Prop died at a site

T Chap

The Prop was fine for days. When I left they were welding steel with an electric system within 6' of the Prop enclosure. Since that day it has not been turned on, today the Prop is DOA. I never ever had a Prop die in my systems. Is it possible the welding killed it?

Phil Pilgrim (PhiPi)

Are you sure it's the Prop that died and not some ancillary circuitry. And if the Prop really is dead, how about the other ICs surrounding it?

-Phil

T Chap

I just had them try to reprogram via PC and my proploader. It would not run via ram or eeprom.

T Chap

The USB chip responds and is working but since it wont even run with RAM then I assume the Prop got killed by the welder proximity.

MikeDYur

T Chap wrote: »

I just had them try to reprogram via PC and my proploader. It would not run via ram or eeprom.

What's feeding it? May have been voltage spikes killed the on board regulators.

EDIT: I'm sure you already checked voltages.

T Chap

The LED's are lit on the main prop board that show 5V in and 3v3 LDO. I never have an issue with this stuff. I am partly asking if the welder could have done it from less than 6' away.

Phil Pilgrim (PhiPi)

I guess the short answer would be "yes," the welder could have caused the problem. But that does not answer what specifically died or how to prevent it from happening again.

-Phil

evanh

Treat this as a learning exercise and go remove the failed unit to diagnose the fault free of charge. Get to the bottom of it ASAP so you aren't building the same problem again.

As for the causes, don't go guessing too much just yet but the two generally tough ones to keep on your mind are:

1) Powerful RFI (From the welding arc itself.) can do it, particularly any stored Flash content is easiest to erase. Usually this is not fatal. Full metal shielding is the only protection from this.

2) But what is more likely is fault like currents in the interconnected metalwork and electrical systems. If there is an external common connection across the circuit board then you are at risk of a high current in that path, this will produce voltages on the I/O pins that can be excessive and therefore damaging. At the extreme it can even blow the tracks off the circuit board.

The typical design avoidance for this issue is to use electrical isolators like optocouplers and transformers. Usually lots of them.

Peter Jakacki

Yes, those arcs can induce a lot of energy into unprotected systems but you have left out 99% of the information.
What board are you using?
How is it mounted? (in a metal case etc)
What leads connect to the box?
Is the welder running off the same mains power as the Prop?
What type of power supply are you using?

If you answer that then that probably covers enough information for the moment, then there's the other 90%......

T Chap

Thanks for the ideas guys. When I left the project last week, it was working and I removed power from the steel enclosure when I left, but the lid was off the enclosure for access to buttons and screen so the custom prop board was not completely shielded. There were several un sheilded cat5 plugs coming off the main board with direct and semi direct access to pins via 1k, 4,7k R's. *The power is now and was from extension cords coming from some place near a pole I assume, there is no main power in the house yet. I assume the welding stuff was coming from the same place, but on different extension cords. My power is an xformer>Brid Rec> 5V switching 2Amp supply into the main board with 3v3 ldo. All the boards have many years and lots of systems installed with 100% up time minus one lightening strike. Since the Prop appears dead as it won't boot and show anything on an LCD nor does anything work, I know the Prop is not running. I tried running off RAM just to see if the eeprom was bad, but no luck on RAM. I will send out a board and it will be fine but I was trying to mainly determine how it failed after not being plugged in. I'll know soon what happened when it gets back on the bench. It is not clear how long the system was plugged into the extension cord during the welding, but at some point the system was unplugged from power and only some CAT5 cables were connected for signals ie motor hall sensors, encoders, switches.

In hind sight I recall turning the system on a few occasions and it didn't boot, this never happens. It is possible the solder was at the end of it's life. So maybe this was a board fault, easy to diagnose when I get it back.

Ltech

We get this some time ago,
System was working for years, but one time we have a failure of the common ground to an rf power amplifier.
First thinking the amp was broken, because low rf output, and blame the propeller sending wrong commands

=> the amplifier takes partial the ground trough the propeller , >15 amps from 12 volt battery....
The propeller pcb was not powered, but trough the sensors and external connection ground, we get a lot of damage.
Eeprom dead, some pins on the propeller, total lose

Guess the grounding of the welding electrode was not always 100% to ground, and passed partial trough the grounding of you pcb....
Welding @ 125Amps ... and get some amps trough the sensors gound

jmg

T Chap wrote: »

The Prop was fine for days. When I left they were welding steel with an electric system within 6' of the Prop enclosure.

Is this something they will do frequently, or was it a one-off case ?

When testing the returned board, I recall mention the PLL is more 'damage-sensitive', and you could test it with code that uses the RC osc first ?