TESTING DIODES FOR CRYSTAL RADIOS

[@ptronix]

Very interesting results Ron, my favorite diode here in the UK is the oa90 & others in that series. Work great in crystal sets. By the way years ago I ran an lcd clock off a crystal set, I had so much signal from the local 198 kHz 500 kW transmitter. It would also light a red led

[@bryanlatimer-davies1222]

Oa91 here in Wales!

[@Rich-on6fe]

A clock powered by the cricket!

[@australischemediengemeinschaft]

Ein sehr aufwendiger übersichtlich dargestellter Diodenvergleich. Die richtige Herangehensweise, um die höchste Empfindlichkeit für die AM-Detektion zu erlangen.

[@rouxenophobe]

I have just built my first crystal radio, which disappointingly didn't initially work with some eBay 1N34A's, but I replaced it with a different 1N34A (from Mike's Electronic Parts, Ohio) and it worked! Really stoked!! This video was really interesting to see the differences between the diodes and to understand the effect of bias.

[@mnpd3]

When I was a kid building crystal radios c. 1960 I would make a detector by tickling tarnished copper pennies with pencil leads. For a condenser (capacitor) I would use sheets of aluminum foil sandwiched between paper. I would have been astounded at your setup here. :)

[@apexmcboob5161]

Very interesting. Though I knew the Ge would win I was surprised at the effect of the bias voltage. Thanks!

[@AliasUndercover]

I've learned more about electronics from you than anywhere else.

[@cyndicorinne]

I knew that this would be interesting as soon as I saw that you brought in the germanium diodes. My dad always said they were better than silicon for the crystal radios that we used to make, and I used to make such projects from the x-n-1 project kits from Science Fair (later just called Radio Shack). Of course with the kits it was clear because the silicon diodes would only pick up a very close radio station we had nearby, but the germanium would catch multiple AM stations in town at the time. Excellent quantitative analysis!

[@MisterTalkingMachine]

About ten years ago when I was in high school I built a small crystal radio with a mini solar panel for diode bias, and I set out to try it with every single diode I could get my hands on. I tried it with all manner of silicon rectifier and signal diodes, shottky diodes, old varactor diodes from TV tuners, cat's whiskers with a variety of minerals including pyrite, galena, chalcopyrite, bornite, and silicon dies from busted chips. I even tried out LEDs and laser diodes. I found that LEDs worked pretty well with the strong stations I had and a AA cell in series with them for bias. I didn't have any germanium diodes at my disposal back then. I am really missing the time when I had the time to just build radios for fun, besides most AM stations worth their megahertz being long gone.

[@Rich-on6fe]

Thank you for this. Always wondered what the 'right' choice would be. I remember getting my first crystal radio working and hearing Billy Jean. I didn't know the difference between Si and Ge back then and was building out of junk and and it rarely worked.

[@SeanBZA]

Yes bought the 1N34A on fleabay, and of course they are Shottky. Still have a few old Sharp Compet boards though with the nice Sharp made Germanium diodes, that still do work for use, the 34A's are used as higher level diodes, though with 40V reverse they do handle high voltage better.

[@borisj4054]

Look for a GEX 34. An English version.

[@glasslinger]

The applications I use the 1N34 are crystal radios, where reverse voltage is not a problem. Ideally you want a point contact diode rather than the modern junction diode. The 1N34 diodes you buy now days are germanium junction diodes, which are the same characteristics as a 1N270 or any other germanium junction diode.

[@franzliszt3195]

Same happened to me. I just checked ebay, all but one are shottky. And that one, $14.

[@gordonwelcher9598]

It would be interesting to try a vacuum tube diode like a 6AL5 or 12AV6. Or maybe one of the 1.4 volt tubes. Maybe a vacuum fluorescent display would do something. The load would have to be higher impedance. The frequency response would be good to check.

[@tipfox9212]

I built one several years ago with a russian 6D6A - subminiature diode. It works ok but I found ge-diodes to be better ;-)

[@johncasteel1780]

I think that a 1.4 volt tube with its directly heated 50 ma filament would not be as good of a detector as the directly heated 6.3 volt tubes.

