Series Resistor and TVS diode placement

You can use the kind that have two back-to-back LEDs at the input.

--sp

YES! Built in to almost ALL IC's terminal accesses is a small protection device. Each protection device to each terminal keeps these chips alive even after poor handling during manufacturing. BUT! these devices are by necessity rather small and won't handle a lot of wattage, so putting a power limiting R between the terminal and the 'outside' world is always good insurance. Or in really robust designs the Engineer will put a whole [somewhat sacrificial] interface IC between the delicate uC/uP and the outside world.

Yes, of course. First digit equals number of leads minus one. ;)

Cheers

Phil Hobbs

So why isn't an optcoupler 5N35? Maybe they don't count N/C leads.

Best regards, Spehro Pefhany

Probably the case. An old tradition: 6C4 is a triode (3) with a heater (3+1 = 4) in a 7-pin base. Though some broke this convention, like 6C5 (octal triode); tetrodes, beam tetrodes and true pentodes were sometimes 5 or 6, etc.

The newer system featured tubes as well as semiconductors, so you have tubes like the 2E26 (beam tetrode). Although the 2 is obviously not the number of active pins.

There are (were?) MOSFETs with substrate (back gate) in the 3N family.

Was there any -N- series dual or multiple opto? 19N35 would seem to be a quad 4N35, which seems silly..

Tim

well

IIRC in tubes the first number was about heater/filament voltage. See the tubes in an old all American five receiver. Three 12 somethings a 35W4 (rectifier) and a 50C5 (audio output pentode power tube).

?-)

Does that count for 0nSemi too?

I don't think so. But the 6N136 has 8 leads with 2 NC (6 leads used) and it's not a 5Nxxx. Go figure.

Best regards, Spehro Pefhany

Be VERY careful with this assumption, as it's only true for relatively high frequencies. See below.

Amen to that.

Best regards, Spehro Pefhany

As a first line of defense against external RF intrusion (to meet the EU EMC directive) is to use a ferrite bead followed by a low inductance "UHF" chip capacitor to ground and use a PCB layout that inhibits any VHF/UHF/SHF energy from cell phones and wireless broadband to penetrate any deeper into the circuit from external leads.

If you do not kill the RF energy as early as possible, it will be quite easily rectified by the protection diodes, which could cause wrong digital state to be read at a digital input.

Not an assumption, rather a sloppy statement. Thank you for the 'correction'. It is a good idea to carefully word statements for the 'soundbite' lurkers in order to prevent carrying away incorrect design rules.

My simple statement was simply refuting the broad statement that the [diode]C and L will resonate somewhere. They resonate somewhere, but often, and luckily, the 'resonance' is up into the frequency where the RF bead has gone very lossy. There is also the matter of semantics between potential resonance AND actual resonant behaviour -- I had already 'assumed' the resonance was too high or was stomped on. Yes, a C and L will resonate somewhere, but its the designer's task to include these parameters and dampen any potential resonance with appropriate [and acceptable to the circuit's performance] R value. L

Yes, RF beads do go through a 'spike' of 'inductance' but so does wire. My point is to account for ALL and DESIGN appropriately. That's why I said, "NO!, not if done CORRECTLY."

Someone had to ask. :)

But..but 3JP1 is a tube and it has a 6.3VAC/600mA filament voltage.

Best regards, Spehro Pefhany

Of course, CRTs go by inches. :)

Were there any CRTs with other-than-6.3V heaters -- or directly heated filaments for that matter?

Tim

See

Yes, there are some exception classes, your CRT being one as well as tuning indicator tubes and thyratrons. Larger transmitting tubes also usually do not follow the rule, e.g. my 4CX600J's. Nor do klystrons and traveling wave tubes. But at least the system tried to be somewhat descriptive.

?-)

The larger transmitting tubes show the number of elements first: 3: triodes 4: tetrodes 5: pentodes

A dash means an internal-anode glass tube, like 3-500Z

A X means an external-anode tube, like 4X150A

A CX means a ceramic-based external anode tube, like 4CX1000A

The number at the end is the maximum anode dissipation in watts.

The letter at the end in usually a version letter, with Z signifying a zero-bias Class B triode.

I don't know of any offhand, but I might suspect some of the miniature ones might.

I took the cover off my ca. 1979 Unisonic XL-990 and found the CRT part number (Orion 60AB4) but didn't find a datasheet on the tube. (I guess the 60 is the diagonal measurement in mm, and it's a B&W P4 type phosphor).

It still functions after 35+ years, however, sadly, there are no more analog TV signals floating around the ether for it to receive.

Best regards, Spehro Pefhany