First, look at voltages induced by the magnetic fields:
The magnetic field from a single infinite long cable is:
u * I where u = 4pi10-7
B = ------------ S is in meters
2 * pi * S
I know you don't have a long cable, nor a single cable, but the
approximation yields a worst case scenario.
Let's assume that you weld with 400A
That gives you a field 1 meter away of 80 uT, not a lot of field.
Let's assume that the maximum voltage you can stand induced from that loop
is 0.3v (so the protection circuitry doesn't kick in and short out a lot of
current.)
The induced voltage for a loop of wire is:
e = B Area dphi/dt = B Area 2piF
Rearranging:
e
Area < ----------- or around 10 sq meters
B 2piF
That's a loop of 10 feet on each side!
Given that your welding cables are probably 0000 Awg and twisted *and* your
sensor and control wires are also twisted pairs, it's hard to conceive that
you'll have difficulty from any voltage induced into your wires from the
magnetic field.
Now let's look at potential voltage transients from the welding:
Striking the arc is not so much, but breaking the arc should generate quite
a voltage. Let's assume 10KV (I know, I know. What about the operator?
After all, we are after the worst case scenario.)
Now you have wires that are within 1 inch of a ground plate that are 3 feet
from 10KV, you will get transients (ratio'd per distance) in the order of
1/30 of 10KV, or 300V! On some circuits that can cause damage, but probably
not 4000 CMOS.
There will be some saving grace that the 300V is a very high impedance
source. However, with power applied to the chips, there is a potential for
the reverse SCR of the substrate to get turned on by the high impedance
voltage source which would be catastrophic to the chips.
What about straight capacitive connection to the circuitry when its powered
off?
Let's assume there's 0.1pF parasitic capacitance between the cables and your
circuitry. And further assume that when the welding arc extinguishes the
voltage rises to 10,000 V with a risetime of around 5 nS.
Current through the shorted parasitic capacitor would be:
i = C dv/dt or around 200mA that does sound
bad.
Notes: Adding caps in parallel to the gates' input capacitances will
benefit nothing since the Thevinin equivalent will simply yield a lower
voltage with a much lower impedance. It would require over 3,000 pF to keep
the transients below .3V
However, adding some "series" impedance between the high voltage spikes and
the circuitry will enable the input capacitance to keep the voltages down.
CONCLUSION
The chips will not be damaged by the magnetic fields whether power is on or
off.
The chips could be damaged by high voltage transients as the arc
extinguishes. And is more likely to be damaged if power is applied. But
simple shielding between the circuitry and the welder will stop all this.
- Robert -
-----Original Message-----
From: Mark <[email protected]>
To: [email protected] <[email protected]>
List-Post: [email protected]
Date: Tuesday, October 12, 1999 1:05 PM
Subject: EMI from nearby welding
>
>The query I'd like to put to the readers of this list is not exactly
>EMC related, but is close enough that I hope you'll forgive me
>asking.
>
>Can anyone give me an idea of the voltages likely to be induced
>in the unprotected input wiring of a process plant control system,
>when welding is carried out nearby? Or point me to a resource
>that may give me some help?
>
>The welding cables will be close (within a metre of the inputs to
>the control system, perhaps even touching the control system
>metal case). The welding transformer will be tens of metres
>away. The process plant control system is an old logic design
>based around 4000 series CMOS.
>
>The control system will not be energised when the welding is
>going on, so my main worry is induced voltages/currents blowing
>up the CMOS. Should I worry, or sleep happy?
>
>All advice gratefully received (except if someone tells me to throw
>away the control system and get something better protected, as
>that's not an option...)
>
>Regards, Mark
>--
>Mark Hone
>
> Wellman CJB Limited Email: [email protected]
> Airport Service Road Tel: +44 (0)2392 629239 (Direct)
> Portsmouth, Hampshire Tel: +44 (0)2392 664911
> PO3 5PG, ENGLAND Fax: +44 (0)2392 697864
>
---------
This message is coming from the emc-pstc discussion list.
To cancel your subscription, send mail to [email protected]
with the single line: "unsubscribe emc-pstc" (without the
quotes). For help, send mail to [email protected],
[email protected], [email protected], or
[email protected] (the list administrators).