[@tipfox9212]

@@johncasteel1780 Actually there is no big difference - tried both. To be honest my tube detectors where pure 'fun projects'. In general it doesn't make any sense due to the heating and because they are simply not designed for that... If you want to use a tube there are many good circuits for regenerative audions even for low voltages. They are WAY better ;-)

[@metalman4141]

We keeping you up Ron !

[@repairitdontreplaceit]

great work ron , always great to see more content from you

[@radiofun232]

Beautiful demo, very informative and convincing!

[@borisj4054]

Those CK705 and 6 are old Russian made diodes. Believe I have a data book that covers them. I collected lots of exotic semiconductors from street markets in Saint Petersburg and Moscow during the 90’s.

[@johncasteel1780]

They also could be old Raytheon parts. The CK705 was a Raytheon equivalent of the 1N34. In the '50s, Raytheon also introduced the CK709 as an RF transistor and the CK722 as an audio transistor to the general public.

[@321CatboxWA]

I was just looking at building a crystal set yesterday ! I was looking at an old tickler and I have a few Nos 1N34's . Thank you ! PS ESP is a real thing . No such thing as a coincidence . 73's You rock ! If you need a vintage component I have , It's yours !

[@flatbrokefrank6482]

Great content as always Ron !

[@ShadowsOnTheScreen]

Did I understand all of it? Nope. Did I learn some stuff? Yup. Thank you!

[@kornami8678]

Loved your video. Well done. I have a Hallicrafters SX-130 shortwave receiver. It uses a 6AL5 vacuum tube detector. I replaced the 6AL5 with a 1N34A germanium diode. I found that with the germanium I had to turn up the audio gain control to get the same audio level as with the vacuum tube. Interesting! I pulled the 1N34A out and put the 6AL5 back in. It seemed to me that the 6AL5 is more sensitive.

[@glasslinger]

The 6AL5 cathode emits with zero volts on the plate so there is no threshold. Thus, the output will be slightly higher. Good ole tube stuff!

[@karlbesser1696]

The 1N60 diode you can buy today is a Schottky diode. This diode used to be manufactured using germanium technology. Many greetings.

[@davidstacy8314]

Hi Ron your tech friend Dave here another good show I enjoy all of them your professional work is amazing through the years I bought a lot of people out I have every tube every dial and transistor out there if you ever need any can't wait to your next video may God bless you and keep you safe always thank you

[@roman_abelardo]

Excelente aporte, muchas gracias 👍 Saludos Cordiales desde Argentina 🇦🇷👋

[@tomstrum6259]

Thx for putting up this really good informative video.....Now I know why the 1N34's have been my best performing DX detector for past 60 yrs !! Glad you ran the foward bias sensitivity measurements also....A very complete test to "settle" the decades of misinformation.....Well done !!

[@priestblood]

Great knowledge thanks Ron

[@cyndicorinne]

This is an excellent reference video. I built a radio just last night using a 130-in-1 project kit, and this kit is new containing to purported 1N34A Ge diodes. The strange thing is that they’re unmarked except for a distinctive opaque bright blue color with black cathode stripe. It actually works pretty well at night and is able to pick up a 750W directional station on MW band. Still, I’m not certain what those diodes are.

[@analog_guy]

You may have a 1N270 germanium diode like the one tested in this video. The 1N270 is electrically very similar to the 1N34A except the reverse voltage is rated at 100 volts as compared to 65 volts for the 1N34A. The 1N270 diodes I have are blue with the black cathode stripe, but they also have ITT printed on them, indicating the manufacturer.

[@cyndicorinne]

@@analog_guy Thank you!

[@moristo]

So now I think a lowest forward voltage diodes is the good as AM detector diode, thank you Ron for the info.

[@ianbutler1983]

Cute cat, Ron.

[@davidpacholok8935]

Dear Ron You might want to add a .002 uf or so across the 100k load resistance to act as a low pass filter. There is RF on the scope as well as audio. The peak to peak audio measurements are made less accurate by the RF. Higher diode Capacitance raises the p-p rf voltage. Love your channel! Dave

[@glasslinger]

Yes. I tried a capacitor and found that the total output signal stayed the same. The chart may be a few millivolts off but the comparison is good. The 1n34A is clearly the champ! But genuine germanium diodes are close to the same.

