View Single Post
  #39   Report Post  
Posted to alt.binaries.schematics.electronic,sci.electronics.design
John Larkin John Larkin is offline
external usenet poster
 
Posts: 1,420
Default High failure rate on 2N7000 - 2n7000.jpg (0/1) - TraceInductanceEffects.pdf

On Tue, 19 Nov 2013 00:00:08 -0600, "Tim Williams"
wrote:

"John Larkin" wrote in message
.. .
He gets abusive every time he figures out that he's wrong.

A PCB trace, or a wire, isn't an inductor: it's a transmission line
with some Zo. The drain voltage won't fly up more than Id*Zo, and will
fly up much less for practical cases.


So let's see how often you specify routed traces by inductance rather than
impedance and length in power circuits. Hmm? Oh, but that would defeat
your argument, so you won't tell. How convenient.

At 1/4 lamba:
Transmission line with Zo Z: inductive
Transmission line with Zo Z: capacitive

Independent of frequency (aside from Xc / Xl dependent on F).

Given the broadband nature of the relay circuit, overshoot will be
something, whether tranmission line or lumped equivalent. Obviously not
25kV, obviously not even over Vds(max), whatever that happens to be. But
consider a trace, to a relay, laid by someone without understanding of
high speed action, and probably not using a ground plane. Zo will
probably be in the 300 ohm range. Spike to avalanche? Easy.

Now shut the **** up, both of you. You're a disgrace to the internet.


You're not making any more sense than Thompson.

The issue was, with a mosfet driving a relay coil, whether the clamp diode is
located at the coil or at the drain of the mosfet. I've heard people insist that
it must be at the coil, and Jim insists it must be at the fet drain. I think it
doesn't matter.

The case in point, that started this, was a 2N7000 driving a clamped 12-volt
relay coil at 30 ma, on a PC board. Jim thinks the inductance of the trace can
blow the fet. He posted a preposterous simulation to prove it.

If the trace were very long, and was 300 ohm Zo (both unlikely) the overshoot
from the trace would be 9 volts max, less in real life.


--

John Larkin Highland Technology Inc
www.highlandtechnology.com jlarkin at highlandtechnology dot com

Precision electronic instrumentation
Picosecond-resolution Digital Delay and Pulse generators
Custom timing and laser controllers
Photonics and fiberoptic TTL data links
VME analog, thermocouple, LVDT, synchro, tachometer
Multichannel arbitrary waveform generators