Why no depletion mode LDOs?

An ECO is a company document that tells manufacturing to do something, like take exception to a drawing, or create a new dash number, or dab some epoxy on a lug, or something like that.

Here's one:

That's different from a PADS ECO file, which is generated automatically by one of the PADS programs.

The first form of ECO vastly pre-dates the second one.

And it has to be signed with this pen:

Did you investigate why this happened? I don't like that sort of "fix" unless the root cause is well understood.

The root cause seems to be bad fuses.

Now *that* sounds like not-well-understood to me.

And an arbitrary value like 1nF is? Come now...

Tim

I've never had to write one (I guess I should I say that with some smugness??). The only direct use I have of ECOs is likely the same as PADS' ECOs. Either could be made into a report as such.

If you need to tack on a "reason" for those squishy readers made of meat, "it's broke" oughta be good enough.x

Tim

My circuit was thought about and then simulated, and the values intelligently (by my standards) tweaked for stability and load transient response. The fets are operating deep in their ohmic regions, so the loop dynamics isn't as obvious as if there were more delta-v available.

Our ECOs often have a longish section explaining the history and details of a product and why it has to be changed. It might say why one customer is doing something unusual and wants a special version. It might give manufacturing options as to whether to apply the ECO at all. We might include before-and-after waveforms. ECOs are an important part of our longterm institutional memory. Not just "do this", but why.

"It's broke" wouldn't get signed by me.

A PADS eco looks like...

*REMARK* Deleted pins: 2, Added pins: 4

which isn't all that useful, long-term.

Work/motion hardening of the fuse element... I see it all the time in automotive applications. ...Jim Thompson

1. These boards are powered up essentially all the time, not cycled like car fuses.
2. We tested a bunch of these fuses at pure DC, and some failed after a week or so.
3. Still can't get your kill file to work right?

unless

Works well... observers are invited to drop by and see just how well.

I just decided I wasn't going to let some little San Fransicko prick exclude me from participating in technical discussions.

As to the automotive fuses... it's not cycling, it's vibration... makes the element look like a crystal... then they crumble. ...Jim Thompson

unless

We can see from here. Either you can't manage a simple kill file, or you're lying yet again. I'd bet on lying.

Technical discussions? Packing material? Tacky candles dripping on tacky decks? Redneck politics? Water softener dilemmas?

Pitiful.

unless

Drop by and see who it is that is lying.

Yep. Pitiful how you can't help but be rude. Must be terrible to feel so insecure that you have to resort to such tactics. But I guess that's all a weenie can fall back on. ...Jim Thompson

Otherwise it's unstable. So much for "designing LDO's". ...Jim Thompson

It's not unstable. The dump resistors greatly improve load-step response.

unless

I've seen pics of your house and its surroundings. No thanks.

Me? Rude? You deliberately, grossly, and repeatedly insulted my wife. I said that those acts make you an asshole forever, and they do.

Put me back in your kill file and keep out of my threads, like you've sworn to do so many times.

Actually, there is one theory: the fuse blow current depends on heat sinking, which depends on pad geometry and board stackup. The rating may assume a different PCB situation than ours.

We strung together a bunch of these fuses, in series connected by small wires, and ran them at 200 mA, and one failed after a few days. That partially confirms the heat sinking theory. But rather than investigate more, we just fixed it with the half-amp polyfuses. We could have put in zero-ohm jumpers, actually.

We've found that the trip current of a surface-mount polyfuse is also very dependant on PCB layout. The thru-hole radials are much more repeatable.

I don't know about "commercial", but there are certainly such circuits in custom ASIC's...

Like RFID tags I designed for Intermec (an IBM spin-off that I'm sure Hobb's knows of)... used a depletion device (Fairchild ASIC process) as a very low power way to start-up bandgaps... turned off as soon as bandgap bias became self-sustaining. ...Jim Thompson

"Your" threads? What a weenie. ...Jim Thompson