[@tomstrum6259]

Big Thx for taking all the Time to share this Rare data !! ....Ch 2 audio looks really "Thick" ...Perhaps a 5 mH + 500 pF LC Low pass Filter to Scope ch 2 to clean up the detected Audio signal voltage ??

[@Joe-ep9lk]

Dang happy to see you again :):)

[@JeanJacquesDuval]

Hi, Really very interesting. I also made some measurements on the diodes but with a slightly different approach. Indeed in France we no longer have AM and the only possibility of receiving something are the very distant stations in MW or SW and BBC at 450km in LW with LF amplification and listening on loudspeaker. These are levels below 10mV in general. For me the threshold voltage measurements generally given at 1mA are not representative of the conditions in the receiver. So I made measurements at 10 and 50 µA. I found a very good correlation between the listening level in the speaker and the results at 10µA. I am a little surprised by your results for galena because for me it gives the same results as 1N34, OA79, BAT85 and this for fairly homogeneous sz threshold values ​​of 0.1 or less at 10 µA. At 0.15 it is weaker (1N60) and very weak at 0.2 (1N1711)If you are interested, a video is on my channel, (The BAT is not there yet)

[@ParedCheese]

It would be interesting to see how a germanium transistor junction would work as a detector diode. 🤔

[@SeanBZA]

They work well, basically a biased diode with the transfer resistance providing current gain. Even silicon transistors work well here, and are the basis of all integrated AM detectors, where the bias voltage is provided by another identical base emitter junction used as reference voltage, and a resistor to provide the input bias.

[@reel2reel.]

Always good to see your videos. Too bad there are no electronic wizards(repair person) down here by Corpus...I have a solid 2 state audio amplifiers that needs repair....and nobody with knowledge to work on them. Wish you were closer.

[@msf60khz]

I have also used a coherer detector, which is a bi-direcrtional non linear device. I found the Mercury-Iron coherer was approx 10dB inferior to a Ge diode.

[@glasslinger]

The coherer is basically a RF triggered switch. It is not meant to be a square law detector device. Any AM detection you happened to get was circumstantial with no good logic to explain it.

[@graemezimmer604]

@@glasslinger Hi Ron, I enjoy your videos very much. As you say, the classic Coherer was intended to act like a relay with a battery, eg ringing a bell when it detected RF. However it is not widely understood is that it also was used as a Square-Law detector in series with headphones. In fact Marconi's notes show that he used a coherer and an earpiece for his famous crossing of the Atlantic. (do a search on Marconi, Atlantic, Collier-Marr earpeice, Italian Navy Coherer) It makes sense, because a coherer is just a dissimilar metal contact (iron filings, mercury, etc).

[@msf60khz]

@@glasslinger There seem to be two modes of operation of coherers, either as a switch or as a "linear" device. Marconi used the mercury-iron coherer in conjunction with an earpiece for detecting radio waves across the Atlantic. I find that most metals will work; we see a bi-directional square law curve, and it is supposed that it is the result of electron tunneling across a very thin insulating layer of oxide.

[@glasslinger]

@@msf60khz Using instruments, I was not able to get any rectification at all. It was non-conductive up to a point and suddenly switch to lower conductivity.

[@dennisqwertyuiop]

great test

[@clytle374]

Good video. Confirmed what I've been reading, hard to get a good modern diode for a crystal radio. Was disappointed you skipped the foxhole radio with a razor blade.

[@glasslinger]

I wasn't able to get one to work. Possibly need an old razor blade.

[@clytle374]

@@glasslinger should have known you tried it

[@doccpu7]

I used green pennies touched by a safetypin in the 60's.

[@401ksolar]

That bias does not have to source very much current, so how about a separate "noisy" antenna connected to a voltage multiplier gm diode/capacitor bank for RPOA remote power over air (similar to other clandestine devices) ?

[@SeanBZA]

Use a solar panel from an old calculator, which, with the on circuit board voltage regulator diode, will work, plus you also get a germanium or shottky diode there as well to use.

