1. Home
  2. Electronics Design
  3. Creating electromagnetic interference on an mbed to mbed SPI link

Creating electromagnetic interference on an mbed to mbed SPI link

Hi,

I have an educational project that uses two mbed devices linked together vi a SPI; the transmitter being configured as a master and the receiver config ured as a slave device. There is no significant length of cabling between the two mbed devices; they will be mounted side-by-side.

For educational purposes we want to introduce interference on the MOSI line to show the detrimental effect on the integrity of the data transmission.

One solution I have come up with that works fairly well is to diode OR the master's MOSI line and the noise source together before connection to the s lave's MOSI line.

The project originator has suggested that he would prefer no direct electri cal connection and have the interference generated via electromagnetic mean s.

I guess I'm going to need some kind of inductively coupled circuit to achie ve this, but I'm a bit stuck with the details of implementing it.

Any ideas for circuits that I could prototype to achieve this would be grat efully received.

I'm currently running the SPI link at its lowest configurable speed of 10kH z.

Thanks.

--
Cheers, Ian
Reply to
Ian Cameron
Loading thread data ...

A Mini Circuits transformer wired in series with the data line will let you put as much or as little crap on it as you like. Make sure that the primary side is always terminated--leaving it open will add a lot of inductance to the SPI side.

Cheers

Phil Hobbs

Reply to
billwether

A piezo spark thing might work, a dead fireplace lighter thing. Zap it close to the chips. Or discharge a charged film capacitor into a few-turn loop. Make sure there's a nice spark.

You could add some resistance to the MOSI line, near the master, to make it more fragile.

I'm guessing the clock line would be more likely to be the noise victim.

--

John Larkin         Highland Technology, Inc 
picosecond timing   precision measurement  

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com
Reply to
John Larkin

+1 for sure m
Reply to
makolber

EMI testing usually does a capactive coupling to data lines. I dont have the IEC specs so I can't say how this is setup.

Cheers

Reply to
Martin Riddle

Thank you all, for your responses.

One thing I should have mentioned is that this unit will be accessed remotely, so whatever is used to generate the noise needs to be activated electrically.

I am currently able to demonstrate the effects of jitter on the clock line and thought it might be nice to use another line for the noise demonstration.

I also had the idea of using a transformer to couple the noise signal into the MOSI line, but wasn't sure about the details for implementation. I was looking at audio transformers, such as these:

formatting link

thinking the DA101 might do the job.

So, transformer pin 1 would be connected to the noise source output via 470R (0-5V), pin 3 to ground, pin 6 to the master's MOSI output and pin 4 to the slave's MOSI input. I guess a fly-back diode across pins 1 and 3 would also be a good idea.

Does that seem like a reasonable starting point?

Thanks again for your advice.

--
Cheers, Ian
Reply to
Ian Cameron

You need to load it harder than that. If you open-circuit the primary, you instantly put 1-3mH in series with MISO, which isn't exactly the unperturbed condition. ;)

Assuming this is a point-to-point SPI link and not a shared bus, I'd hard-wire 50 ohms in parallel with the primary, and use a bit more noise power. That way you can unplug your test set without breaking anything.

Cheers

Phil Hobbs

Reply to
Phil Hobbs

It is a point-to-point SPI link. Ah, OK, thank you. I will look into driving it with 50R across the primary.

Thanks again.

--
Cheers, Ian
Reply to
Ian Cameron

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.