CoilX troubleshooting for the not necessarily clever

jakekooser · Post by **jakekooser** » Tue May 06, 2014 9:31 pm

I got my CoilX in the mail and in the process of connecting the wiring to it, the U2 IC fell off the board. I re-soldered it, I believe in the correct orientation, and temporarily connected it to the terminal on my coil that is fed from the ECU and a ground point on the engine, then to the RCP as shown in the examples on the wiki.

It didn't work. I am not sure if that's because of my soldering skills, or lack thereof, or if the signal I am trying to connect to is not good. I may have access to an oscilloscope here at the office but I am not 100% certain and I have to be in a training class, so my time today will be limited.

I just wanted to open a thread so I can get a documented step by step troubleshooting guide together, if that is possible. I'm attaching a picture I just took of my CoilX in the condition it's in right now.

Where should I start troubleshooting?

Thanks,

Jake

jakekooser · Post by **jakekooser** » Wed May 07, 2014 4:48 pm

So far, step 1 is: discover we don't actually have a working o'scope at work. Going to try and pass by the local makerspace after work to see what they have.

Post by **brentp** » Thu May 08, 2014 12:39 am

Hi Jake, first step is to verify the board is getting 5v and ground as expected on the RCP side.

Then, using a scope, verify that the coil input is seeing that 400~ V high voltage spike.

Then, using a scope, probe the output and ground to see if you're getting a square wave output.

Then, using a scope, probe the coil input side working your way around the components you see going towards the opto isolator; D1, D2, R2, R3.

Become familiar with the theory of operation here so you understand how the voltage clipping and filtering works. http://autosportlabs.net/CoilX

let us know what you find out!

jakekooser · Post by **jakekooser** » Thu May 08, 2014 1:45 am

I had my car re-dyno tested this afternoon because they didn't give me the A/F ratio info I paid for, and I was talking to the technician there, he feels, based on his experience, that it is probably a software / configuration thing rather than a wiring issue, so I reconnected it temporarily with the wiring outside the car and tested again. These tests were done with the CoilX connected to the green wire from the DME to the coil, and the EXT ground connected to the nearest ground point on the intake runners. Diagram attached.

First, I started with remembering to set the pulses to 6 to match my # of cylinders, with no result. I also configured all of the pulse inputs the same way, trying both 10 and 20 hz sampling rates, on the off chance that I had misunderstood which input was which, even to the extent of using the wire coming from the CoilX like a probe while watching the live readings on my tablet (via bluetooth) to see if there was any response and there never was. A couple of times while saving changes, I saw some gibberish readings, like "13" for RPM while the engine was idling at ~900 RPM, but nothing that kept displaying.

It was about then that I gave up hope of configuration fixing it and got out the multimeter myself.

Hi Jake, first step is to verify the board is getting 5v and ground as expected on the RCP side.

That is where I started tonight, and I see a solid 5V coming from the 5V reference and I also see that on the CoilX board.

Then, using a scope, verify that the coil input is seeing that 400~ V high voltage spike.

I was able to lay eyes on a scope at the local hackerspace, but the people there didn't know where the probes were, etc, and I am not a member yet, so I wasn't able to use it. The dyno guy has a scope, but as mentioned earlier, he didn't think it was wiring and nothing was connected at that point of the day anyhow. I will probably end up going out there to continue. I have one of these little guys http://www.gabotronics.com/development- ... otolab.htm that can allegedly do 20V input, plus I have a 10x probe, that should theoretically get me to 200V, but it needs a firmware update before it wants to talk to my laptop in the garage and that is a whole 'nother ordeal.

I did use the meter and see 13 VDC there, I am aware that's not much use when we're looking for pulses.

Then, using a scope, probe the output and ground to see if you're getting a square wave output.

I may be able to do this with the xprotolab scope after I update the firmware.

Then, using a scope, probe the coil input side working your way around the components you see going towards the opto isolator; D1, D2, R2, R3.

I'm with you, just trace the path to make sure it has continiuty. U2 is the optical isolator?

Become familiar with the theory of operation here so you understand how the voltage clipping and filtering works. http://autosportlabs.net/CoilX

I fear I will soon be VERY familiar with how this works. This really should not be as difficult as it has been, but I don't fault anyone but myself.

Post by **brentp** » Thu May 08, 2014 1:50 am

Thanks for the update and being willing to work through and learn, which will help us learn too if there's something fundamentally different with your engine.

which connection in that diagram did you connect CoilX to on your engine?

jakekooser · Post by **jakekooser** » Thu May 08, 2014 1:55 am

I connected the input to the CoilX here where it is highlighted. If you look at the larger diagram, it is fed from the ECU.

jakekooser · Post by **jakekooser** » Thu May 08, 2014 11:50 pm

Okay, got home tonight and reconnected everything. Ohmed out all the diodes along the path and it seems like everything is cool. Hooked up the little Xscope and started at the input to the RCP, remembering that it's only capable of ~200V if it's lucky. Nothing there. Stepped my way to the board and wasn't sure which was the output of the optoisolator, so I went to the obvious input of the opto and found this: It's pretty small, but it's there and I don't see anything coming out of the 2 pins on the other side of the opto so I am removing it and going to go double check my solder joints with the microscope camera and see if I missed something when I reattached U2.

jakekooser · Post by **jakekooser** » Fri May 09, 2014 2:22 am

I figured out that U2 wasn't soldered fully. It works, but initially it showed 1/2 of the correct RPM on the android app, so I changed the pulses from 6 to 3, which corrected that, but when I look at the data on RaceAnalyzer, the RPM value is roughly doubled. I pulled it off again because I have to figure out how to route the cables, but it is functional now. There may be an issue with one of the apps it seems, though.

Post by **brentp** » Fri May 09, 2014 2:28 am

thanks for the update! Sorry you had to go through that due to a part not properly soldered. If you have screenshots of the ignition signal going into CoilX, that would be great to see here.

Also, try monitoring RPM again in both apps after power cycling RCP, there might've been an old config value stashed.

jakekooser · Post by **jakekooser** » Fri May 09, 2014 2:37 am

No problem, Brent. Half of the bad soldering was mine, but I think it's on there now. Heat shrink is cheap, so I'm not worried. I'll hook it up again probably this weekend and get some shots of the signal going into the RCP. It's a bit difficult to manage with so few hands. I'll sort something out and get the wiring run permanently, too. That should make things a bit easier. So glad I bought this scope on a whim a couple of years ago!

Post by **brentp** » Fri May 09, 2014 2:47 am

Oscilloscopes are super important for car electronics. Glad it's working with your ignition system!

Let us know how stable the RPM reading is -especially iif you see any spikes in the reading.

jakekooser · Post by **jakekooser** » Fri May 09, 2014 11:00 am

Brent, what O-scope would you recommend for general purpose automotive use? I'm not on a NASA budget, but I'm not afraid to buy the correct tool for the job. The PC-based scope I have is functional, but it's a hassle to manage the PC and all the cabling. I'd prefer a more conventional solution.

Post by **brentp** » Fri May 09, 2014 4:32 pm

Any of the Owon scopes would be a good choice for a budget:

http://www.amazon.com/s/ref=nb_sb_ss_i_ ... cilloscope