 for my project (BMW R100RT, 980cc air cooled boxer engine, Mitsu turbo, Saab cooler, etc).
I actually was going to attempt to use all the removed kit from the donor bike before I gained enough insite as to how things should work, plus how Honda had controlled those bikes (see "plethora of sensors and reference points to assist the ecu in handling things without O2 reference). That and also the vast difference in CC's influenced those decisions.
I ended up just using the exhaust "Y" casting & the resonance air chamber with its twin throttle body and included TPS. I reasoned it would be best to target modern, easy to locate/ purchase temp sensors, injectors, and the like.
How's your IHI turbo, those are a bit harder to find now and I'm sure parts are right up there $ wise.
With the high boost those developed and no aftercooling you might want to also consider water injection as a strategy to get everything available on the table and avoid the "pinking".
Cheers, Lorne.

R100RT wrote:Not sure you had your post moved over to the megasquirt side yet.
Thought I would add a comment, as I started out with a large box full of borrowed parts from a CTX650 for my project (BMW R100RT, 980cc air cooled boxer engine, Mitsu turbo, Saab cooler, etc).
I actually was going to attempt to use all the removed kit from the donor bike before I gained enough insite as to how things should work, plus how Honda had controlled those bikes (see "plethora of sensors and reference points to assist the ecu in handling things without O2 reference). That and also the vast difference in CC's influenced those decisions.
I ended up just using the exhaust "Y" casting & the resonance air chamber with its twin throttle body and included TPS. I reasoned it would be best to target modern, easy to locate/ purchase temp sensors, injectors, and the like.
How's your IHI turbo, those are a bit harder to find now and I'm sure parts are right up there $ wise.
With the high boost those developed and no aftercooling you might want to also consider water injection as a strategy to get everything available on the table and avoid the "pinking".
Cheers, Lorne.

Post hasn't moved yet.

And I have read many of your posts... ones where people would gravitate towards one (was it a brother?) bike, then drift towards yours when they realized something about yours was just not quite right. That was yours, no?

As you have probably learned during your adventures, the EFI system on the 500T and 650T were the pre-cursors to Honda's PGM-FI system used extensively on their automobiles and many motorcycles from the 80's through present. And you were correct in not trying to implement the 650T system to control yours. Granted it could fuel the 650T upwards of 15 psi boost, but it nonetheless programmed for the 650 with its unique fuel injectors and such. The EFI system would likely not even work well on the other similar Honda 650's (CX/GL) of the time. Kudos for not wasting much time on that endeavour.

Anyhow. I've spent weeks going over every bit of information I can find: I've literally read the service manual 3 times; read the CFI manual [sic] half-a-dozen times; stared at the wiring diagram so long I see schematics when I close my eyes, and scrutinized the other CX500T/MS attempt. I fully understand how little there is in the public arena to help one understand how the 500T worked- Honda's "troubleshooting" was "test this sensor for this", "test for that" and if it fails replace the sensor, fix the wiring or replace the computer. There was no information beyond that and little, unscaled, un-indexed 3d graphs showing what the "fuel map" or "ignition map" was.

The 650T was much more simple than the 500T in comparison; basically a throwback to the earlier models with transistor ignition and mechanical advance on the ignition side, with a little ECU involvement and a decent bump in compression. The 500T was a "dual fuel map" system where one map operated basically as speed / density with one pertinent MAP type sensor after the throttle body and operated at lower throttle opening angles, which then transitioned to a different fuel map using alpha/N calculations at a certain throttle angle with a different MAP type sensor in the surge tank (what you called the resonance chamber) for boost enrichment. And the 500T had an odd ignition system in that advance and retard were purely electronic:at 0 vacuum, timing is approximately 25BTDC, at normal "idle" vacuum, timing is around 28-30, and as vacuum increases the timing advances up to around 45. Off idle, around 2500 RPM's, the turbo is starting to boost and the timing sticks mostly at 20-25 all the way up to the 9000 RPM redline / 18-19 PSI boost. Frankly, I think the fact that Honda kept the ignition seperate from the injection could actually help me do this.

