View Full Version : small signal schottky diodes
04-17-2004, 04:25 AM
I was just planning to make up a parts order for some VHF, UHF, and low microwave design projects.
One of the first devices I need to design is an RF detector probe.
In the past a diode like a 1N34A would be used for this type of application. This and most germanium devices are going the way of the do-do bird. I have run across the BAT46 schottky diode and wonder if this might be a modern new and improved device that could fill the void.
There is one trouble however. I can't find any data on response time or usefull frequency. I have found that a limiting factor may be the aprox. 10pf junction at 0 Volts.
How high a frequency can this 14 cent device be used for?
Are there better alternatives?
Here is the Vishay data sheet for the BAT 46 Schottky diode.
ST Microelectronic data sheet (Mouser web page)
The BAT46 is DO-35 150mA 100 Volt component.
It is regualrly marketed as a "better" alternative than a 1N34 germanium diode. #W8ZR uses them in his newest project (May 2004 QST) to protect the inputs for an intergated circuit by limiting peak-to-peak RF voltages.
Consider using the 1N914 or 1N916. #These are a small signal, high conductance fast switching Schottky device in a DO 35 package. #The 1N914 is 4 pf & 1N916 is 2 pf at zero volts. These are available for pennies each in surplus stores. #The BAT 46 has 6 to 10 pf of capacitance. #Here is an interesting article on extrapolating information from diode data sheets. #
04-17-2004, 06:35 PM
The 1N914 is a small signal silicon diode and is NOT a suitable replacement for the germanium types like the 1N34 series. The voltage drop across the diode is in the 0.6 to 0.7 volt range whereas the germanium types have a 0.2 to 0.3 volt drop.
When used as replacements in things like balanced modulators, mixers, detectors, the 1N914 does not usually function very well. Before I realized the difference I did try replacing the germanium diodes in the balanced mixers of Heath and Collins transmitters. Big mistake! The carrier supression was almost 20 dB less than with the germanium diodes!
The same thing is true if you try to make a crystal receiver out of a 1N914 instead of a germanium diode. Yes, you will be able to hear the pretty strong stations. However, the volume will be greatly reduced and the weaker stations will not be able to be heard. Same thing is true when used as a detector. The 1N914 requires a much stronger signal to be applied before it can function properly as a detector.
I have seen circuits for crystal radios that add a small amount of forward bias, just enough to bring a silicon diode to the brink of conduction, and then the RF will be properly rectified. Doing this with even the germanium diodes results in a more sensitive circuit!
However, the cost of the little germanium types is so low that it would not be economically feasable to do it with all the extra parts hung on the circuit in the commercial world. Of course, if you are just experimenting, have fun !
73 from Jim
For the sake of clarity, I am sure what STH meant was that in applications that were designed when germanium was the only low threshold voltage game in town, (before 1970), that they were used in mixer and detector circuits and as he said, they are still the best choice for crystal sets. #The signal levels of those old, mostly tube, receivers were designed to tickle germanium and will not easily overcome the higher voltage threshold that silicon (0.7v) or Schottky (0.3v) have. #Ion implantation of silicon has helped to bring this threshold voltage closer to germanium, but has not yet reached the 0.2 volts mark. #
Today virtually all mixer and detectors are Schottky devices. #The circuit designers compensate by driving the diode slightly harder compared to germanium, but just right for Schottkys. #The result is a much more reliable, less static sensitive, lower cost device to produce. #Hewlett Packard has been using matched sets of Schottkys in their spectrum analyzers, network analyzers, sweepers and scopes and other products for many years. #HP manufactured their own line of balanced mixers and detectors for low microwave applications using HP Schottkys like the 1N914. #There are a lot of different part numbers for Schottky diodes and many of them will work for you, don't be afraid to try some. #Use what is available and inexpensive.
Virtually all current ham radio equipment employ schottkys as mixers and detectors with a great deal of success. #Consider that the RF and IF amplifiers prior to the diodes, be it mixer or detector, #make up for any small diode junction voltage losses. #If you intend to measure all but very low signals, then Schottky will work fine. #Try it if it works in your Rf probe application, great!, if not you could always use the 1N34A. #Let us know.
04-18-2004, 12:53 AM
I was not questioning the use of the schottky diode in place of the 1N34 series, etc. Often they work even better than the original germanium types. There have been all sorts of articles, "tips", etc., about replacing the germanium diodes in Heath, Collins, etc., balanced modulators with schottky diodes. However, the 1N914 diode just isn't a good replacement in most situations for a germanium.
As for adding a bias to a crystal set, there are some "purists" who say that such a receiver is not a "true" crystal set since it requires a voltage other than that which is derrived from the r.f. signal to run it. Since my "main" crystal set is a Mengel Etherion that was manufactured on 12 October 1922 (I have the original box, original manual, and original packing slip!), I have enough problems finding a "sweet spot" with the cats whisker that I wouldn't worry much about putting a bias on the crystal!
There is a photo of this receiver on one of my websites (under the radio collection series of photographs)
04-18-2004, 01:25 AM
Thanks to all who have replied so far. I have been able, from the get go, to find the spec sheets for the BAT46 and have been researching other similar diodes.
So far I have found that the 1N5711, or the SD101 may work better than the 46 at low level. (up to 1 mA).
I am still looking for other similar devices.
About 8 years ago I designed and built an A/V signal and RF power control box for one of our TV news helicopters where I was taking signal from a directional coupler(-20dB) from a 2gHz 10 W RF line. I don't have the company reference books available anymore, but it seemed that the diode number I used was an HP-xxxx
and it worked fine.
04-18-2004, 02:00 PM
I have trimmed many a cat's whisker, and a few cats. I don't have a galena any more. Does anyone know the V-I characteristics?
