#1752

Поделиться
HTML-код
  • Опубликовано: 28 ноя 2024

Комментарии • 120

  • @w2aew
    @w2aew 10 месяцев назад +45

    Sometimes you have to watch out for the clear glass package - some characteristics like leakage, etc can be affected by ambient light.

    • @rfburns5601
      @rfburns5601 10 месяцев назад +6

      Now that would be cool, rerun the test while scanning a laser pointer across the diode.

    • @hardrocklobsterroll395
      @hardrocklobsterroll395 10 месяцев назад +7

      Every diode is a photo diode if you use the right light!

    • @egeotomasyon
      @egeotomasyon 10 месяцев назад +1

      Alan, I completely agree with you.

    • @Oldclunker-ge5zp
      @Oldclunker-ge5zp 10 месяцев назад +4

      In the old days (60s), the glass diodes (both Germanium and Silicon) were painted black. Even zener diodes were black. This seems to have vanished since the 70s. Perhaps we should all pay more attention to this and always have a sharpie/Edding with you :-)
      Now that there are so many fake parts around, wouldn't it be possible to tell apart Schottky-Silicon from Germanium by their temperature coefficient and rectification efficiency over frequency, e.g. 50/60 Hz vs. 28/145/433 MHz? Whatever you have available...

    • @IMSAIGuy
      @IMSAIGuy  10 месяцев назад +4

      yes, I did a couple videos about photocurrents. At HP we did not have the clear glass package 1n4148 diodes. but a custom -0033 part in the stockroom that was the same part but in a black glass package

  • @davidwillmore
    @davidwillmore 10 месяцев назад +15

    You had me at "it's an excuse to use my curve tracer".

    • @lmamakos
      @lmamakos 10 месяцев назад +1

      That curve tracer sure seems to be more useful that I expected such a thing would be..

    • @davidwillmore
      @davidwillmore 10 месяцев назад

      @@lmamakos for such a simple device, they are so handy so often.

    • @Oldclunker-ge5zp
      @Oldclunker-ge5zp 10 месяцев назад +1

      Not quite as simple actually. They can produce up to 10 or 20 Amps at least temporarily and up to 2000 V and up to 200 W. They can measure down to 100s of microvolts and 10s of nanoamps. Supervise power and current limits for the device under test. Safety lockout for the operator. Plus GPIB control on the more recent ones. A behemoth of relays switches transformers opamps fans. The manual says it's a 2-person lift. Right they are.
      Watch the video of the vintage Tek museum in Oregon. Or better go there in person. It's a dream come true.

  • @carl-l3z
    @carl-l3z 10 месяцев назад +26

    If memory serves me, 1N34 and their germanium cousins are point contact diodes, and can be identified thru the glass case by mechanical differences.

    • @Homer19521
      @Homer19521 10 месяцев назад

      ruclips.net/video/3ao0-gRYzBU/видео.html&pp=ygUdMU4zNCBhbmQgdGhlaXIgZ2VybWFuaXVtIHRlc3Q%3D

    • @Broken_Yugo
      @Broken_Yugo 10 месяцев назад +2

      This is also my understanding, anything in that small glass package that 1N4148 uses is a silicon diode of some description. Germanium RF detectors in glass have the larger package and visibly point contact arrangement

    • @d614gakadoug9
      @d614gakadoug9 10 месяцев назад

      My apologies. I just wrote the same thing, before reading your comment. I didn't mean to rehash your remarks.

    • @bayareapianist
      @bayareapianist 10 месяцев назад +1

      ​@@Broken_Yugonot only that... If I remember correctly the direction marking for silicons were black and germanium ones were red. But I am sure about the Japanese germanium transistors were made in aluminum cylindrical with dots near their collectors. A blue dot means an NPN (very hard to find) and a black or red means a PNP (very common).

    • @carl-l3z
      @carl-l3z 10 месяцев назад

      No worries ...

  • @jameswkirk
    @jameswkirk 10 месяцев назад +1

    Years ago a company (Germanium Power Devices, I believe) was trying to develop a germanium Schottky diode, with something like 0.2 V forward voltage to reduce losses in switching power supplies. I think they also looked into germanium FETs. They're still around doing Ge and InGaAs IR photodetectors.