[@danhorton6182]

Awesome series. I’d love to see something similar with a bunch of different Russian germaniums. I’m sorry that your friend didn’t have the best of luck with his. I bought several from a fairly well known and trusted seller out of Ukraine, I got their equivalent of the 1N34, 1N34a, 1N60, and a few others that folks on the forums say are exact equivalents of some of the ones we use. I also got several D18/1S307 blue tip “holy grail” diodes. Not sure what makes them “holy grail”, but many call them that. Anyways, maybe the difference was the price as the ones I picked up were far more than 20 for $1. They were about $12 after shipping and you got 30 of them, the D18s were a bit more. Since I got several different kinds he did combined shipping and they ended up being closer to $9 for 30 diodes. I tested every single one and not one dud. I measured them at different currents and they all had fairly tight forward voltage groupings, meaning per device number. Much better pricing than the $12-$15 EACH for the 1N34 and 1N34a that people are asking. That’s crazy pricing. I was able to get 40-50 diodes for the price of one essentially. I wish the authentic ones that were used in the US could be found cheaper. Anyways, would love to see a comparison.

[@glasslinger]

Note that all modern diodes (except microwave diodes) are grown junction diodes. The original early diodes were point contact diodes where the "junction" was a microscopic point where the needle tip touched the germanium. Thus, you will not find any of the modern diodes that will have the sensitivity of the original diodes. Point contact diodes are no longer made because each diode has to be individually adjusted. Not practical in production quantities at any price you would pay. There are no modern uses that demand the sensitivity of point contact diodes to make production practical. Thus, buying Russian diodes are no different than buying diodes from Mouser.

[@danhorton6182]

@@glasslinger yes, that is what I purchased I believe. I wish I could post pics. Magnifying the diode you can see through the glass body a very thin wire that kind of curly q’s as it comes to the very fine point that touches what I’m guessing is germanium on the other side of the diode. I’m using the microscope I use for smd soldering to see it, it’s a very tiny wire. Could I possibly send you some photos that you can confirm? If they are good early ones I’d be happy to send you some if you have any desire to test them. These are definitely old stock. Their build suggests several several decades ago, but I know Russia was making this stuff after us.

[@OverUnity7734]

Years ago I heard that Phosphor bronze was best for a cat's whisker. Phosphor bronze can be found in some guitar strings.

[@glasslinger]

I have some wire. I'll give it a try.

[@franzliszt3195]

I would think silver wire would be best, not really due to the lowest resistance, but it conducts well even if oxidized to black.

[@KB4QAA]

@@franzliszt3195 Phosphor Bronze is very 'springy" and won't deform like copper or silver wire.

[@tipfox9212]

Nice demonstration of the usabillity of different diodes for crystal radios - thank you! I liked especially comparing the results at 1/10 MHz. The resullts (absolute values) clearly will change with every different setup of the crystal set (due to several reasons) - but the tendency seams clear: Ge is the best ;-) May be you could use an old Ge transistor (EB/CB-Diode) if you don't have any Ge-diode ...

[@johncasteel1780]

Probably not. For detection, a point contact diode should work better than a semiconductor junction.

[@moristo]

it's very important things how to use diode

[@jozef1961ful]

Robiłem podobne doświadczenie i w.g nich, najlepiej w obwodzie detektora sprawdziła się dioda germanowa serii DOG 58 polskiej produkcji. I did a similar experiment and, accordingly, the germanium diode of the DOG 58 series of Polish production worked best in the detector circuit.

[@graemezimmer604]

Interesting video Ron, but I'm afraid you rather missed the point with bias. All forward bias does is changes the impedance of the diode (eg moves the operating point). So for instance a silicon diode (which has a very high impedance) can be forward biased to operate in a lower impedance circuit. And the 1N34 (which is fairly low impedance) can be steered to better match the impedance of your test set (and headphones). To put it another way, if you arranged impedance adjustments on the input and output, you could get a 1N34 to operate at optimum sensitivity with zero bias. Consequently it is misleading to test a number of diodes in a given circuit. All that does is show which particular diode suits your circuit best. A more useful test is to design a circuit with impedance taps (in and out) and find which diode likes which impedance. There was an exhaustive study on this on TheRadioBoard, but unfortunately that has long gone.