The CX500T has two crank sensors (one each cylinder for ignition) and two cam sensors (one each cylinder for injection). Since they are electrically and functionally independent, I can start with Microsquirt on either one and keep the other half functioning as stock. I can start fuel only, get that tuned it to work with the stock ignition system over the entire operating range, then incorporate ignition, starting with what's known about the stock system's curves, or I can do the opposite- get ignition working first, then bring injection in later, when I've dialed in the ignition enough that fuel is the limiting factor.

I know I am ambitious here being the what, 2nd known person to try this on this bike? I wonder what happened to the first? Admittedly, it is a rare bike so why even try? That's a subject for the future and those who need to purchase parts for a bike made only 1, maybe 2 years when you consider what parts can cross from the 650T to the 500T. But when you consider even fewer 650T's were made than the 500T's, the imperative becomes clear.

But I recognized immediately that I do not have all the information available and therefore I am trying to be as methodical as possible even in the planning stages, even if it's not apparent in the posts I make. I've also has some communications with people who have used CX series engines in other arenas and am tapping their knowledge (when they let me) trying to establish a reasonable and attainable goal. Which, off the record, is 10-15% improvement in peformance and throttle response up through 4000 RPM, and 10% above.

The bike has sat awhile but the turbo was fine when I parked it. Before I start the engine, I'm going to pull the turbo's oil fittings off and drip some diesel into the housing and spin it a bit by hand, and verify what I drip in the top comes out the bottom. An inter/aftercooler (US/UK nomenclature) is in the "what if" planning stages too- just have to find one small enough to fit in with the bike's styling yet big enough to not heat soak over a mile of wide-open-throttle operation.

My intentions and other things have changed, if even ever so slightly.
I mentioned re-using as many sensors as possible, and that is still a goal, but I have expanded that to include minimal modifications to the factory harness.

Two of the 500T sensors (P1, Pb) can be used as barometric compensation as they are is 0-5v and responsive from -1 bar to 1.5 bar (relative atm)- P1 happens to be plumbed between the air filter and turbo so it's nominated for duty; Pb has an identical (within tolerance) pressure / output curve as P1, but it is located after the throttle plates. Since it has problems with reading boost, it's unacceptable for use as a MAP/boost sensor with the Microsquirt in spite of it's location between the throttle and the engine. The other two 500T sensors (P2, Pign) are not suitable for manifold MAP sensing as they are out of range at -0.5 bar (relative atm), though they can read in excess of 2 bar (relative atm). See http://cx500forum.com/index.php?/topic/ ... entry83238 for a graph of the sensors output versus vacuum / pressure. Pign may be an option as it is spec'd for input > 8V, but the output curve is spec'd to 5V- still researching that one, and it's probably not a good candidate anyways since it is the one that the ignition system uses for reference. My solution is to use the plumbing for the Pb sensor to actuate a generic MAP sensor capable of reading -1 bar through at least 1.5 bar (relative atm) within a 0-5v range.

Hopefully this project will progress from thought to implementation in the next few weeks. In the meantime, I'm open to discussion.

Actually, maybe I should shut my mouth: if I get this to work reliably, I could make a fair dime from it.

Not sure of the projects that you might be referring to (I do have a brother with very similar BMW airhead without any modification or MS equipment on it). I had a posting on the "MS Success Stories" under the motorcycle section AKA "R100RT", but then reposted on the MS2 Extra forum @ the below link half way down succession of projects (as I felt things had changed enough on the project) http://www.msextra.com/forums/viewtopic ... &start=220

I'm an only aware of one other BMW airhead boxer twin going the MS route (Les Garten) but there was an individual that modded his K75 with MS kit back along the way.
Its nice to keep things stock, but if you invest as much time as it seems you will (based on your extensive posting and investigative work) it becomes stressful when older parts on the bike fail and you can't find replacements so easy. For that reason I shifted away from a number of items I might have utilized from the 650 (fuel pressure regulator, map sensors, coolant sensors, injectors, and eventually the turbo which wasn't really a proper match for the cc's and such involved.
If you do add a charge cooler (inter/after what have you) the resonance chamber might be sacrificed as the air will calm moving through a core anyways. I can attest to the challenges of adding parts to a "faired" bike where there isn't very much room. I fabbricated my cooler from 1/2 of a Saab 900 assembly less its plastic tanks, and I log very modest intake air temps while under boost (generally max's to 100F when ambient is mid 70's). Pumping hot air just isn't a good thing.
Good luck on your project.
Lorne.