Newbie: ignition only setup on a small 4-stroke single

[mfro]

Hello everybody!

I'm new on this forum (although registered already for quite some time and busy swallowing as much of helpful information on this forum as brains allow). I live in Germany so please excuse my English which is pretty wacky at times.

I'm currently trying to attach a Microsquirt to a small (153cc) 4-stroke single. Although it might seem counter-intuitive, I have to start with an ignition-only setup. The stock generator (about 40-60W AC output) is much too small to deliver enough power for a decent fuel pump so I had to adapt a more powerful alternator first. With that change, all the stock ignition parts are gone and I have to use the Microsquirt to drive my coil before I can seriously start with the EFI setup.

I was able to attach a 180W generator to the engine which will hopefully able to deliver enough power to feed a complete Microsquirt setup with carefully selected parts, but now I have to setup ignition first (fuel still on carb). I have manufactured a 18-1 trigger wheel and will attach a VR sensor to it (mounting bracket yet to be finished).

DSC_2757.jpg (28.71 KiB) Viewed 1820 times

My first questions: what do I have to do to the (at first) unused "EFI-only" inputs/outputs if I intend to start with ignition only setup to avoid confusion or even damage? What do I need to enter into Megatune to temporarily disable most of the EFI functionality?

Thanks in advance for your help.
Markus

[800vtwin]

You don't need to disable anything, just don't hook them up (and protect the free wires from shorts.)

All you need to start is the vr input (#1), power, grounds and ignition 1 out. You can run wasted spark and not used a cam sensor. You will need a map sensor or tps sensor (with alpha-n settings) for the timing table use. And the fuel pump out to run a relay to the pump. And sensor power supply.

So pins 1, 8, 12, 21, 22, 23, 32, 33, then 24 or 27 with 20 & 28.

The manual is your best friend.

[mfro]

The manual is your best friend.

Thank you. It's under my pillows ;) .

Another question (I guess from your nickname you're familiar with bike applications): many motorbike ignition coils do not have a negative terminal (they're just grounded to the frame). The manual says the Microsquirt should drive the negative terminal.

Do I need to look for another coil?

[24c]

Do I need to look for another coil?

I would... On most motorcycle that have EFI, it is normal to have a 12V positive feed to the coil, and the MicroSquirt equivalent switching the negative on and off to make a spark. OK some have stick coils, plug top coils, CDI systems rather than the more normal inductive coil and HT lead to the spark plug, but you will need two inputs on the primary side. I assume your bike, monkey bike? is 12V.

[mfro]

I assume your bike, monkey bike? is 12V.

Thank you. Found a coil that seems suitable and ordered last night.

Yes, we're talking about a Honda Monkey here and yes, it's 12V.

[800vtwin]

yes, and cdi's usually put out high ac voltage (100v ++ ). This is what your old coil needs to work.

I just reread my post. You do not need to run a map or tps sensor.

You just need a static input. A simple resistor across the map sensor wiring would be the easiest solution. The timing table only needs to be 2d. - Use one cell across the rpm range.

[mfro]

You just need a static input. A simple resistor across the map sensor wiring would be the easiest solution. The timing table only needs to be 2d. - Use one cell across the rpm range.

Wouldn't it be enough to just put the same values into every row of the timing table?

[mfro]

I'm back from testing. With mixed results, some of them I don't understand.

DSC_2759.jpg (30.16 KiB) Viewed 1768 times

I've put my trigger wheel on the lathe, hooked the Microsquirt on hoping to see some meaningful values. After initially disappointing erratic display with ridicolous RPMs in the 50000 range, I managed to adjust everything until values seemed reasonable. My lathe has a switch where I can change two speeds on the fly, nominal between 330 and 660 RPM.
Megatune shows 364 and 737 respectively which seems pretty close.

I only got it to work by entering strange values as "Skip Teeth" the "Advanced Ignition Settings" dialog:
Trigger Wheel Teeth: 18
Missing Teeth: 1
Skip Teeth: 36
Delay Teeth: 0

As I understand the manual, for a single cylinder, wasted spark engine, this should be either 0 or a multiple of "Trigger Wheel Teeth".

