I'm building a cable tester for 8 wire UTP, cat III - V (normal 10Mb/s
network cables) up to 100m.
Using a 16F84 I am wondering if I should use any ESD prevention.
My setup looks something like this: (plz. use a Courier font)
----+ -- +5V -- +5V +---
16F + + + +16F
84 + Diode Diode +84
+ + + +
RBx +--1K--------Twisted cable-----+---1K--+Rbx
+ + + +
+ Diode Diode +
----+ + + +---
-- GND --GND
The 1K resistors are not for ESD, but I may "short-circuit" 2 or more pins.
My question is: Are the diodes needed?
I have read (page 23, note 2 in Microchips F84 manual) that
"I/O pins have diode protection to VDD and VSS".
But can I trust that? Are they strong enough for an ESD discharge?
Holger Morgen wrote:
> I'm building a cable tester for 8 wire UTP, cat III - V (normal 10Mb/s
> network cables) up to 100m.
> Using a 16F84 I am wondering if I should use any ESD prevention.
Yes. Of course. Cables are large capacitors +/- piezoelectric
> The 1K resistors are not for ESD, but I may "short-circuit" 2 or more
Of course the resistors are for ESD.
> My question is: Are the diodes needed?
Yes. Split the resistor preferably, 470 ohms between terminal and
diodes, another 470 between diodes and PIC.
> I have read (page 23, note 2 in Microchips F84 manual) that
> "I/O pins have diode protection to VDD and VSS".
> But can I trust that? Are they strong enough for an ESD discharge?
Maybe, maybe not. That's why you use the extra.
The only limitation I can see is that I can't see a cable as fully
"tested" with such a resistor in series!
|part 0 2112 bytes content-type:application/x-unknown-content-type-asm_auto_file; name="Cable2.asm"A few months ago I made similar simple tester for RJ45 network cable.
without searching the net looking at other designs I figured the
quickest way to check wires will be by pair.
we have 8 wire that means 4 pairs.
this is how I did it .
attached is the source code
I didn't even use crystal RC is good enough.
4 green LED's and 4 red LED's
power on reset all LED's will turn on to make sure all of them ok
to determine the start 1p, 2p, 3p, 4p I turned all green LED's on then
it will turn on one by one if polarity is wrong red will
turn on if open then no light. the schematic I have in ORCAD format.
all together took me about 2 hours to do it. as far as hardware you
do not need any diode.
Let me know about it.
Holger Morgen wrote:
Content-Type: application/x-unknown-content-type-asm_auto_file; name="Cable2.asm"
Content-Disposition: inline; filename="Cable2.asm"
Attachment converted: wonderland:Cable2.asm (????/----) (0002D27D)
Yes, you can make several possible arrangements with the wires, and one
good example is just using always return jumpers at the end of the cable
(as a loop back connector), so you can control everything from this
side. The use of a nice multicontact relay at this side can allow you to
simple connect all pairs in a single long pair that loops back as
another pair, so you can measure also losses in high frequencies and
more using a simple comparison of amplitude/frequency received. Remember
that electric signals travel aprox at 5ns/m... you could measure the
Andre Abelian wrote:
Thanks for all your help!
> Paul B. wrote:
> The only limitation I can see is that I can't see a cable as fully
> "tested" with such a resistor in series!
"My" will only check if the RJ45 connectors are installed correct. I will
not even check if the indivual
twisted pairs are mountet the "right" place. (Too dificult).
More... (looser matching)
- Last day of these posts
- In 1999
, 2000 only
- New search...