  • @d614gakadoug9
    @d614gakadoug9 10 месяцев назад +3

    The very fact the the the supposed 1N34s are in DO-35 packages with "red" wire apparent inside the glass says they are not 1N34s.1N34s are point contact diodes. To the best of my knowledge you simply can't construct a point contact diode in a DO-35. The sort of $-shaped contact wire should be visible in the real thing. The real package is a tube, not a solid piece of glass.

  • @jweismiller
    @jweismiller 9 месяцев назад +1

    Can you show how the old copper oxide meter rectifier diodes compare to Schottky & Germanium? I've heard that there is no modern device equivalent to them.

  • @d614gakadoug9
    @d614gakadoug9 10 месяцев назад +3

    Reverse leakage of Schottky diodes at low voltage is a very important and VERY limiting factor in lots of circuits.

  • @PeopleAlreadyDidThis
    @PeopleAlreadyDidThis 10 месяцев назад +1

    For what it’s worth, I checked an ST Micro 1N5711 on my 576 just now, same instrument settings. Its curve is more vertical just beyond the knee, closer to 30 degrees from vertical where yours is more like 45. It is then almost ruler straight resistive. It doesn’t have the “it’s a resistor, but now it’s being a diode” characteristic seen here.
    It’s in a blue package with black cathode stripe.

    • @IMSAIGuy
      @IMSAIGuy  10 месяцев назад

      I'm sure mine date way before ST Micro starting making them

    • @Oldclunker-ge5zp
      @Oldclunker-ge5zp 10 месяцев назад +1

      Keep in mind that anything that meets the 1N5711 Data sheet can legally be sold as 1N5711. Unfortunately for some users, the data sheets are pretty vague or ignorant about some characteristics that are important to some off-label usages like guitar effect pedals.

  • @johnwest7993
    @johnwest7993 10 месяцев назад +4

    I have a very large bag of old HP 1N5711's, so I find that they work well for everything. :)

  • @v8pilot
    @v8pilot 10 месяцев назад +1

    Using a high gain germanium transistor with collector connected to base will give you a lower forward voltage drop than a plain germanium diode.

  • @HA7DN
    @HA7DN 10 месяцев назад

    Just tested a few different diodes as peak detectors last week. We also had a hand full of unmarked glass germanium point-contacts, now I suspect they might as well be 1N34s.
    They performed the best at the highest test frequencies and smallest signal levels, but after that, suprisingly, LEDs worked well with high level RF.

  • @glasslinger
    @glasslinger 10 месяцев назад +2

    None of the diodes in the video are genuine 1N34. The 1N34 original point contact diodes were not manufactured past the early 1960's. The original diodes had to be individually adjusted so there was an access hole in the side of the case. The later genuine glass-metal used a low melting solder to hold the leads. The solder was heated to softening and the point contact was made and then the solder cooled quickly to lock it in place. Modern germanium diodes are made via a diffusion process that does not require individual adjustment of the point contact. All the diodes you have are "equivalent" germanium diodes.

    • @Oldclunker-ge5zp
      @Oldclunker-ge5zp 10 месяцев назад

      Does it mean that point contact Ge are no longer made? They had some unique features at low amplitudes. You could rectify a small RF signal without an amplifier and power supply.

    • @glasslinger
      @glasslinger 10 месяцев назад

      @@Oldclunker-ge5zp The original point contact 1N34 has not been made for over 50 years! The "equivalent" 1N34 mow made are all modern diffused junction germanium diodes. The original metal-germanium point contact had a noticeably lower threshold than the diffused junction type.

    • @edmaster3147
      @edmaster3147 8 месяцев назад

      You never cease to amaze Ron. Enjoy your video's very much. I find the old manufacturing and material choices are extremely interesting. And learning from the people who understand.

  • @M0UAW_IO83
    @M0UAW_IO83 10 месяцев назад +6

    1N5711 is a favourite of mine for building mixers but looking at that perhaps it might not be the best option.
    I don't have the luxury of a curve tracer so I might have to see if I can measure differences in mixer performance with different diodes.

    • @chodnejabko3553
      @chodnejabko3553 10 месяцев назад +2

      That is interesting. It seems like the weak knee would give you a soft-clipping effect, which might be a benefit in audio applications. What part of the circuit you use them for?

    • @inductr
      @inductr 10 месяцев назад +1

      You might have heard of homebuilt xy tracer and use 2ch scope.

