esd diodes as diodes

per

run,

gear

i/o.

with

either

input. If

calibration

every

Again, Captain Obvious sticks his neck out; "if it's a better solution, it might be better".

Again, Captain Obvious makes an incredible statement.

???

ote:

a relay per

aceable

test run,

all the gear

ributed i/o.

ables with

e I look

se I was

puts,

tan, either

then

longer

bles

tizing

o each input. If

libration

muxing of

ce one at every

bvious one

Ah. I did not notice I was talking to AlwaysRight.

Dimiter

relay per

traceable

run,

the gear

distributed i/o.

cables with

look

was

either

longer

digitizing

input. If

calibration

of

at every

one

You were the one trying to tell John his business, then defending the position with tautologies that certainly deserved the Captain Obvious tag.

ote:

nd a relay per

traceable

ry test run,

th all the gear

istributed i/o.

g cables with

here I look

cause I was

inputs,

kistan, either

ask then

no longer

cables

igitizing

into each input. If

calibration

g_

le

he muxing of

Place one at every

n obvious one

, it

ition

I was asking about his point of view. It happens that this sort of business is similar to mine, and the issue under discussion is certainly part of my business.

And yes, I happen to have the acquisition _and_ networking technology under one hood, so I was curious what can be the reason for others not to move this way.

He was not interested in answering which is something I can live with.

Yet I was not after your "know better" general comments which asked for the "AlwaysRight" tag.

Dimiter

relay per

traceable

test run,

all the gear

distributed i/o.

cables with

I look

because I was

inputs,

Pakistan, either

then

longer

cables

digitizing

each input. If

calibration

muxing of

one at every

obvious one

Yawn.

He told you. I can't help it if you didn't/couldn't read.

He answered before you asked. Perhaps he doesn't like repeating himself.

Wrong, bucko. You're the one telling others they don't know what they're doing with tautological proofs.

re

Right, AlwaysRight. Just keep on echoing my posts every time to show you have grasped some of the point, what would the world be without your invaluable insights. Never mind your tech posts are party talk grade at best, keep on delivering. I'll skip the rest.

I wanted yours, which you proved were nothing more than platitudes and gibberish.

You almost got me, AlwaysWrong. I almost didn't recognize you there.

That's the beauty of the daughterboard-for-termination-and- conditioning concept. It's been around since at least the 70's and

These daughterboards stick up perpendicularly from the "mainboard" and have the ESD and other per-channel components on them.

I could imagine, but I don't think anyone sells them, IDC connectors with built in ESD diodes etc.

Note that D-sub connectors with RFI feedthrough filtering (sometimes just capacitor to ground but in some instances little pi networks with tiny little series inductors) and/or ESD diodes have existed for decades. Goddamn expensive though. I remember seeing prices like $30 each. I get the impression these are used as band-aids over poor designs that neglected RFI etc. to begin with.

Tim.

But they greatly increase costs.

These are plug-in anti-aliasing filters we added at a customer's request. But we charge for them.

ftp://jjlarkin.lmi.net/Filter1.jpg

ftp://jjlarkin.lmi.net/Filter2.jpg

John

We are going to order them. The 0.3mm PCB (or FPC) can be added to the back of the existing PCB. 0201 ESD diodes are around 6 cents each (Qty 15K). Adding around $1 for bonding, 26 pins can be done for around $3 ea. Setup cost is high, but per unit is not too much.

We are constantly changing the main PCB, so we can't build too many at a time. But we are OK with 576 (15K/26) ESD add-ons immediately.

More connectors/complexity => higher cost and lower reliability

Adds capacitance to the I/O and all sorts of other problems, too.

Recently NXP proudly announced in an internal bulletin the design-in of no less than 73 (or was it 83?) of their ESD protection diodes in a cheapo Chinese MP3 player. Can you believe it? They must have used 'em for some other purpose besides blocking off the USB and headphobe jack.

robert