Are these real 1N4007?

[Alex Eisenhut]

The diode with the longer thicker leads is a vintage 1N4004 from my crap.
The one with the shorter thinner leads in a 1N4007 from some random eBay seller.

Both diodes test as diodes, 0.55V forward, open in reverse with a DMM.

Those thin leads look more like cheap resistor leads. Are there legit 1N4007s built with such skinny legs? Surely there can't be any money in making counterfeit 1N400X diodes ??

[bob91343]

Take a sample of the questionable diodes and measure the leakage at the specified voltage, either 1000 V or somewhat higher, depending on the details in the EIA specification.  If they pass, you have good diodes.

I can think of no better way to verify the quality of what you have.

[Alex Eisenhut]

Yes, I'll have to try that next, only with 100V to start. But these skinny leads might also be legit cost reduction. I've seen new TO-220 packages from legit sources with a very thin metal tab, compared to the almost 1mm thick of my vintage parts.

[coromonadalix]

To me the printing is ugly and not defined on your "thin" leads diode, i never saw any "bad" printing as this,  i use in4004   and the leads are "thick" 

[Alex Eisenhut]

Yeah, the printing is thicker than the leads on the 1N4007. It might be a 1N4148 with an epoxy fake body molded over it... 

[Teledog]

Careful!
Picked up some "supposedly" 10 amp diodes at an electronics surplus place in Toronto years ago.
They certainly looked beefy, but..no,..all were junk..shorts &/or no rectification whatsoever!
Also,.. ordered "supposedly" germanium diodes from china..not!!!
Unless you're ordering from Digi-key ...Test them all!
(even at that, have had a few parts packaged/mislabeled..really!)

[trobbins]

The die may pass all vintage 'datasheet' specs, except the package may not exactly align with the thermal ratings (due to the change in conduction by the leads).  As suggested, use an insulation resistance megohmmeter at 1kVdc to confirm no measurable leakage from PIV.  So I'd suggest they are legit, but like any design, you need to relate the spreadsheet average curremt rating to the applications current waveform and then die temperature, which is not trivial for many rectifier applications.

UF4007 from ebay certainly have the slightly reduced lead diameter.

[0culus]

Careful!
Picked up some "supposedly" 10 amp diodes at an electronics surplus place in Toronto years ago.
They certainly looked beefy, but..no,..all were junk..shorts &/or no rectification whatsoever!
Also,.. ordered "supposedly" germanium diodes from china..not!!!
Unless you're ordering from Digi-key ...Test them all!
(even at that, have had a few parts packaged/mislabeled..really!)

Nobody makes germanium diodes anymore (as far as I'm aware), so any you see for sale "new" are usually Schottky diodes.

[rsjsouza]

Nobody makes germanium diodes anymore (as far as I'm aware), so any you see for sale "new" are usually Schottky diodes.

NTE Does (NTE109, NTE110A, NTE110MP, etc.)
https://www.nteinc.com/diodes/general-purpose.php

[bob91343]

I was gifted with a box of resistors and diodes mostly.  All look good and clean.  A bunch of 220 Ohm resistors all measured way out of tolerance.  Included were perhaps 50,000 diodes, 1N4148 which can serve as rectifiers in many places.

Years ago I built a replica of a decimal counting unit to plug into an old counter.  It used 10 tubes and many other parts.  When finished, it didn't work.  I was gobsmacked at the thought of troubleshooting it but bit the bullet and started testing.  I found one diode marked backwards.  Replaced it and the unit worked, nothing short of a miracle.  I was torn between reversing the diode and replacing it and decided replacement made sense to simplify future troubleshooting.

So don't assume if things look a bit odd that someone is trying to cheat you.  The world is filled with surprises, many of which almost refute logic and common sense.

[Pawelr98]

I have had some 4007 or similar diodes go short circuit in circuits where the backwards voltage was around 800-850V or so, still under 1000V.
They could in fact work for several days before shorting.

So try and apply a reverse voltage close to 1kV and let the diode sit there for several hours.

[0culus]

Nobody makes germanium diodes anymore (as far as I'm aware), so any you see for sale "new" are usually Schottky diodes.

NTE Does (NTE109, NTE110A, NTE110MP, etc.)
https://www.nteinc.com/diodes/general-purpose.php

Hah, TIL. Nice. 

[amyk]

At the scale these things are made, I suspect the reduction in packaging and lead diameter made a significant reduction in cost.