[@radiofun232]

The impedance adaptation (to get the best AM audio signal out) regards all radio detection diodes. The test setup in this case is/was based on practice and shows convincing results on the oscilloscope. It is in my opinion a proper test setup. The 100 K resistor at the output of the test circuit is a well chosen and practical value, just like the used (AM modulated) HF test frequencies.

[@graemezimmer604]

@@radiofun232The problem is that those "convincing results" are misleading. For example it shows that a 1N34 needs bias. But it wouldn't if the input and output impedances had been optimised. If a diode benefits from bias, it merely shows the impedances need adjusting. All that bias does is change the impedance, it doesn't (and can't) change the sensitivity. The other issue is that without a tuned circuit, the capacitance of the diodes can not be tuned out, which can make a big difference. Whatever, these points can be easily demonstrated with a better test set.

[@radiofun232]

@@graemezimmer604 In my opinion and radio experience (made tens of SW radio receivers, watch e.g. my YT Channel) there is nothing wrong with this test setup. The same setup is/was used for testing different diodes, that makes a comparing possible. 100 K is a good value as "end-impedance" parallel to which the probes of the oscilloscope are connected. When you make a crystal radio and give the detection diode some bias (like Glasslinger did) you can hear (and see on the oscilloscope) that the AM radio detection can be optimised. So it can change the sensitivity of the AM radio detection. My advice: make a crystal radio, or a simple superhet, test & try. Tune in e.g. to strong and weak radio stations and vary the bias, see/listen what happens. 9 sept. 2022.

[@graemezimmer604]

​@@radiofun232 But again you miss the point. Every diode has an optimum impedance, and if you set input and output impedances to suit, you will achieve optimum detection sensitivity. And that will be the best sensitivity that you will achieve with or without bias. But if you test a group of diodes in a fixed test set, then all you will find is which diode best suits that particular impedance. And you will miss the fact that some diodes will perform even better at their optimum impedance. Bottom line, this test can not show which diodes are the most sensitive, because it does not optimise the impedance for each diode. And BTW, I have build many, many radios and conducted many tests over 65 years.

[@radiofun232]

@@graemezimmer604 OK, good. But even with perfect impedances at the output and the input of the detection diode circuit there are situations where the signal that has to be detected is very weak. In such a case forward biasing can give better detection and thus reception results, by moving the operating point of the diode. That is my experience. Furthermore, it is a practical and handy way to get all kinds of AM detection diodes "working" in their optimum way. That is, of course, not your point, anyway.

[@valentinocolaon6060]

great video. thank you

[@sheep1ewe]

Thank You Ron, i really do appreciate this!

[@dmpi483]

What a diode collection. And I though I had a good stock.

[@glasslinger]

I have dozens more types which are all silicon, not suitable for crystal radios.

[@franzliszt3195]

I must say, this video is just plain fun!

[@silverXnoise]

Those 1N34s make great guitar fuzz/distortion pedals.

[@tucker934]

Pretty sure that’s do to one of it strange effects germanium diodes, and maybe a few others allow current to reverse in the opposite direction a bit. Please correct if I’m incorrect

[@markbarrett2225]

Thank you for your videos. I enjoy your channel a great deal.

[@andymouse]

Nice one Ron !....cheers.

[@johncasteel1780]

The '70s Yaesu FRG-7 uses 1N60s in its detectors.

[@MVVblog]

Very interesting!!!!!!

[@6F6G]

The forward voltage drop is proportional to the reverse breakdown voltage in Schottky diodes. A Schottky with a low voltage drop will have a relatively low breakdown voltage and high reverse current leakage like germanium diodes.

[@glasslinger]

Interesting observation. However, for crystal radio operation they still would need a small bias to eliminate the threshold.

[@6F6G]

@@glasslinger A schottky diode is a metal-semiconductor junction. Crystal detectors are primitive schottky diodes.

[@radio-pirol]

Thanks for making this video. I'm currently playing around with shortwave crystal radios and got stuck at using a east German WF OA645. Some people online recommend modern schottky diodes like HSMS-2860 or 5082-2835. I did not get hold of these for now but I may try.

[@glasslinger]

Note from the experiment that the schottky diodes work great if a small forward bias is used to bring the threshold down to zero. They will rectify up into the high megahertz, even the low UHF range!