However, I don't get tach sync at all if I enter 0, get a very erratic result with 18 and values seem to be reasonable at 36. Can anybody explain this behaviour?

It might or might not correspond with the other, more serious problem I have: when I move the belt, my lathe can do 1400/2800 RPM. I didn't manage to get tach sync at those speeds at all, no matter what I tried. The engine where I want to mount the microsquirt is supposed to do close to 13000 RPM, so a reliable tach is essential to me. I know my trigger wheel teeth aren't optimal according to the manual (it recommends square teeth if I remember correctly, but the way I have it was much easier to fabricate) - would the tooth shape have so much influence on the signal quality or is there anything else I could try to get reliable results?

I'd appreciate any helpful response.

[grippo]

The ridiculous rpms are probably the result of noise which you have fixed, at least at low rpms. The fact that you had to use 36 skip teeth (meaning 1 tach pulse per 720 deg) indicates you are using a sequential dual spark mode, which means you should have a cam synch, but I am guessing you want to run wasted spark. If you look at microsquirt.info, and select dual spark on the left sidebar menu, you will see a lot of good information and illustrations on setting up your wheel, even for 1 cylinder engines, and at the end there is another link to even more information which explains what all the terms mean.

We have had all sorts of tooth shapes and the car manufacturers use all different shapes. The important thing is that the sensor match the wheel, which means the small detection part of the sensor should be smaller than the tooth, and the distance from the tooth is also dependent on the the wheel size and sensitivity of the sensor. A sensor that works with a large wheel will see a large signal because of the large metal mass and therefore needs less sensitivity.

It is extremely difficult to design a circuit that works with all wheel and sensor combinations. Especially when you factor in the fact that many people make their own wheels. Generally the problem we have had at higher rpms is that the missing tooth gap produces a small bump in the middle where it should be flat. We have put in software to throw out this false bump, but it is probably not perfect, so you may have to play with resistors and diodes as other people have done. There are numerous posts on this. We are working on better diagnostics and fixes for this, but it is a hard problem and all aftermarket ECUs have problems with this. Many are sold with a wheel and sensor so they can be sure it works.

One thing that will help is to look at the rpm gauge in TS. If it jumps to 0 for an instant, it means the ECU is seeing noise on the tooth and is losing synch, then quickly regains it. There is also a trigger counter (I think there is a gauge for it) that counts up if noise (extra tooth) comes in, and counts down if a tooth takes to long to come in, based on the time between previous teeth. The last would mean the signal is too weak, so move sensor closer to wheel. In the datalog, there is also a deltaT that measures the time between teeth in microseconds - this is what the ECU is seeing, so on a lathe at steady speed the times should be close.

[mfro]

We have had all sorts of tooth shapes and the car manufacturers use all different shapes. The important thing is that the sensor match the wheel, which means the small detection part of the sensor should be smaller than the tooth, and the distance from the tooth is also dependent on the the wheel size and sensitivity of the sensor. A sensor that works with a large wheel will see a large signal because of the large metal mass and therefore needs less sensitivity.

Thanks a ton. Advice from the guru is gold, obviously . I changed sensors from this

DSC_2761.jpg (23.61 KiB) Viewed 1759 times

(automotive sensor probably meant for cars with larger wheels)
to this

DSC_2762.jpg (19.1 KiB) Viewed 1759 times

(Nippon Denso sensor used in small motorbikes)
and that fixed sync problems instantly at all RPMs my lathe is able to do (even 2800 RPM is now steadily displayed at 2820 without any counter reset when changing speed).

The fact that you had to use 36 skip teeth (meaning 1 tach pulse per 720 deg) indicates you are using a sequential dual spark mode, which means you should have a cam synch, but I am guessing you want to run wasted spark.

It still only works with skip teeth = 36 (18=RPMx2, 0=erratic behaviour), although I double checked that I have disabled dual spark mode (I do in fact want to run wasted spark). Do you have an idea on how to fix this? Does it matter at all - what if I just leave it since it seems to work?