[T3sl4co1l]

That might well be out of spec.  At least, one or the other, the wire diameters don' t look like they can both be in tolerance:
https://en.wikipedia.org/wiki/DO-204
DO-41 specifically (DO-204AL).

The leads dissipate a significant amount of heat; I'd be worried the thin one can't handle rated current at max ambient.  Easy enough to check, with a power supply and ballast resistor.

Tim

[Tom45]

Regarding testing the diodes with 1 KV.  These days I wonder how many of us have a power supply that can output 1 KV.

[trobbins]

Need to use a 'megger', or alternately known as an insulation resistance tester.  Some countries require that type of test on any mains ac circuitry as part of verification of installation, so the meters are quite common (but the meters are not common in countries that don't require much in the way of testing  , such as USA).  The modern testers typically have 250, 500 and 1kV DC levels and are current limited, and will confirm diode has reverse leakage below at least 0.5uA for a 2Gigohm max reading (vintage datasheet nominal level is 0.3uA at 1kV at 25C).

[exe]

These diodes are produced by so many manufacturer (and I suspect many of them just rebranded), so just by appearance it's hard to judge. I suggest just buy parts from a reputable distributor and don't bother with guesswork and testing. Although, I like testing components, it's fun by itself, but sometimes gets boring when I want things to "just work".

[Pawelr98]

Regarding testing the diodes with 1 KV.  These days I wonder how many of us have a power supply that can output 1 KV.

A lot of folks have it in their homes without even realizing it.

European folks can just use a doubler from 380-400V 3 phase mains which are very common.
US folks can get away with a tripler from 240V mains.

From there just make a voltage divider using resistors or use series zener diodes to adjust to 1kV.

You can build that in several minutes.

[0culus]

Regarding testing the diodes with 1 KV.  These days I wonder how many of us have a power supply that can output 1 KV.

A lot of folks have it in their homes without even realizing it.

European folks can just use a doubler from 380-400V 3 phase mains which are very common.
US folks can get away with a tripler from 240V mains.

From there just make a voltage divider using resistors or use series zener diodes to adjust to 1kV.

You can build that in several minutes.

I hope you have some kind of current limit on that, like at least a fuse that won't let you dump 15 or 20 amps (or more) through a short.

[T3sl4co1l]

Well, an isolation transformer might be highest priority!

Half-wave type multipliers are AC coupled, i.e., mains flows through a capacitor, so the value of that capacitor limits the peak current even under short-circuit conditions.  Using say 10nF caps limits it to a few mA.

You need a series resistor anyway to limit avalanche current, in case it breaks down at a lower voltage; some megs is a good idea at least as a starting point.

You probably also need a specialized meter, or a dropping resistor for a more conventional one: most don't read over 1000V, or are rated to tolerate such.  A 10M DMM (or VTVM if you're old school) with a 90M 1% resistor in series will give a 10x reading easily enough.

Tim

[Kleinstein]

The thinner leads may be just cost reductions. However much of the heat has to go through the leads. So I would not use the thin lead version for higher currents, like > 500 mA.

There is a chance they are fake ( someone here even reported fake 0 ohms resistors - so there is nothing to cheap to get faked). If  fake, it could be lower maximum voltage, more series resistance (so compare at some 500 mA with good cooling) and also more leakage. The original 1N4007 are usually quite slow, which helps to get low leakage. One could do a quick reverse recovery check  (scope and a +-2.5 V with 50 Ohms impedance rectangular signal): 1n400x are more like 1-10 µs, most small diodes like 1N4148 are < 50 ns.

[CJay]

I dunno, I've seen weird off brand 1N400x diodes with thin legs like that, question is, would you want to take a chance on them or is it better to just dump them and buy something from a reliable distie?

Especially as they're mains rated.

[Haenk]

European folks can just use a doubler from 380-400V 3 phase mains which are very common.
You can build that in several minutes.

Erm, at least by german standards (and probably EU standard by now), 3-phase stuff requires handling by a certified electrician. For a reason.
However a used DIN tester is rather cheap on ebay, and even if out of certification, it will still give you 1kV without the need to tinker around and try to blast yourself directly to Heaven's Gate.

[cdev]

 I bought a bunch recently to experiment on as they can sometimes sub for PIN diodes in circuits. I was thinking they could be used as RF switches, in things like preselectors and similar, Maybe there would be a way to exploit that property, if present, to evaluate the likelihood of them being bogus without needing the high voltage?

[T3sl4co1l]

Well, for PIN purposes, you can test them to only whatever voltage you need, and can measure reverse recovery easily enough, and capacitance.

Tim