TunerStudio disconnecting when 2k+ rpm

[Ferrazzo]

Hello, i have a MS installed on my monocylinder 35 cm³ engine for a university project.
When it is running, after i connect to the TunerStudio and speed up to above ~2k rpm, the TunerStudio loses connection to the ECU and some seconds after going back to slow rpms (800~1k), the connection comes back.

I am using a RS232 to USB cable. I know i should go wireless, but i live in a country where things are really expensive and this project should be the cheapest possible.

I am guessing the problem is EM noise from the coil. After this guess, i measured the output signal of the coil (pin 36 on my MS1 V2.2) with an oscilloscope and before running into some other problems, i was able to notice that the signal was pretty shitty (wish i could have took a photo). But when activated, i could notice that the signal had really high up peaks (as some down too).

So, i would like to ask you all, is this the, in fact, the problem ?
Is my guess correct ? If so, what could i do to correct it ? Money is a problem, so wireless is not a possible solution, unless it is arduino related.
Should i use a snubber on the coil signal output ?

I know i should upload logs or at least images to help, but im really asking for guesses and if my prediction is correct and what could i do to fix it.

Really thanks in advance.

[Matt Cramer]

What sort of USB adapter are you using? Also, are you using non-resistor spark plugs by any chance? Non resistor plugs are nothing but trouble with EFI.

[Ferrazzo]

Hello and thanks for the answer, the cable i am using is similar to this one:

http://mlb-s1-p.mlstatic.com/cabo-conve ... 2013-F.jpg

The spark plug i am using is a resistive one with about 5-6k ohms.

[Matt Cramer]

Not good - that long USB cable will act as an antenna for any stray electrical noise.

[Ferrazzo]

I think i might have a smaller one, but as i remember, me and my colleagues from the university did some tests (with a smartphone) and found that the EM noise near the coil was much bigger and since the ECU is near the coil as well, we tried using a long cable to prevent that (and that looked like it helped). But we will try that and see what happens.

[Matt Cramer]

Use a long run of RS232 cabling and keep the USB cabling to a minimum. RS232 is way more noise resistant.

[Ferrazzo]

Did it. But we also got a new coil and placed a ferrite core on the coil signal wire. Apparently, the problem disappeared.
Here is a video:

https://youtu.be/52yxLWdaljs

The TPS is a little buggy, we will probably fix it today.