    • @Oldclunker-ge5zp
      @Oldclunker-ge5zp 10 месяцев назад

      Curve tracers like the one shown here use the mains frequency 50/60 Hz and also at DC. They are originally intended for characterization of small batches of diodes and transistors at semiconductor companies for the purpose of writing the data sheet. I have not seen a curve tracer running in the kHz or MHz range.

    • @analog_guy
      @analog_guy 10 месяцев назад +2

      The 1N5711 diodes are good for high-speed mixers and other high-speed applications. They have much lower capacitance and much lower carrier lifetime (low carrier lifetime is useful for high-speed switching) as compared to the BAT85 or BAS85 diodes. Which diode is "best" depends on the application.

    • @keithking1985
      @keithking1985 10 месяцев назад

      Yes what part of your circuitry are you using them??

  • @AC9BXEric
    @AC9BXEric 10 месяцев назад +1

    Old parts like that are interesting. Imagine those old 5711s are in a guitar distortion effect thingy. And then one day it fails with a bad diode. And you go buy a 1N5711, pop in in there, the circuit comes back to life but it doesn't sound much like it used to. It was depending upon the characteristics of the original part and it just isn't available anymore.

  • @MrVeryCranky
    @MrVeryCranky 10 месяцев назад +2

    Before the 1N34 was the original OA91 germanium diode.

  • @tomstrum6259
    @tomstrum6259 10 месяцев назад +1

    I have an Elinco 550TK 7 Transistor AM Superhet demo board I built with Sockets for the Transistors & the 455 kHz IF detector Diode....I tried Various Diodes including severalb color Led Diodes....The Leds look Really cool functioning as IF audio Detectors & Sound great !!

  • @ahmadhasan8607
    @ahmadhasan8607 10 месяцев назад

    I found this video very useful. I've heard that we can measure the turn on voltage of a diode like a glass diode by using a digital multimeter, which is set at diode metering position. Whatever value the DMM reads corresponds to turn on voltage of the diode. Is this a reliable method for this purpose. However, I don't have any curve tracer at home.

  • @chaosopher23
    @chaosopher23 10 месяцев назад

    I have a rather large assortment of diodes. I noticed that the old ones are clear glass with/without a stripe, while the new ones are pinkish with a black stripe. I need a 30x magnifier to see the numbers, or I might use a 1N914 when I need a zener. They look the same! I wonder what use that weak knee is good for?

  • @Yuēhàn24
    @Yuēhàn24 10 месяцев назад +2

    The large glass diodes look more like an OA90/91.

  • @ericmc6482
    @ericmc6482 10 месяцев назад

    I really appreciate your approach and your presentations, thank you so much.

  • @KeritechElectronics
    @KeritechElectronics 10 месяцев назад +1

    Real deal germanium diodes, with a soft knee and all... Good to have a curve tracer around!

  • @chickenbiscuit4525
    @chickenbiscuit4525 6 месяцев назад

    IN34's are basically an antique.
    Might be easier to hunt for lost treasure.
    BATS are silicon, like pretty much everything in between.
    So hard to control a knee.
    Perhaps without the varicap/diode.
    But they become mlre expemsive than the fundamental components.
    D18's xX are still pretty source reliable.
    Today 30c tomorrow 30 bucks...

  • @Chris_Grossman
    @Chris_Grossman 10 месяцев назад +1

    The bad 1N5711 diodes have a high series resistance with the shottky portion, when the voltage gets high enough the protection ring diode turns on.

    • @paul_AAA
      @paul_AAA 10 месяцев назад

      Yeah, its reverse characteristic definitely looked like floating guard rings were turning on, but as for the forward curve it was more like a bad quality Schottky contact. An influence of floating rings on forward curve is negligible.

    • @Chris_Grossman
      @Chris_Grossman 10 месяцев назад

      @@paul_AAA The knee in the curve at 0.68V is the guard ring turning on. The portion from 0.4V to 0.68V is the high resistance Schotkey.