[@radio-pirol]

@@glasslinger Just finished watching. I will try a little bias shortly! This might get my receiptions to a higher volume. I'm using 4kohms headphones as I don't want to spend like 60€ on a 50K:5K transformer or something similar just to get the load off that diode… that's not worth it…

[@SeanBZA]

@@radio-pirol If you are adding a bias battery anyway just add in a 1 transistor class A amplifier as well, to use the bias supply.

[@radio-pirol]

@@SeanBZA that would be too easy 😅 I want it to work without any active amplification.

[@seandoofer5720]

Interesting, the 1n34 has been a favorite for detecting for a long time, I'm surprised the schottkey works so well with a little bias. Have you tried this with a thermionic detector diode?

[@jozef1961ful]

Niedostępne są napisy. Zawsze były. Subtitles are not available. They have always been. Regards!

[@LheodaDjTechTv]

nice sharing idol

[@robnic52]

Great video thank you. I recently read a Jan 2007 article by Bob Culter describing a high sensitivity xtal set using a zero threshold double MOSFET chip ALD110900A for the detector. I built his circuit but could not get it working, I could not find the ferrite rod he specified. Tried a basic Chinese ferrite of similar size but possibly very different construction, also I am unsure if i soldered the litz wire properly. Further unsure whether the eBay litz was the RF enamelled strand type or the uninsulated type used in record player tonearms for flexibility (allowing free groove tracking). A bit clueless I'm afraid. :) Mr Coulter obviously had a working radio so a circuit to test this detector option might be interesting? Googly the article, very interesting stuff.

[@glasslinger]

The whole premise of a crystal radio is to not use power, such as a battery. Thus, using the FET (which requires bias voltage to operate as a detector) would not meet the requirements. Also note that a biased standard diode would not meet requirements either.

[@graemezimmer604]

@@glasslinger Ron, actually the Bob Culter design used a "zero Bias" FET, so no battery was needed. The problem with that article is it tries to use just a ferrite rod antenna, so it is wildly optimistic unless you live very close to a BC station. There has been a huge amount of work done since then (especially by the Chinese enthusiasts) and they get amazing results, but our tests on TRB showed that a FET detector (while excellent) is only about as sensitive as the best conventional diode detectors.

[@glasslinger]

@@graemezimmer604 Looking at the characteristic curves it looks like a small bias would be necessary to really get it working better than a point contact diode. It's really hard to beat a 1N34!

[@graemezimmer604]

@@glasslingerFWIW The original FET detector article was "Nostalgia for the Future" by Dave Cripe in 73 Amateur Radio Today, Dec 1995 . the other was the "Hi Sensitivity Crystal Set" by Bob Cutler in QST January 2007. If you go to youtube and search for "Testing the Double Tuned AM and FM Crystal Radio" you'll see an updated version using a 3Sk143 FET. Billy gets quite extraordinary results on both AM and FM in a portable set, but he is quite close to powerful transmitters.

[@donwright3427]

Have you ever smelt a burnt out selenium rectifier

[@andymouse]

You can't forget it if you have !

[@gordonwelcher9598]

It is worse than the byproduct of Taco Bell.

[@ingussilins6330]

:D :D yes..

[@tenlittleindians]

If these modern diodes were all we could find in the trenches today how could we figure out the forward bias needed without an oscilloscope? The needed voltage could be produced with homemade batteries of all kinds. Your 500 ohm variable resistor could be made up of individual scrap resistors to get the amount needed for a given diode. The 4.7 mf capacitor is a common enough value to be found in almost any scrap circuit board for salvage. Would you simply drop in a 500 ohm resistor and test it? What would you be listening for to determine if you needed more or less bias to tune such a circuit by ear?

[@migalito1955]

Thanks....

[@georgegonzalez2476]

Nice work. However your test circuit needs more capacitor to filter out the RF. The results with a bigger capacitor will be a bit more dramatic.

[@glasslinger]

The results in the voltage test were not as good. We were measuring the peak to peak output voltage. The capacitor reduced the peak voltage. The results in the headphone test had no discernible difference when a .002 uF was used as the filter. Surprised me a bit too!

