Power supply protection networks

John Larkin a écrit :

You could also write from right to left, just to not be stuck with some

1000s years old conventions.

--
Thanks,
Fred.

Ever tried to lay out nice fat power supply traces to an 2-dimensional array of chips on a non-multilayer board? You could really get to hate those few irregular chips without power supply pins in the corners (eg. SN7490).

Sheesh, kids nowadays!

Best regards, Spehro Pefhany

--
"it\'s the network..."                          "The Journey is the reward"
speff@interlog.com             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com

sage

Sure. Use square lists of the resistors in question with a cell formula for whatever ratio (divider, whatever) you're looking for. Us conditional formatting for the ratio you're looking for, and copy that cell to the square. The real power of spreadsheets is the copy and paste function.

And why are quad things so screwed up? LM358: makes sense. LM324: WTF? LM393: consistent. LM339: WTF? TL084 is yet another pinout.

Tim

-- Deep Friar: a very philosophical monk. Website:

Derives from the circuitry... optimized metal routing _on_the_chip_.

...Jim Thompson

-- | James E.Thompson, P.E. | mens | | Analog Innovations, Inc. | et | | Analog/Mixed-Signal ASIC's and Discrete Systems | manus | | Phoenix, Arizona 85048 Skype: Contacts Only | | | Voice:(480)460-2350 Fax: Available upon request | Brass Rat | | E-mail Icon at

| 1962 | Isn't the definition of "totalitarian" when the government owns significant manufacturing; and controls the major media outlets?

I have a hunch that this may have been done (in part, at least) to try to maintain some degree of footprint and routing compatibility, when early analog DIP parts were made as substitutes for older metal-can (e.g. TO-5) parts.

The TO-5 parts number the pins "going around in a circle". By numbering the pins on (e.g.) a DIP-8 as they are, the pins are in the same basic orientations as the equivalent pins on a TO-5. This would allow a DIP-8 footprint to be placed on the PC board, and stuffed with either a DIP-8 ceramic or plastic package, or a TO-5 part standing up above the board with its leads bent slightly to enter the rectangular pattern of holes.

--
Dave Platt                                    AE6EO
Friends of Jade Warrior home page:  http://www.radagast.org/jade-warrior
  I do _not_ wish to receive unsolicited commercial email, and I will
     boycott any company which has the gall to send me such ads!

"Jim Thompson" wrote in message news: snipped-for-privacy@4ax.com...

So? I don't give a shit what's in the black box. ;-)

And speaking of power lines: they'd be even better all the way to one side (pins 1 and 14, or 7 and 8). There's lots of precedence for this. Tons of tubes have adjecent heater pins (3 and 4; 4 and 5; 7 and 8; 1 and 12; etc.). Only very few chips -- usually the big ones, like DIP40s -- came anywhere close, but they dropped them in the middle (10 and 30 or so), which is just as useless in through-hole wiring (though it does work well for SMT caps).

Aren't some of the high speed logic chips doing that nowadays, because they simply can't afford to be bypassed any other way?

Tim

--
Deep Friar: a very philosophical monk.
Website: http://webpages.charter.net/dawill/tmoranwms

You would if the guy that did the stuff inside the black box, couldn't.

How do you plan to wire that when you have chips in two dimensions?

Tubes had the advantage of multi-layer wiring. ;-)

When they came out with the 74S and 74AS series one company threatened to do that because of package inductance. The others laughed them out of business.

Why would you do something that stupid? You do the hard part and leave money on the table because you didn't do the easy part.

Still stupid.

I looked them up on DigiKey. They seem to be a lot more than that. Not a very good selection, either.

TI did some center-power-ground parts, for a while.

John

It's not unreasonable for a control system to have a high-resolution, high-accuracy sensor and a relatively crude actuator. Lots of processes are slow and essentially lowpass the process input. An on-off (1 bit) triac drive can do excellent temperature control.

Oversample and conquer!

John

There is this thing necessary for high speed parallel busses called line length equalization. I have even seen it done for faster analog stuff. It gets real clear when you think it through with the why in place.

All the so called ergonomic keyboards piss me off, most especially the idiot flat versions. Somebody needs to beat these twits to the edge of death to make them actually study the physiometrics of typing.

That you don't understand it doesn't mean that it's stupid.

The input signal (from ADC) follows several path, some with integrators, others without. Then filtering, digital gain and RMS over some some BW. We have to see tiny signals, under the output LSB which will be scaled for some outputs, and dithering is not an option there.

And 20 bits ADC doesn't necessary mean 20 bits accuracy. We are using the 12b ADC in a multi-bits 3rd order SD configuration to get resolution with low OSR. And the ADC comes for free with the DACs and uC. Just some opamps. And digital gain means less calibration, so what?

Yeah...

--
Thanks,
Fred.

Then his company fires him and gets a better one.

Doesn't bother the FPGA guys. Why should it bother a measly LM324?

Remarkably, chips do too, last I checked.

Tim

--
Deep Friar: a very philosophical monk.
Website: http://webpages.charter.net/dawill/tmoranwms

Some of the 54xx/54Hxx chips were like that.

Physics rules.

When was the last time you saw an FPGA on a two-layer board?

Stupid beyond belief.

WTF?

=A0 =A0 ...Jim Thompson

Maybe not. The (V-) connection can be made just about anywhere, there's no 'metal routing' issue in connecting to the substrate.

The LM324 design clearly uses reflection to make four amplifiers from one design, like bookmatched veneers, while the LM339 uses stamp-a-copy repetition instead. Looks like a personal-preference issue to me.