cancel
Showing results for 
Search instead for 
Did you mean: 

Electromagnetic protection for JTAG interface?

manuel239955_stm1
Associate II
Posted on August 10, 2012 at 13:41

Hi, im not sure if this is the correct subforum for this question, if not please move this subject to the correct spot.

I am wondering if it is necessary to protect the JTAG interface of the STM32 processors from electromagnetic emission?

If the JTAG interface is not protected, EM emission should be able to set the processor in an unwanted state under circumstances, right?

Thanks!
10 REPLIES 10
Posted on August 10, 2012 at 14:07

I'd think naturally it's pretty immune to EMI, perhaps you're thinking of ESD? It's not radiation hardened though.

Tips, buy me a coffee, or three.. PayPal Venmo Up vote any posts that you find helpful, it shows what's working..
raptorhal2
Lead
Posted on August 10, 2012 at 15:00

10K pull up and pull down resisters on the board provide good protection. If you need more than that, find an EMI expert.

Cheers, Hal

manuel239955_stm1
Associate II
Posted on August 13, 2012 at 07:55

Hi clive,

ESD is part of our EMV ( thats the name of it ) and that means electromagnetic compatibility.

I have thought about a ESD-diode and a resistor (few 100 ohms) in each of the JTAG signals.

emalund
Associate III
Posted on August 13, 2012 at 14:54

ESD is part of our EMV ( thats the name of it ) and that means electromagnetic compatibility.

I have thought about a ESD-diode and a resistor (few 100 ohms) in each of the JTAG signals.

I just wonder.

while It is prudent to consider ESD when the unit is ''out there'' why do you need to consider it when the unit is in the ''programming environment''

PS beware of the caopacitance of an ESD-diode, it can be significant.  There are ESD devices that are very low capacitance and, anyhow, a diode due to how slow it is, is very poor ESD protection.

Erik
manuel239955_stm1
Associate II
Posted on August 14, 2012 at 07:37

Well, i have thought that if we want to update the processor in field via JTAG, there could arise problems when connecting the JTAG adapter to the JTAG interface, therefore protect it from ESD impulses.. or is this assumption incorrect?

emalund
Associate III
Posted on August 14, 2012 at 15:34

<i>Well, i have thought that if we want to update the processor in field via JTAG</i>

It is very uncommon to do field updates via JTAG, usually that is done with a bootloader.

Erik
Posted on August 14, 2012 at 16:15

It is very uncommon to do field updates via JTAG, usually that is done with a bootloader.

Indeed, unless you have technicians running around with JTAG adapters updating things, field upgrades are usually pushed via simpler means.

I usually have several means to update parts, including connecting to a server and pulling a new image if you have network/internet connectivity. JTAG represents the course of last resort when you've bricked the device to the point where external connection are non-functional.

Using the built in system loader, or your own boot loader, an RS232 port would provide a more robust, and ESD protected interface method than connecting directly to the pins of a CMOS processor. The processor has some ESD protection diodes on the IO pads, which clamp over/under voltages to the rails, but if you dump enough energy into it you'll just evaporate the silicon.
Tips, buy me a coffee, or three.. PayPal Venmo Up vote any posts that you find helpful, it shows what's working..
manuel239955_stm1
Associate II
Posted on August 15, 2012 at 08:18

Hi guys,

Updating via JTAG is our alternative.

Usually we use our self-written bootloader and we update the firmware via a CAN-similar network (CAN-similar protocol)

But haveing an alternative is important if our first option fails, and since its not possible to disable JTAG, why not..? 🙂

Are there any shematics or layouts available where i can see if these JTAG pins have ESD-diodes already implemented?

Thanks

Edit : Ive found a layout, there are no ESD diodes on the JTAG pins implemented..

emalund
Associate III
Posted on August 15, 2012 at 13:44

Updating via JTAG is our alternative.

 

Usually we use our self-written bootloader and we update the firmware via a CAN-similar network (CAN-similar protocol)

 

The use of JTAG in the field should be such a rarity that you do not need worry about it

But haveing an alternative is important if our first option fails, and since its not possible to disable JTAG, why not..? 🙂

 

as above

Ive found a layout, there are no ESD diodes on the JTAG pins implemented..

 

probably because the capacitance of such diodes would slow it down too much

Erik