Surge Current and Power in FETs

Why argue with it? Most of the boards of any complexity I've ever designed were not 100% right first time, even if I thought they would be. So let it go and make sure the test spec has plenty of battery connect/disconnect tests. You might find some unrelated issue during testing which no amount of simulation will find.

What test spec?

I hope that says it all. I am constantly on the verge of quitting as this is so poorly run. I'm not going to make any further efforts on this issue. I completed a simulation with every resistance in the path I could think of and it still stresses the FETs to the limit. If the board designer doesn't want to consider this something to be fixed... especially when it can be done with an external connector, I am going to throw in the towel.

In my world, engineers tell PCB designers what to do.

Most of our boards are sellable at rev A, the first PCB to be fabbed. And we rarely prototype. Just review hard. If you assume that the first board will have serious bugs, it will.

I wonder if this is still the ventilator project.

In my world, circuit designers are engineers and PCB designers are engineers and they work together to produce a successful PCB. Sometimes they are the same person.

Pretty much anything is sellable if the customer doesn't notice its defects or doesn't care.

Probably

>

Sounds reasonable to me. If this is the ventilator then I don't want it tested on me. So I'd prefer it didn't get as far as testing.

=============================

** A 1 ohm NTC thermistor would help a lot, one rated at say 6 amp would drop to under 0.1 carrying 4 amps rms.

Arcing damage to the contacts on you connector is another concern with "hot plugging".

.... Phil

What kind of parasitics? It is important to determine worst case numbers even if some of that has to be a SWAG.

That is likely going to result in field failures and an increasing warranty overhead, upon which the CFO will become unhappy.

? ... Even adding a 0.1 ohm

Not sure how the hierarchy goes in that company and whether you are an employee or self-employed. My answer to this as an employee would be that I will not sign the ECO. If things could become dangerous you may have to go a step farther.

As a consultant I make it clear, in writing, that I consider the design too marginal.

Hazard as in "three alarm fire"? That would put a different spin on things. If it's just a failure then, well, that company will see a lot of warranty claims. "No, no, I did not leave it on the wrong connector, honestly".

That's usually worst case and often there isn't any leakage to write home about. Why no discharge resistors?

I don't think any are important other than the resistances of all the wiring connectors and components in the current path. There is a resettable fuse, but it takes some 100 ms to activate even at 100 amps. The circuit draws 4 amps avg with peaks up to 10 amps, so the circuit needs to tolerate that.

The guy designing the power board decided that rather than kill the main power rail other than the supervisory circuit (a couple of diodes could power that from the two power sources), the items on the 12V rail can be individually shut off. The battery has a self drain current around 100 uA, so the circuits are well bellow that when the machine is off, at least in theory. Adding a bleed to the caps drains the battery when off as the caps are always under power. Even if they got turned off, we would have the surge issue when turned on. It was a poor solution in this regard. Hand waving and bluster rather than analysis.

.> >>>Same with the FETs, they offer typical and max, but no min for the channel resistance. This is also impacted by temperature.

PTC thermistors are sometimes used - their resistance goes up as they get warm.

Wait until a unit blows up on switch-on

That became practical in the 1990's when, printed circuit layout programs became tolerably user=friendly. Lay-out drafts-people are cheaper than engineers, but if the board layout is tricky for electrical reasons - lots of cross-talk, or controlled transmission line impedances - it can make sense to get the board designer to do the job. Telling the layout draft-person exactly what to do takes time, and they don't always get the whole message.

John Larkin does seems to go in for incremental design and development. Two weeks from start to completing all the documentation (including layouts) doesn't accommodate a lot of innovation.,

If you aren't changing much, there's not a lot that can go wrong.

In my world there are 7 guys to do back-end web development, 4 guys to write node.js, half-dozen guys to do the Android app. 4 managers, several marketing people, etc. all making like 75k/year in a startup environment

The hardware design for their new product idea like "design and build a novel motion-control camera platform" gets put on Upwork at a starting bid of $7,000 flat fee for everything from concept to finished PCB and enclosure design and if you can't do it all yourself f*ck you we'll find someone who can we're not paying two hardware contractors, lol

at least that, I should say

Is the designer allowed to change the circuit on his own? That could get chaotic.

Most PCB designers don't understand electronics. All the best ones that I've worked with have been women who don't. I've never met a PCB designer with an EE degree.

I'm working on a board layout right now. Critical parts are easier to do myself than to explain to someone else.

It kinda fun, once in a while.

We don't knowingly ship defects. Again, if you expect defects and expect to find them by testing, you'll have them. I know one organization that defines six formal iterations of PCB layout, and aways uses six or more. It takes them years to get anything into production.

....

Only 4 managers?

IKR?

It's bad when they admit upfront in the solicitation their last engineer left with the project "near complete" but then they describe what needs to be done and it's like "by near-complete we mean it don't work at all"

....How do you argue with that sort of reasoning?

Some do.

I have.

Why? I thought you said only women could do it.

Oh. That's why.

-------------------------------------------

** No fooling ??

Dogs bark and cats meow - incredible stuff.

Inrush suppression requires an NTC. See my other post here.

..... Phil

That's insane. I said no such thing.

Sorry I forgot that those from your part of the world have their sense of humour surgically removed at birth.

Illogic is not humor. Not in this part of the world.

I was born in a 1936 Ford. Nobody was available to remove anything, for which I'm grateful.