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

eestrine wrote:Al, since the cam VR sensor is putting out such a low signal, we are considering replacing it with a Hall sensor. Hall Sensor to OPTIN+/- and crank VR sensor to VRin2. Do you forsee any problems with this solution?
Elliott

The 1.4 V from the cam is definitely too low to trigger. But as 24c says, you must wire the crank VR to VR1 and the cam/ Hall sensor to VR2.

Al, added some steel to the cam wheel. With 0.030 gap had 1.4 volts peak-to-peak. With 0.007 gap had 4.0 volts peak-to -peak. Still no spark or rpm in the falling input ... mode. Tried other modes. In "Dual inputs timing from 1 cam tooth" mode got 482 rpm but spark on only #1 cylinder. This is the correct rpm when cranking with the starter. What do I try next?? Also where do I find the software for checking the input tach pulses??

Thanks
Elliott

The "Dual inputs timing from 1 cam tooth" is probably getting its rpm from the VR1 sensor. To me it looks like the cam sensor is not showing up on the output of VR2IN. If you can scope the input of VR2IN at R50 and the output (that goes to the processor) at the junction of R43 and R58, that would be the quickest way to tell. Or you can find the tachref code under the headings Ignition and under it Tooth Analyzer on the left side scroll menu at microsquirt.info.

Elliot, are you still using two VR sensors (I reread your posts, but IIRC, you are still using VR1 crank & VR2 for the cam), or have you tried the Hall on the cam?

FYI, early on I didn't always see the cam sensor using TachRef.exe when cranking(this was a voltage thing), but always did using a two channel oscilloscope. If you are using two VR sensors, one on the crank, and the cam, you are probably suffering from the shared sensor grounds issue that plagued me.

Yes I have a VR crank sensor with a nine tooth flywheel and a VR cam sensor with one tooth. I looked at both with a scope and the signals look good. VR2 has a 4 volt peak to peak signal. I put a 2k resistor across the signal and it seemed a bit cleaner. I do see some very sharp spikes among wondering if these could be from the starter. I do get spark now on both cylinders, but the timing is way off and the only mode that has spark is the "Single Cam Input" mode. Will try the tachref software next week.

Attached is a log file from a plugs out run.

Thanks for your response.

Elliott

Try the tachref, but when you do use the only configuration that will work properly with your engine. That is the falling (maybe rising) cam synch with crank wheel. That is what you have. Another config may give you spark, but it won't work properly. As to the rest, you must have negative trigger offset. Again you may get it to spark, but it won't be proper timing and you will have problems at high end. The way I would set the configuration is to assume and even fire engine, get it to spark properly on the scope. Compare the cam signal to the spark outputs. The sparks should be 360 deg out of phase and one spark from each cylinder between each cam synch. Of course this won't run the engine, so fuel should be disconnected, but if this isn't right, you will unlikely get the Odd angle to work either. Once you do have it right, then you only have to play with the odd angle.

On the good side, the datalog shows that you have solid synch in the sense that you are not losing rpm and trigger count is stable. But your spark advance shows -26 deg. This may be from using trigger rise. I would use calculated to start and open up the pulse tolerances, if needed. It works on the bench, so you should be able to get both spark outputs. Then its a matter of adjusting trigger offset and delay teeth to put the sparks where you would want for an EVEN fire engine. The next step is to put the second spark output where you want with odd angle offset.

Al, I did as you said, downloaded Tachref. Tried it in several modes. No rpm, spark or pulses in the Falling signal with wheel.. So, just to make sure it was working, I tried the single crank wheel and single cam wheel modes. In the signle crank wheel mode, got spark, rpm 482, and pulses in Tachref. Disconnected the crank sensor and connected the cam wheel sensor to VR1. No spark or rpm but I did get pulses in Tachref. Appears to me that the MPU can recoginze signals from both sensors. I looked at them on the scope again and they look very clean. Seems like not recognizing cam signal on VR2, which is 4 volts peak-to-peak. What is the level of the signal on your bench test? I did use calculated and varied the percent from 25 to 75 with no effect. I'm stumped.

We are considering going to a 18 tooth crank wheel with 1 missing tooth, but I'd like to try your new beta version before I give up on this configuration. How do I get the software?

Elliott

I'm not sure what you tried. The tachref program wipes out the ECU code and just looks at the tach inputs and dumps out what it sees to the PC. There is no mode. You can set the ECU to whatever mode you want before you load tachref, but this is gone once you load tachref. So I suspect something else was wrong. It looks like tachref is seeing VR1 but not VR2, but for some reason didn't see either on the first attempt. If you confirm that with Tachref you CAN see cam signal when put on VR1 (it will show up as blue) but can't see it on VR2 (should show up as yellow), then the problem lies in the differences between the VR1 and VR2 circuits. The TachRef is useful because it rules out any software code or config problem. It is like a scope, but shows what the ECU sees after the tach signals go through on board processing.

You might also try putting VR1 signal on VR2 - if that doesn't show up either, it may be the port is burned out.

Please believe me there is no sense in trying an 18 tooth wheel, the problem is in the cam signal. I would switch to the 18 tooth after I got everything working, but only because it will give better timing accuracy - or I guess you could use it in wasted spark mode with no cam.

eestrine wrote:Yes I have a VR crank sensor with a nine tooth flywheel and a VR cam sensor with one tooth.

Excellent, you have the same issue I had with TachRef.exe not showing a cam sensor plot on VR2 (yellow)...when I did sort out the voltages which were lower than yours, I had to use two diodes one on each VR input to get my bike to run with B&G 2.890.
I used a couple of 1N1007s, and I will take a picture of the orientation, but the unit is not accessible at the moment. From what I remember, if I placed the diode one way round it blocked the signal (polarity issue I think) and the other way it made no difference on the crank VR, and the same for the cam VR2 too. They have to go back to back a bit like half a bridge rectifier and the common connection is the VR sensor ground pin 33?.

Good luck buddy.

I have found a picture of the two diodes, when I was getting pretty pissed at sticking various components on the co-ax with the soldering iron. Some healthy snot on there...
[img]two_diodes.jpg[/img]
You can make out the orientation re the VR sensors and the co-ax ground wire.