    • @paul_AAA
      @paul_AAA 10 месяцев назад

      @@Chris_Grossman Are we talking about the same guard rings? What I meant is there are floating guarding rings which are close to the Schottky contact but not on it or inside or whatever. The point is they are not under potential when forward current flows, so they have no influence on forward curve (they are just outside). Those p+ areas are under potential only when electric field reaches it after definite reverse voltage. But I saw such curves many times when we couldn't make a good Schottky contact formation process. If the process was dirty, there were always defects in metal structure and a thin layer of SiO2 between metal and Si. it doesn’t make too many problems in high current range because it’s mostly defined by wafer resitance, but it reveals itself on small forward and reverse current modes like it was shown in the video. If reverse current scale was less than 100uA (maybe 1uA), it would’ve been clearly seen I’m sure.
      But I might be completely wrong and maybe there are some guard rings that I haven’t seen yet which turn on in forward voltage mode.

  • @1kreature
    @1kreature 10 месяцев назад

    You should trace the curve of the diode in clear package when shielded from light and when exposed to high levels of light.

  • @nreitcheck
    @nreitcheck 10 месяцев назад +1

    Check them against the datasheets. You might be surprised.

  • @etmax1
    @etmax1 10 месяцев назад +2

    A true 1N34 is a point contact germanium diode and if you look through the glass at magnification you should a very fine S-bend of gold wire contacting the diffusion side. The other side should be bonded to a flatter surface.
    Schottky diodes are typically manufactured in a planar process so the gap between the 2 wires is in the order of the thickness of the die (plus bonding, and note this is for signal diodes).
    Schottky diodes leak a lot but notably they have higher capacitance than a germanium point contact diode.
    Germanium diodes virtually have to be done in glass, injection moulding is not supportive of point contact diodes.

  • @AB1Vampire
    @AB1Vampire 10 месяцев назад

    I had to substitute standard zeners for the BZX series on a recent project. I do not have curve tracing ability but the standard zeners did not work well in the circuit. (BZX79C15 for 1N5352 and BZX79C12 for 1N5349

  • @jensschroder8214
    @jensschroder8214 10 месяцев назад

    It is known that germanium bridge rectifiers are not only diodes but also resistors.
    Therefore, a silicon bridge rectifier and a resistor should be used as a replacement.

  • @robertclark8351
    @robertclark8351 10 месяцев назад +1

    Love this. I know it is mundane, but what is the test gear? (In case we see a curve tracer surplus and want to pick it up, etc.)

    • @milantrcka121
      @milantrcka121 10 месяцев назад +1

      I believe it is Tek 576 type. Rare and coveted. Bring Money...

    • @IMSAIGuy
      @IMSAIGuy  10 месяцев назад +1

      it is a curve tracer: ruclips.net/video/BabCunP9ib8/видео.htmlsi=jX6xTmGt796P2Yj5
      this one is a TEK 576 that I bought for $200 and took about a month to get it working again. I have videos of the repair on my channel

    • @Oldclunker-ge5zp
      @Oldclunker-ge5zp 10 месяцев назад

      Curve tracers are not indestructible. When testing 20Amp Motorola power Schottkys as an apprentice 40 years ago, magic smoke came out. It didn't break though. I think I had a 576. It wasn't until Recently that I read a warning in the user manual PDF...
      You should know what you are doing or carnage will ensue.

    • @IMSAIGuy
      @IMSAIGuy  10 месяцев назад

      It's only rated to 10A

  • @t1d100
    @t1d100 10 месяцев назад +1

    Any excuse to use the curve tracer is a good excuse.

  • @Madness832
    @Madness832 10 месяцев назад

    I was ponderin' how to pronounce the name of that store that sold the BAT-85s.

  • @ccshello1
    @ccshello1 10 месяцев назад

    At 7:50, a close up should be able to show the "point contact" structure (a.k.a. cat whisker) of the classic diode design.

  • @LutzSchafer
    @LutzSchafer 10 месяцев назад

    You're Schottkys look like going into negative resistance at the reverse breakdown and even show oscillation. Or was this an issue with your curve tracer?

    • @IMSAIGuy
      @IMSAIGuy  10 месяцев назад +1

      Breakdown is part avalanche and tunneling.

  • @herbertsusmann986
    @herbertsusmann986 10 месяцев назад +2

    Crystal radios work best with real Germanium 1N34 or similar. Next compare Ge to Schottky at very low fwd voltage levels. I believe you will find the Ge will act like diodes at lower fwd voltages than the Schottky.

