1N4007 quandry

[alexradium]

A friend of mine works in the semiconductor industry as a chip designer. When they started making multi-core processors they started off getting quite a few failures from fabrication. As an example, if they were making quad-core chips, some would pass the 4-core test coming out of fab while others would not. Those quad cores that had a bad core or two used to be culled. Then he came up with the idea of rebranding those culled multi-core processors as good processors with a smaller number of cores. Many dual core processors are actually 4-core rejects. He got a patent for that re-binning idea. If you ever wondered where a 6-core CPU came from, it was supposed to be an 8-core, but it failed so they binned it as a 6.

a lot of the time "lesser" rated products are just performance variants of one product that meet different perf levels, so they get binned into specific performance levels. All RAM chips are sorted this way. They don't bother to make low-frequency RAM chips. they try to make all high-frequency RAM chips, and the ones that don't perform up to spec get rebranded as lower performance products.

same with tubes...just think of all the grades and colors....that was GT business.

[R.G.]

When the 1N4000 series was first conceived, it was >hard< to make 1kV junctions. So the high voltage ones did sell at a premium, and you only bought as much high voltage as you really, really needed. As processes got better, and making 1kV junctions got easier, I ..suspect.. the manufacturers of only trying to make 1N4007s and sorting out the lower breakovers to sell as the cheaper parts. Eventually, they may only try to make a few voltage grades, or maybe only one. Those datasheets ever mention a maximum breakover voltage. It may be that all 1N400x parts now do over a kV. Or a couple of lower ones. Hard to tell.

It's important to remember that we're looking over a gap in time about the same as the gap between the first electrical utilities EVER and the second world war. Lots of things change.

[Randall]

Historical accounts aside... OK nevermind. I'll just use 4007's and not bother with it further.

[loudthud]

I'll bet if JM measured the breakdown of his 1N4002s they'd all be over 1KV.

[Enzo]

Randall, there was never a reason you as an individual needed to buy 1N4001 or whatever, 1N4007 will be fine. Some industry contract though could specify a "low voltage" type, so they make them available.

[Justin Thomas]

This sounds similar to what I had heard was done with CC resistors back when... So maybe it's not that new?

Justin

[darkfenriz]

Aeons ago the 1N4001,2,3 had lower capacitances and more importantly lower reverse recovery charges than 4007, so it made sense to use lower volatge rated one in some applications.
After the switched mode revolution of the '80s and '90s all of those landed in the ultra-slow department and are only used for up to 400Hz switching.

[picturevalve]

Temperature coefficient

I seem to recall an amp. I fixed which used 1N4004 diodes as bias compensators, they were thermally connected to the driver transistors. I forget the amp or the config., but the spec was 1N4004 due to their temp. coefficient.
I bought quite a few to match a pair.

[Mick Bailey]

I wondered about junction capacitance, but some spec sheets show the lower voltage devices have higher capacitance (15pf against 8pf). If I was guessing I'd have put it the other way round. Other sheets show an average of 15pf for all types.

[darkfenriz]

I bet modern 1N400x have nothing to do with originals from '60s except for being 1A and slow. Ditto for between manufacturer's.

Recovery times (for example) will vary much more with the manufacturer than with voltage rating.

[Dave H]

I seem to recall an amp. I fixed which used 1N4004 diodes as bias compensators, they were thermally connected to the driver transistors. I forget the amp or the config., but the spec was 1N4004 due to their temp. coefficient.

The temperature coefficient of a PN junction (Vf or Vbe) is about -2mV/deg C. Either Diodes or a transistor can be used for bias compensation. I've used a TO92 transistor stuck with thermal compound in a hole in the heatsink. It's easier to fit than a string of diodes.

[picturevalve]

The temperature coefficient of a PN junction (Vf or Vbe) is about -2mV/deg C. Either Diodes or a transistor can be used for bias compensation. I've used a TO92 transistor stuck with thermal compound in a hole in the heatsink. It's easier to fit than a string of diodes.

Diodes with higher reverse voltage ratings are intentionally manufactured with lighter doping so that the depletion region for a given reverse voltage is wider than it otherwise would be. The disadvantage with lighter doping is that the forward resistance and voltage drop for a high-voltage diode is higher than it would be for a lower-voltage diode.
Whether this is the case for modern manufactured devices, I do not know.

[R.G.]

The temperature coefficient of a PN junction (Vf or Vbe) is about -2mV/deg C. Either Diodes or a transistor can be used for bias compensation. I've used a TO92 transistor stuck with thermal compound in a hole in the heatsink. It's easier to fit than a string of diodes.

More importantly, a single bipolar can be used as an amplified diode, producing a bias voltage that can be adjusted for the exact voltage you need for compensation, as well as tempco, etc.

[Dave H]

More importantly, a single bipolar can be used as an amplified diode, producing a bias voltage that can be adjusted for the exact voltage you need for compensation, as well as tempco, etc.

That's what I did