[@georgegonzalez2476]

@@glasslinger Yes but pls keep in mind that the whole point of this circuit is to extract audio. If you measure the peak-to-peak signal, with a lot of RF feedthrough, you get a misleading reading-- you get the higher peak-to-peak voltage but only about HALF of that is really audio, the rest is the thick green fuzz of inaudible RF feedthrough. A capacitor will reduce the voltage on the scope, but you get a truer reading of the actual usable audio detected. Regards.

[@kensmith5694]

What diode is the schottky? I found that a 1N5711 works somewhere between my better 1N34 and my worse 1N34. I had to put a 2K across the crystal earphone to make it work.

[@joostderidder]

I wonder what a 1n4007 would do. I suppose that it is unusable also, just as the first SI-rectifier.

[@celsoneves2368]

Great!.

[@AureliusR]

Definitely would have been worth turning on averaging for the measurements! Also, I don't think I've ever seen someone use the resistance ranges for diode checks. At first blush, that seems to be like it wouldn't be a very good way to compare diodes. Why not use the diode check range on your meter? Also, during the 1MHz/10MHz checks around 25:30 in, I'm not seeing the waveforms change period when you switch. Were they still connected to the same circuit?

[@glasslinger]

It circuit was adjusted for most output and best view. Using a meter to test diodes is only for leakage and forward voltage. A curve tracer must be used for serious comparison.

[@infraceptor4749]

Among the most popular in this series were: 1N34A / 1N270 (germanium signal)_

[@TravisTellsTruths]

Wonderful cat

[@tenlittleindians]

32:25 minute mark begins the forward bias circuit and information on these diodes with bias added.

[@TheTjopp]

Would be interesting to see this kind of test with a copper oxide diode, and with an ideal (active) diode as comparison.

[@glasslinger]

A while back I tested a small meter rectifier copper oxide diode and found the frequency response decreased rapidly above the audio range. No good at all for RF. Not sure why, possibly the capacitance of the junction.

[@SeanBZA]

Should try to get some of the more exotic diodes, like a backwards diode, where forward voltage drop is actually higher than reverse breakdown. But those are hard to get now, such highly doped silicon and also so touchy in use.

[@glasslinger]

Motorola used to make these but they have been obsolete for over 20+ years now. Hen's teeth may be easier to find! :)

[@SeanBZA]

@@glasslinger Try Unijunction transistors, they also make a pretty heavily doped junction, though they are now impossible to get, only source being old russian parts on surplus, unless you find a stack in some dusty corner of a military warehouse surplused.

[@glasslinger]

@@SeanBZA I have some 2N491's. I'll see if they do anything.

[@ButtonsKing]

You should include a copper oxide element (Cuprox, Sirutor). It also was used in AM detectors, too.

[@glasslinger]

They were used back when the carrier frequencies were in the 20 KHZ TO 100 KHZ range. The efficiency drops drastically above about 30 KHZ and is for all practical purposes unusable at even broadcast band frequencies.

[@ButtonsKing]

@@glasslinger Thank you. My father told me that he had used it at about 700 KHz, but it was probably in the vicnity of a broadcast transmitter. I have never played with it.

[@fourfortyroadrunner6701]

All well and good HOW DO YOU BUY GENUINE 1N34 / 1N34A DIODES?????????????????

[@glasslinger]

It is definitely difficult with the scammers selling modern germanium diodes "as" 1n34 diodes. The modern diodes are not point contact. Keep an eye on ebay. Most are fakes but occasionally real 1n34 diodes show up. Look at the photo to make sure the package style is not modern!

[@ingussilins6330]

Next try direct current biased silicon diodes and LED's as a AM demodulator.

[@chriscimino7854]

Ecg126 base emitter junction works good. I am thinking of using a tuning coil to tune in and rectify with a 1n34a and filter a strong radio station to power the collector through the headphones and another variable tuning coil for the actual radio reception for weak or strong stations. I haven't tried it yet but its a theory I have on my mind or a solar panel but that's cheating hahahaha. Maybe just a long wire outside going to a rectifier will pick up any stray RF and rectify but I don't know

[@violinist3]

Have you tried biasing a silicon diode til it reaches conduction? I know purists would say that's not the classic way.