    • @Bob-1802
      @Bob-1802 10 месяцев назад +1

      Diodes like 1N34 do not need any biasing to start demodulating a weak RF signal, that's why they are prefered in crystal radio.

  • @krisraps
    @krisraps 8 месяцев назад

    I Have MANY MANY OLD SOVIET Surplus Diodes, All Kind Of Them, Selling. I Have NOT Only That But Many Radio Things

  • @chrisscott1547
    @chrisscott1547 10 месяцев назад +1

    Your 1n34 package looks like 1n914 diodes to me.

  • @roboanalogtom
    @roboanalogtom 10 месяцев назад +1

    Hey, can you do a simple crystal receiver and show the difference of putting a slight bias on the diode? Many designs use a 1-10 meg resistor to bias the diode and I don't understand why. edit- I don't understand why the slight bias improves the performance of the crystal receiver

    • @M0UAW_IO83
      @M0UAW_IO83 10 месяцев назад +4

      If you pre-bias the diode it takes less received signal to get the diode into conduction so it makes the receiver more sensitive.
      Interesting experiment though.

  • @SkullDraizer
    @SkullDraizer 9 месяцев назад

    Many old things have weak knees.

  • @chodnejabko3553
    @chodnejabko3553 10 месяцев назад +1

    I'm really intrigued by the 1n5711. I'm sure there is an application where such soft knee would be desirable.
    I wonder how 1n34 compare to red LED's, since those are also used for audio clipping.

    • @hardrocklobsterroll395
      @hardrocklobsterroll395 10 месяцев назад

      I think they just might be old. He mentions in the video that they’re old so the technology process could have been less refined and introduced a resistive region from the junction. I don’t remember enough semiconductor physics to propose a mechanism though

    • @keithking1985
      @keithking1985 10 месяцев назад

      ​​@@hardrocklobsterroll395that could be a number of factors from less purely refined elements to the dopping process or even material thickness from machine calibration differences over the years.. would be interesting to find out the actual reason.. 😊

  • @Chupacabras222
    @Chupacabras222 10 месяцев назад

    Is it possible to differentiate those by reverse leakage current?

  • @silverXnoise
    @silverXnoise 9 месяцев назад

    Does anybody know what kind of matrix displays they’re using for the DRO on these old Tek curve tracers?

    • @IMSAIGuy
      @IMSAIGuy  9 месяцев назад

      custom from TEK, they are not fully matrix, but are plastic optical fibers that spell out things and illuminated with bulbs at the other end.

  • @microflite
    @microflite 10 месяцев назад

    Schiottky vs Germanium look inside the diode with a microscope. With Ge, you should be able to see the whisker and the point contact.

  • @adamlazarowicz9136
    @adamlazarowicz9136 10 месяцев назад +1

    Video focus on static characteristic of diodes. In radios or mixers capaticance plays rule too... There is a very good realy germanium diode , made in Soviet Union ( Russia ) name GD507A . It is GOLD BONDED , cat viskers HF diode. It easly work up 1GHz . In my opinion its same as 1N34 or better.Long life and rock solid. Rest of it come now from Ukraine ...

  • @michaellundsrensen2292
    @michaellundsrensen2292 10 месяцев назад

    The diodes you name baS85 - is that baT85's? I can't find any BAS85 diodes.

  • @iblesbosuok
    @iblesbosuok 10 месяцев назад +2

    06:28 is that noise at the falling edge?

    • @vikenemesh
      @vikenemesh 10 месяцев назад +2

      Yes, reverse breakdown is a very uncontrolled process in diodes not explicitly designed for this operating mode (zener diodes are optimized for this, for example). Every trace the curve-tracer takes results in slighty different characteristics (due to noise in the curve-tracer or self-heating of the diode) near the beginning of this breakdown region.

    • @iblesbosuok
      @iblesbosuok 10 месяцев назад

      @@vikenemesh thank you

  • @ThinklikeTesla
    @ThinklikeTesla 10 месяцев назад

    There are a lot of listings for "1N34" (some even saying "germanium") that are really Schottky, all in the smaller case with a black stripe, exactly like what you showed. I've always wondered whether these fakes would perform as well in situations where you'd typically really want Ge instead of Si, like the classic crystal radio circuit…

    • @bobweiss8682
      @bobweiss8682 10 месяцев назад

      They don't work as well as Ge diodes do. The Vf is comparable, but the Schottky diodes have much higher junction capacitance.

