Cam wheel speed error in code 3.83

[indrek86]

Hello to everyone!

I'm working on my first EFI motorcycle project (Yamaha XS400).
Mechanical part is almost done and ECU with all sensors is under testing right now.
I have microsquirt with V3.83 code.
I use 24-1 tooth cam trigger wheel.

I found error in rpm calculation, when "single cam wheel input" is selected.
Everything is working, but speed in tunerstudio MS is wrong. It is the same as cam wheel speed, but it should be multiplied by 2 to get the crankshaft speed. I can fix it in tunerstudio, but I think that all rpm based calculations in microsquirt are still wrong.
For example: fuelcut function. I adjusted fuelcut to trigger when speed is going above 3000 rpm, but it cuts fuel when speed is above 6000 rpm.

ignition, injector and tach output(injector LED) signals are OK. In tach output is crankshaft speed.
I didn't found any information how to correct this problem.

Is there any software fix for that? Or do I have to make all tables and other rpm based settings with actual engine speed divided by 2?


regards,
Indrek

[Matt Cramer]

Can you post a copy of the problem MSQ and a data log of the engine running on it? This will give everyone more to go on.

[indrek86]

Engine is not running yet. I made testbench run.
Trigger wheel is on the camshaft. Tunerstudio & MegaLogViewer show correct camshaft RPM. But problem is that there should be crankshaft RPM because it is 4 stroke engine and camshaft is moving at 2x lower speed than crankshaft.
Correct engine speed is only in tach out (injector_LED). I added picture of injector_LED signal when my camshaft trigger wheel is running at 3000 RPM. There you can see the real engine speed (6000 RPM). At the same time tunerstudio show me 3000 RPM.

[Matt Cramer]

What are you using to measure the camshaft RPM?

[indrek86]

I'm using oscilloscope.

[indrek86]

All other ignition setups where primary trigger wheel is on the crankshaft are OK, because there is no need to multiply RPM value in uSquirt firmware.

XS400 ignition sensors use camshaft speed and there is no room to add some trigger wheel to the crankshaft. It can be done, but it need a lot of mechanical work. So the easiest way was to use original pickup sensor and build new trigger wheel to replace original one pulse wheel. But now while testing I discovered that all RPM based things (timing advance, rev limiter....) not working correctly.

[24c]

Hi Indrek,

I hope I'm not being dopey, but I've used a twin sensor camshaft single tooth based set ups, and a 12-1 tooth single camshaft sensor set up, and I think I had RPM issues too, because I was using a two speed cordless drill with known specs. However, somewhere along the line, I changed some settings and it resolved itself.

Your post has got me thinking to dig out the test rig again and just double check. I just looked at my desktop, so now I look at the laptop.

Mike

[indrek86]

Hi Mike,

Which firmware version did you use for camshaft based setups? Maybe some earlier version is better. I have checked all ignition related settings in tunerstudio, but still no luck.
I haven't found any good information for camshaft based trigger setup.
Maybe you can give me your tunerstudio files? Then I can check is there any difference.

My test rig is a DC motor with variable power supply to adjust RPM. 24-1 trigger wheel is connected to it. I check all signals with oscilloscope. I use LM1815 conditioner for pickup sensor and signal in MC9S12C64 MCU TachIC input is good.

I haven't tried to connect trigger wheel signal to 2nd input (TachIC2), but I think then it will not work at all.


Indrek

[24c]

Indrek,

I was switching between different firmwares, MSExtra & B&G, and another product from Imfsoft. However yesterday, I was looking for the Tunerstudio files on my laptop to help you, when my brother popped round, and then I had some Video8 movies to convert, put a USB power supply on his motorcycle, plus somebody popped round with his car electric window not working, so I got sidetracked. I will try to go through it today. I have my rig, found a v3 MicroSquirt, so I'll try and find out what I did.

Most folks poo poo cam based set ups, because of the lack of precision, but they forget most applications are older sloppy based Kettering solutions, so there is a marked improvement anyway over original.

I was using a 12-1 wheel with a Hall effect sensor, and another one tooth wheel with two Hall sensor (dual spark), all operating on the camshaft. I too have a variable speed rig, but it's AC washing machine motor, with drill speed controller, does 1000-23000 rpm using polyvee drive system. The cordless drill, does really slow speeds 0-650 rpm, and being low powered you can load it at "cranking" to simulate compression effects! If I was making a rig again, I'd use a variable speed router (woodworking type), because they work great for two stroke set ups.

If you use the other Tach on it's own, it won't work AFAIK.

I'll update this post later when I find my Tunserstudio files and double check with my rig, because like you, my stuff wasn't put on a running engine, just bench testing.

Mike

[24c]

Indrek,

I have just set up my rig again, and got a RPM output using my Hall sensor and a 12-1 wheel. Quite funny really, so used to messing with Module boards and v2 Microsquirts, it took a slight time to remember how to hook the Hall sensor up!
I have used my cordless drill (known max speed in LOW or 1 is 440 rpm max), and I am seeing around 840 in my TunerStudio gauge, which equates to 2x camshaft speed or crankshaft speed.

My MicroSquirt v3 is loaded with B&G 3.76
My basic .msq settings in Dual/Multi Spark Options are "Single Cam Wheel Input", and in Advanced Ignition Options Trigger Wheel Teeth = 12, Missing Teeth = 1, and Cam teeth = 0

I haven't loaded 3.83 in yet, but 3.76 seems to be showing the right RPMs. I really don't know if this helps you.

Mike