EEPROM being destroyed during processor programming

[rogerjef]

Hi

We have a very occational problem that results in a M25P16 EEPROM being destroyed during the programming of the associated dsPIC30F6015 microcontroller.

We have monitored the voltages around the EEPROM and all are within the normal operating voltage limits during the programming.

The board in question is one being used for development, and has had the EEPROM changed twice in the last 4 months. The board, installed in a test rig, has been re-programmed probably about 50 times without problems.

Has anyone else experienced this type of failure??

Thanks

 

[zappedagain]

That doesn't seem like enough resistance to cause issues.

How long is your cable? You can build up a lot of voltage on an unterminated cables as it swings through the air. Once you plug it in that voltage surges through your circuit.

 

[rogerjef]

The RS232 serial lead, used for powering the board during programming, is approx. 2m long, as is the lead to the Real Ice programmer. The lead from the programmer to the header is approx 400mm.

The serial lead is always plugged in first.

One interesting thing is that the SCLK pull down resistor is physically sited next to M0.

I can very vaguely see a mechanism for failure, but it is clutching at straws.

During programming there is pulsatile current flowing in the 0V return to the programmer. If that line is high(ish) resistance, and maybe with some inductance, then possibly programming current could flow down the RS232 gnd.

This may possibly cause a spikey voltage to be developed across the pull down resistor directly connected to M0, stressing and maybe killing it.

The other memories might be protected by track inductance.

As I say, straw clutching at it's maximum

 

[MacGyverS2000]

Nah, that dog won't hunt...

Dan - Owner

http://www.Hi-TecDesigns.com

 

[IRstuff]

Blowing an input requires either a forward-biased junction running amps, or the standard reverse-biased junction running 10+ kV.

An induced EMF in the ground leads are usually less than 1 V.

TTFN

FAQ731-376

 

[rogerjef]

As I said, Gents. Straw clutching at it's peak.

I guess I will now have to wait until the next occurance.

Thanks for your suggestions

Roger

 

[notaguest]

Do you have passive resistor pullups on the Hold and Write protect inputs? I would also consider pullups on the chip select for that M0 chip. Are you using 3V3 for the entire board or is there also 5V powered devices associated with the serial bus?

If M0 is the only one blowing, then the chip select is the only unique line for that chip. I would also check the actual board layout to ensure there is nothing connected to the DNC pins of the chips etc. Do not rely on a schematic drawing.

 

[rogerjef]

Pullups on chip select lines only. Pull down on SCLK.

The whole of the digital section runs from 3V3. There is +/-5V supplies for the analogy bits.

There are no DNC connections on the SOIC8 version of the M25P16.

Just M0 failing is the confusing bit.

 

[notaguest]

I'm not sure of the SCLK 'stuck'. If there is hardware damage to the M25P16 clock input, then try adding a small input resistor such as 10R to the SCLK input and CE input so that you can see if the inputs are sourcing fault currents.

Like you say, there must be some definite reason for only M0 to be affected so far. Usually power-up transients are where damage like this are liklely to occur.

 

[zzzjim]

For some overkill monitor Airwaves with a spectrum analyzer.

could a long unscreened track be grabbing local RF tuned to its length , Mobile phone / TAXI pickup .

They stop people transmitting on oil-rigs !

 

[EddieCurrent]

There's no way for me to know if this is your issue, but some manufacturers are notorious for having issues at "unspecified voltage" that can occur during shutdown (sometimes startup) between the time the voltage drops below minimum operating voltage but there's still enough voltage to operate. Shutdown is usually worse since its often slower. Low power devices are worse yet. If this is a test setup, there's probably a nice big power supply that can run the system a long time as the voltage drifts down making things worse yet again. You can get into weird situations where some op codes work,others don't, etc...

In many instances a robust reset signal is all that is required to keep the system from running, but in others it may be necessary to take even more precautions. A watchdog on the power supply should be sufficient but check with the manufacturer or test it yourself. Failure to do so can result in some very odd behavior...The system works great until you shut it off, then intermittant issues pop up depending on where you were in the code when you start shutting down and how quickly you actually turn off.

I have seen systems with bootcode programming capability take themselves out during these transients. We could never see how we hit the code, then the mfgr informed us that at some voltage level the branch instructions no longer operated reliably and we just blew through the bootloader and erased program memory.