BTW, we used to have a lady Dr. from Russia named Galena, said it was common in Russia.
Anyone pick up xtals from the railroad tracks? Cheap. http://www.qrz.com/iB_html/non-cgi/emoticons/smile.gif
I'm sorry, I Gave you a more generic response on inexpensive detectors which did not do justice to your level of technical expertise. #You had a newer call sign and I mistakenly thought it was a younger, less experienced person with limited resources. #It has been a while since I worked for Hewlett Packard, but as I recall there were several detector diodes that were useful well into the microwave region with a few performing at more than 10 Ghz. #I will list a few here, because I am not sure exactly what you had in mind for your RF probe design or the maximum frequency of interest. #
Microstrip line # #= 5082-2207 or 5082-2755
Surface mount # = #HSMS-2850
Beam lead type #= HSCH-5300 (1-26 Ghz)
Above in matched pairs is the HSCH-5500 #
These are all RF detector diodes. #The beam lead package may be used in stripline or microstripline applications or surface mount (if you clip the ribbon leads short), of course "short" is always required for microwave anyway. Hi Hi.
I have used the beam lead package detectors and PIN diodes in a number of my designs because surface mount packaging mounting is difficult for me on engineering breadboards. #Surface mount is more geared for a Panasonic Pick and Place robotics PC board assembly machine. #I have had some success using silver paste with surface mount devices and occasionally small gage silver solder and a fine tip low wattage soldering iron. #
Hp had some application notes on Schottky detector diodes in microwave applications printed in the late 1970's or early 80's. #One as I recall covered impedance matching them in circuits and used a spice model up to 26 Ghz. #I do not know if they are still in print. #Try your local HP sales office. #
You also might consider purchasing used or surplus RF test gear if you do not want to reinvent the wheel. #I hope this is is helpful, if I can be of more assistance, please let me know.
ALL Electronics is selling HP 5082-9496's -- matched quads of HP 5082-2835 Schottky diodes -- for about three dollars a quad. A GREAT deal. I bought some for an application at work and paid for them myself; that way I could keep a couple for upgrading mixers.
04-21-2004, 02:20 AM
Thanks K7FE and KA5S. Cort I am fairly certain that the diode I used before was the HP 5082. At least I recall a 5 and 8 in the part number.
As far as the newish ham call goes, I started studying for my ham ticket in 1964 and by 1967 I had a 6 meter beam on top of my folks house.I listed to 2 and 6 on a 1949 Dumont TV chassis (continous tuning from channel 1 to channel 13) Somehow I got distracted by "social life"
school, and working to make ends meet. By 1974 I had my first class phone and was working for a radio station before the ink was dry on the FCC paper.
It took a serious back problem for me to have the time to myself to buy the latest license manuals from the ARRL and bone up. Within a few months of my $56,000 operation I had my extra class ticket and an Icom 706mk2g, and was having a blast.
</span><table border="0" align="center" width="95%" cellpadding="3" cellspacing="1"><tr><td>Quote (kd5rpo @ April 20 2004,19:20)</td></tr><tr><td id="QUOTE">#Cort I am fairly certain that the diode I used before was the HP 5082. #At least I recall a 5 and 8 in #the part number. #
It took a serious back problem for me to have the time to myself to buy the latest license manuals from the ARRL and bone up. #Within a few months of my $56,000 operation I had my extra class ticket and an Icom 706mk2g, and was having a blast.[/QUOTE]<span id='postcolor'>
Yeah, 5082 was the part number prefix for their diodes. Now I am looking for 5082-0180's -- step recovery (or "snap") diodes.
Congratulations on getting back into ham radio - it's a shame it took a $56K operation to get you back!
Just to add another crystal diode you might try:
The 1N90 is recommended by some of the crystal
radio builders as being a good replacement for the
1N34 and is supposed to be more sensitive.
I have not tried this myself, just read about it.
</span><table border="0" align="center" width="95%" cellpadding="3" cellspacing="1"><tr><td>Quote (kd5rpo @ April 16 2004,21:25)</td></tr><tr><td id="QUOTE">I was just planning to make up a parts order for some VHF, UHF, and low microwave design projects. #
One of the first devices I need to design is an RF detector probe. #
In the past a diode like a 1N34A would be used for this type of application. #This and most germanium devices are going the way of the do-do bird. #I have run across the BAT46 schottky diode and wonder if this might be a modern new and improved device that could fill the void. #
There is one trouble however. #I can't find any data on response time or usefull frequency. #I have found that a limiting factor may be the aprox. 10pf junction at 0 Volts. #
How high a frequency can this 14 cent device be used for? #
Are there better alternatives? #
John, By now this has probably been answered.
Do a search for zero bias microwave diodes. There are very low capacitance zero bias detector diodes available. If you load them with a high resistance circuit and compensate them, they are reasonably linear down to a few hundredths of a volt.
W7EL wrote some articles about compensation systems for QRP power meters.
Cortland, What are you building that you need harmonics from a step recovery diode? #What is the desired output power, multiplication and freq?
K8ERV, Tom - #To answer your question about Galena, I did not put it on a curve traces to see the junction voltage drop, it is mostly lead or lead sulfide. #
For the ultimate in low voltage thresholds, try an inverted transistor shunt detector. #This could be done with a small signal microwave transistor so it would be useful at microwave frequencys. #Ground the collector of an NPN transistor, apply the signal to the emitter and supply about 50uA into the base thru a resistor. #The emitter-base is acting as a diode junction that is biased up to ZERO threshold. #Here is an article in EDN about it which appeared a few years ago.
Inverted transistor shunt detector (http://www.reed-electronics.com/ednmag/contents/images/60701di.pdf)