[@glasslinger]

Watch the video when the 1N914 is being tested with bias. Pay attention this time! :)

[@violinist3]

Yes..I put your video on pause for later viewing. I always wondered about the biasing approach.

[@graemezimmer604]

Yes. A Silicon diode has very high impedance and is pretty much unusable in a normal circuit. However with forward bias, the impedance can be brought down to usable values where it can work surprisingly well. Unfortunately (even with bias) a Silicon diode isn't as sensitive as a good germanium diode due to the different chemistry.

[@Super8Rescue]

I wish I understood all of this.

[@swirlingabyss]

The less voltage it takes to make it chooch, the better it will work for a crystal radio.

[@attilarivera]

👏👏👏👏👏👏👏 Thank u 💚💛💙🇧🇷🇧🇷🇧🇷

[@pjmelect]

I always use a 0A81 as a detector diode in crystal sets, I would have thought that you would have tested the range of the 0A series of diodes in your tests, or are the 0A series of diodes not commonly available in America as they are in the UK?

[@glasslinger]

They are available on UK ebay but the shipping is horrible. (way more than the items!) The original point contact diodes (1n34, 1n82, 1n60, etc.) have the same sensitivity as the point contact transistors. Note that none of the "modern" diodes are point contact any more. They are all diffused junction now, which have noticeably higher threshold voltages. This is why modern germanium diodes are not as sensitive as the originals with the same part number.

[@NiHaoMike64]

For the diodes with transparent casings, what happens to the sensitivity if they're illuminated with light? I'm thinking the photoelectric effect would help the electrons move.

[@SeanBZA]

Yes the light will provide bias, and with silicon you can use IR LED's to provide enough bias voltage to get them running. Then makes me think an optocoupler with plain transistor output will work as the right device, as you get a base emitter junction for the diode, and the right light source coupling.

[@franzliszt3195]

Saw in a 100 year old radio magazine that gelena detectors work better is bright sun light an in low light.

[@eDoc2020]

I'd like to make a crystal radio but with a spectrum analyzer I've found that the two strong stations where I live are only 30 kHz apart. I doubt I'd be able to get usable separation.

[@franzliszt3195]

You can make a notch filter to take those stations out.

[@diegogarciamedina6363]

Hi did you measure the voltage across crystal earphone with actual radio transmission? thanks

[@rustyrobinson8027]

Thanks

[@kenromaine2387]

Why was the cap added to the test PCB during the test but now shown on the test circuit drawing?

[@glasslinger]

Video editing mixup. Oh well. It should be there.

[@tenlittleindians]

I think people are forgetting the whole point was to be able to make a radio from commonly available parts. In todays world finding a small battery to provide the bias voltage would be the simplest component of the radio to source! Anyone that's ever seen a clock running off a potato for power could build the battery. Last time I checked, potatoes were easier to find than diodes that didn't need a little biasing. Can't wait for your potato powered diode radio plans! Even in the trenches of WW1 they had potatoes.

[@ikeabuchi1]

ck706 raytheon epoxy sealed verson direct competitor to the sylvania 1N34

[@glasslinger]

There are several other metal-semiconductor junction diodes as well, such as 1N60 and 1N82. All are not in production any more. All equivalents are semiconductor grown junction types now.

[@charl8764]

1n34a oa91 signal diodes

[@miroslawkaras7710]

You should not redeem the scaling so low, as the noise of the scope abscuring the signal shape.

[@glasslinger]

I wanted to see for sure what the low end sensitivity was.

[@kensmith5694]

@@glasslinger A useful way to cut down the noise is to add a very small capacitor on the load side. A lot of that noise is up in the 1MHz or 10MHz range.

[@paulziminskin2ghr282]

OK

[@wagnerfratelli3500]

hey friend enable the languages

[@qzorn4440]

and i thought i knew something about diodes... 🧐 extremly well done thank you. 🤩 it seems like the russians have a truth problem.. 🙄

[@IU3EVR]

Maybe you have a new oscilloscope .Lecroy I guess

[@stdorn]

Keysight

[@grampysmagic3327]

THANK YOU !!!!!!!!!!!! de AA9GG

[@phonotical]

For a 1N34A(g) I've found a D9D works