General > General Technical Chat
Dumbest "equivalent schematic" from a datasheet - how many errors can you find?
<< < (6/6)
magic:

--- Quote from: Zero999 on March 29, 2022, 10:35:19 am ---I admit, the only error I saw was the inputs reversed.
--- End quote ---
Well, wrong transistor type was the first thing I noticed too and VCC the last.
But for me it was familiarity with LM358 internals followed by "no way that they really changed it, no one ever got three stages compensated at low cost".


--- Quote from: Zero999 on March 29, 2022, 10:35:19 am ---why didn't they simply copy and paste it? It's probably even the same die as the LM358, just with only one half connected.

--- End quote ---
Part number sounds like it's second source of STM product. For whatever reasons (aka NIH), schematics of such things are usually redrawn, by all manufacturers. The poor intern tasked with this job got the STM schematic and a firm instruction to bring it to TI standards, including elimination of all 4-way junctions, which he did ;D

Regarding the internals, STM actually manufactures a single channel version. There is also LM321 made by National Instruments and ON Semi, likely a similar thing.
fourfathom:

--- Quote from: magic on March 29, 2022, 08:30:06 am ---Which is funny because TI is normally very religious about avoiding four-way junctions.
--- End quote ---

I am violently opposed to four-way schematic junctions, but they show up in so many places I was beginning to feel like I was out of touch.  I even wrote a polite and well-reasoned (IMHO) letter to QST (the ham radio magazine) in which I suggested that 4-way junctions should not be used in their schematics.  The letter wasn't published, and they still use the 4-way style.

Hey, you kids get those 4-way junctions off my lawn!
T3sl4co1l:

--- Quote from: Zero999 on March 29, 2022, 10:35:19 am ---The drafting errors are more obvious and indeed understandable, after seeing the original, nut why didn't they simply copy and paste it? It's probably even the same die as the LM358, just with only one half connected.

--- End quote ---

Well, C&P won't do for updating it to whatever the latest documenting style is.  Or maybe they want to have a vector or EDA version of it, who knows.

I suspect they don't have any standards as far as what schematics are drawn in, as long as they meet some minimum format and appearance standards.  Possibly vector format is included in that.  I've seen Altium, OrCAD, etc. in schematics, at least in the application sections / notes, so they're not picky about those at least.  Probably block diagrams, inline figures, etc. are done in just, InkScape or any other variety of vector-format graphics tools.  I've seen so many goofy, avoidable errors with shifted lines and inconsistent and dumb symbols, it can't be drawn with library symbols, at least it's not all the time.

Here's a particularly dumb example:



They also have a tendency (but, again, no consistency) to use the stupid emitter-arrow "MOS?" symbol, and I've seen it wired in reverse before (some LDO, don't remember offhand what).  Which is exactly how I'd think you'd wire such a bastard of a symbol, but nah, the convention is apparently analogous to an NPN and the "emitter" is supposed to be the source.

And as you can see above, sometimes they draw an insulated gate, sometimes not.  Hm, don't think I've seen -- wait wait wait hold up I've got another one.
https://www.ti.com/lit/ds/symlink/ucc256304.pdf
The HV startup circuit, they consistently(!) refer to, throughout the datasheet (and its relative the -404), as using a "JFET".  These are CMOS chips.  There's no fukkin' way they put an HV JFET in there.  The block diagram shows a MOSFET, as you'd expect.  Is the block diagram actually lying then?

Anyway, I was going to say, while they've sometimes removed the gate oxide, I don't think I've seen a gate junction (arrow to/from substrate, i.e. a JFET) before, just, at the simplest, the IC version of symbol (symmetrical D/S, SS left off, gate circle/not indicates P/N channel).  Without it actually being a JFET, I mean (and at that, mainly just among the classic op-amps; very little indeed uses JFETs anymore? at least on block level).

Tim
magic:
Of course vector formats are used, for inclusion in PDF datasheets if nothing else.

TI has always had a consistent style and symbol set for transistor level schematics, so did others like National. I think I could recognize these two 90% of time if not more. Well, at least for so long as they published schematics at all, so up to 2000~2010 or so. Their contemporary "block diagrams" are becoming less consistent.

This botched TS321 is obviously done in their standard style. Another example here; you just know it's TI.




--- Quote from: T3sl4co1l on March 30, 2022, 12:54:25 am ---They also have a tendency (but, again, no consistency) to use the stupid emitter-arrow "MOS?" symbol, and I've seen it wired in reverse before (some LDO, don't remember offhand what).  Which is exactly how I'd think you'd wire such a bastard of a symbol, but nah, the convention is apparently analogous to an NPN and the "emitter" is supposed to be the source.
--- End quote ---
It's a fairly widespread convention in CMOS IC design to draw things like below, by analogy to bipolar.

The N-MOSFET symbol in your schematic looks like somebody got creative combining both approaches. Well, at least the body-channel arrow is shown correctly.

Navigation
Message Index
Previous page
There was an error while thanking
Thanking...

Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod