what is WRONG with these people?

Their native language has scripts that read from left to right? Too many schematics seem to be like that nowadays :(

Have you never got a footprint wrong and then been thankful that an alternative fits?

--
Mike Perkins 
Video Solutions Ltd 

Sometimes you can flip and/or rotate a part to make it work, but the pick-and-place machine doesn't know to do that. So I can't qualify pinout-different parts for the same application, even if it's just rotation, because that might put mixed parts into the same stock bin.

These phemts are just crazy. Most of them seem to be pre-flipped. The big pin is randomly located.

That Skyworks part is electrically amazing. The Avagos are probably OK for RF, but they are mushy and leaky, especially a heap of drain-gate leakage. We use RF parts in switching apps, which they weren't intended to do.

Avago (now Broadcom) avoids the various bizarre pin numbering schemes by not numbering their pins at all.

--
John Larkin         Highland Technology, Inc 

lunatic fringe electronics

And here I've been wondering for decades why they don't just adhere to one or two standard pin-outs for TO-92 BJTs.

Every possible SOT-23 pin numbering scheme has been used. And I think every possible SOT-143 (4 pin) permutation, too.

--
John Larkin         Highland Technology, Inc 

lunatic fringe electronics

I'm pretty sure there are some BCCxx transistors that don't have the same pinout between manufacturers, even on the same part.

"SKY65050" would be a good name for a top-shelf premium vodka

I remember ordering two batches of the plastic version of

2N2222As (originally metal TO-18). One batch was OK but the other one, stamped PN2222A, seemed to be all defective. It took me some time to figure out that they'd swapped C and E. That was before everything was available online in pdf.

Hmmmm, SKYY75040 (750ml 40%) would be more conventional!

If they wanted the free marketing, to an incredibly exclusive subset of engineers that actually knows what a SKY65050 is, and who mostly otherwise drink beer and whiskey and not cocktails... sure, they could use that number. :-)

Tim

--
Seven Transistor Labs, LLC 
Electrical Engineering Consultation and Contract Design 

After mostly getting SOT-23 pinouts right they felt they had to exercise their creativity more.

--sp

--
Best regards,  
Spehro Pefhany 

A reminder to all RF manufacturers (oh yeah, like they're even listening...) that the IPC defines consistent pinouts. SOT143 is listed under IPC-7351 section 8.9, and references IPC-7352 SOT143 and JEDEC TO-253.

Not sure if this is the official JEDEC print (as of the date given), but the pinning is consistent with every other damned standard in the world.

The tolerances are mostly tighter than SOT-23, which is nice.

Somewhat sadly, IPC is only voluntary standards, nothing required.

Tim

--
Seven Transistor Labs, LLC 
Electrical Engineering Consultation and Contract Design 

SOT343N (~SC70-4L but unregistered) with the wider pin on the left originates in Europe, though the inverse pinout, created by final leadforming alone, was a common European option (philips and Zetex)

SOT343R (SC82AB) with the wider tab on the right, originates in SEAsia.

Pinout is vendor-driven and depends on both wafer substrate polarity and the end-use intended application.

Pinout of current parts may be simply legacy, as the RF houses have passed through many hands under corporate restructuring. No-one in this field intentionally produces a part simply as second-source competition.

RL

Actually that was 2mm bodies........

The earliest parts were actually the 3mm versions; SOT24(right)/SOT143(left) in Europe and SC61(left)/SC61A(right) in SEAsia.

Jedec only registers TO253(left)

RL

Yup. If you don't care about slightly worse flatband noise (0.4 nV vs 0.3) and higher 1/f corner (30 vs 10 MHz), the Skyworks part is the bee's knees--it looks like a JFET that got bitten by a radioactive spider. :)

What do you like about it specially?

Cheers

Phil Hobbs (by the pool today due to having sanded my feet a bit raw on the sand yesterday)

I remember the earliest inline plastic transistors being BCE. Now EBC is more common.

--
John Larkin         Highland Technology, Inc 

lunatic fringe electronics

Ultra low drain leakage, sub-uA. If I use it to discharge a capacitor, to make a super-linear timing ramp, I don't want a lot of drain current when it's off. And I'd like to be able to ramp to +5.4 volts (current source clamped to +5) without worries.

Some phemts have so much d-g leakage they can sort of latch up. The Skyworks is amazing, for a phemt.

Us mouse-clicking tenderfoots need to wear flipflops or something outdoors. Ski boots work for me.

--
John Larkin         Highland Technology, Inc 

lunatic fringe electronics

Interesting, thanks. I used a couple of ATF55143s in that sub-$3 TDR because it was all 50 ohms and single supply. You can patch up the Broadcom parts with a cascode for RFish things, but it doesn't help leakage!

Cheers

Phil Hobbs

EBC is usual, BCE was Japanese (2SA, 2SB, 2SC, and 2SDnnn parts) while BEC is used for RF parts (less Miller capacitance, is the theory).

You do a bit of artistic leg-crossing when a 2SC transistor comes out, and a PN2222 has to replace it.

Ge alloy transistors had the base in the middle (mechanical reasons), and the JEDEC plastic parts were one base-lead bend from fitting the old metal-case footprint.

The big enhancement parts, like ATF50189, are great for power switching. That one is good for 1 amp. It also makes an outrageous diode.

--
John Larkin         Highland Technology, Inc 

lunatic fringe electronics