    • @Oldclunker-ge5zp
      @Oldclunker-ge5zp 10 месяцев назад

      For small AC amplitudes, eg some millivolts, Ge point contact works better. My circuit ran at approx. 200 kHz.
      The old HP RF rectifier probes 11096 for multimeters and oscilloscopes use Ge upto 500 or 700 MHz.

    • @bobweiss8682
      @bobweiss8682 10 месяцев назад

      @@Oldclunker-ge5zp According to the 11096 manual on the HP Archive webpage, the diodes in the probe are silicon, not Ge. With 2 different part numbers...

    • @Oldclunker-ge5zp
      @Oldclunker-ge5zp 10 месяцев назад

      I saw multiple versions of 11096 manuals, also with A and B suffix. There are subtle differences in the 2 diodes used. The Ge diode is used as rectifier and the Si diode is used to linearize the DC output over a wider range of amplitudes by introducing a non linear voltage drop, assuming you are connecting to a multimeter with 10 M input if I remember correctly.

  • @rfburns5601
    @rfburns5601 10 месяцев назад +1

    Hey, now do a tunnel diode, and see if you can keep it from oscillating in the test fixture.

    • @stamasd8500
      @stamasd8500 10 месяцев назад

      I was going to suggest exactly that.

    • @IMSAIGuy
      @IMSAIGuy  10 месяцев назад

      I've got some on order

    • @Oldclunker-ge5zp
      @Oldclunker-ge5zp 10 месяцев назад

      Will oscillate. Even DIACs , 4-layer diodes and high-gain bipolar transistors will oscillate.
      Good luck though ;-)

  • @mk6595
    @mk6595 10 месяцев назад

    For those without curve tracers, see "How to discern Germanium-Diodes from Schottky-Diodes" here on YT: ruclips.net/video/Lyos1D7_2OI/видео.html

  • @keithking1985
    @keithking1985 10 месяцев назад

    I subscribed to this channel a few days ago and I'm very impressed with the content. I'd like a curve tracer too.. 👍🇮🇪🙏👍🇮🇪🙏

  • @fromgermany271
    @fromgermany271 10 месяцев назад +2

    So for clipping where the reverse leakage is out of range, modern Schottkys are a cheap replacement for germanium.
    BTW, Si and Ge differ in the absolute values of the curve point, but not in shape, so guitar pedals just need to be adjusted in range to replace Ge by Si.
    Purity extremists might disagree. But they also use mp3. 😂

  • @W4BIN
    @W4BIN 10 месяцев назад

    Should have looked at several 1N5711s, you could have picked a bad one, that one was NG.
    I never saw one that acted that way. Ron W4BIN

    • @analog_guy
      @analog_guy 10 месяцев назад

      I also at first thought the 1N5711 might be defective, but I looked up the data sheet (HP data sheet) and found 410 mV max forward voltage at 1 mA, 1V max forward voltage at 15 mA, so it looks like the diode meets spec. Its forward conduction curve also appears to be reasonably close to the typical curve shown in the data sheet.

  • @pibbles-a-plenty1105
    @pibbles-a-plenty1105 10 месяцев назад

    1N7511?

  • @pvc988
    @pvc988 10 месяцев назад

    PER
    V
    E
    R
    T
    DIV

  • @robinbrowne5419
    @robinbrowne5419 10 месяцев назад

    I ordered 5 1N34 diodes from amazon for my crystal radio and got 5 unmarked schottky diodes in clear glass packages. This is the classic fraud. I should have known better since I now know that real 1N34 diodes are as rare as hen's teeth :-)

  • @RichardFraser-y9t
    @RichardFraser-y9t 10 месяцев назад

    I still prefer the geranium diodes

  • @AnalogDude_
    @AnalogDude_ 10 месяцев назад

    1.5 cents each?

    • @IMSAIGuy
      @IMSAIGuy  10 месяцев назад

      yup

    • @AnalogDude_
      @AnalogDude_ 10 месяцев назад

      How do they do that?
      the gas to melt the glass would be worth more, i guess.
      @@IMSAIGuy

  • @Hopeless_and_Forlorn
    @Hopeless_and_Forlorn 10 месяцев назад

    I'm old. Do an OZ4.