Keyboard Shortcuts
ctrl + shift + ? :
Show all keyboard shortcuts
ctrl + g :
Navigate to a group
ctrl + shift + f :
Find
ctrl + / :
Quick actions
esc to dismiss
Likes
Search
tunnel diodes retrace lines in curve tracer
|
Hi Miguel,
toggle quoted message
Show quoted text
This is precisely the curve a Tunnel Diodes creates on a curve tracer. From this image you can measure many of the important tunnel diode (TD) parameters. The curve tracer typically applies a rectified full-wave sine wave to the device under test. Assume we begin with the voltage at zero. As the voltage rises (we won't need more than a volt or two) the current in the TD starts to rise until it reaches a peak. This peak on the left of the curve tracer shows the maximum current of this part of the curve known as Ip. Ip is typically 1mA to 10mA and when you replace a TD you always want to match the new one to the Ip of the old one. As the curve reaches Ip drop down vertically to the horizontal axis and read the peak voltage. This parameter is Vp. This is typically 0.05V to 0.15V. Once the TD reaches Ip it can't go any further without switching in less than 1nSec to a new state. That happens so fast the curve tracer can't follow it and you see a horizontal line going to the right until it stops at a new point. The current hasn't changed but the voltage is much more now. It may be as much as 0.3V to 0.5V at this point. The current can now continue to increase and the voltage will increase with it but this part of the curve is not important. What happens next is the voltage from the curve tracer will start to drop. Remember we switched to the far right part of the curve and although the current is the same or more than the peak current, Ip, the voltage is more on this part of the curve. Now the voltage from the curve tracer will begin to drop. As it does the current can almost drop to zero and the voltage is dropping as well. Eventually we reach the next critical point where the voltage cannot drop any more without the tunnel diode jumping again. This point is called the Valley Voltage, or Vv. The current through the tunnel diode when it reaches Vv is called Iv. At this point the TD switches back to the left part of its curve in less than 1nSec. Once again this is too fast for the curve tracer to follow so as the beam moves as fast as it can you see a horizontal line going from the Valley voltage to the left where the tunnel diode is back on its original part of the curve. Now the process repeats. As the curve tracer voltage rises the current increases until it reaches Ip at which point the current stays the same but the voltage increases from Vp to a higher voltage. As the curve tracer voltage decreases the current drops and the voltage begins to drop until the voltage reaches Vv. At that point the TD jumps to the left part of the curve, the current stays the same at Iv but the voltage drops to almost 0V. The negative resistance can be calculated (approximately) if you draw a line between the two points on the curve tracer where the curve seems to disappear. Since your photo shows the vertical and horizontal sensitivity of the TD curve it is easy to calculate the approximate negative resistance of this TD. R = E / I = (0.12V - 0.45V) / (1mA - 0.2mA) = 0.33V / 0.8mA = -412ohms. Notice that is negative resistance. Negative resistance is an impossibility but it can exist within a very small region of the otherwise positive resistance of the complete tunnel diode curve. For more information Jack Rogers of Tek wrote a 13 page explanation of all of this with actual circuits used in Tek instruments showing how they perform. It is on the vintageTEK at Dennis Tillman W7pF -----Original Message-----
From: [email protected] [mailto:[email protected]] On Behalf Of Miguel Work Sent: Wednesday, February 17, 2021 12:28 PM To: [email protected] Subject: Re: [TekScopes] tunnel diodes retrace lines in curve tracer Thanks Jeff, I have found some information in internet, I was stubborn /g/TekScopes/album?id=260881 -- Dennis Tillman W7pF TekScopes Moderator |
|
You can actually trace the full TD curve, including the negative resistance region, with a special curve tracer or fixture. It's all about keeping the impedance levels very low. As I recall, there's a good example DIY one shown in the old GE Tunnel Diode manuals, from the 1960s. I even thought about making a TD fixture for the 576, but then I got back down to earth and considered why I would want to do that - other than just for the hell of it. I'd use a regular curve tracer really only to see if a TD is good, or to determine the peak and valley info on unidentifiable TDs, or sometimes to see if an unknown part is even a TD in the first place. All of these things can easily be determined without seeing any detail about the negative resistance region. The curve there is well known and predictable, so even if you don't see it, you can picture it. There is one area that could be an interesting use for a TD fixture, not just for them, but in designing low level negative resistance circuits, where you don't already know the expected shape of the curves.
Ed |
|
There is a straightforward modification to a 576 to get it to display the full tunnel diode curve. I did this a couple of years ago when I was measuring a bunch of tunnel diodes. The fixture gets a shunt resistance. The resistance of the shunt should be small enough to keep the parallel combination of the shunt and the negative resistance of the tunnel diode positive at all points in the curve. An opamp measures the voltage on the diode and applies a proportional offset current to the vertical amplifier inside the 576. The offset cancels the current in the shunt resistor so that the displayed current is just the diode's current. The modification was completely reversible. As I recall, the only reason the side cover had to be removed was to hook a clip lead (or maybe two leads) to a particular point in the vertical amplifier. There was no soldering involved. I removed the clip lead when the measurements were done. If I can find my notes from that, I will scan them and share.
|
|
OT: Fig. 1 of the Tektronix Jack Rogers doc....
Tunnel Diodes and the stock market ?? { apologies ... really } I was looking at Fig. 1 in the Jack Rogers doc that Dennis graciously posted, ... and was immediately struck, ...had just been doing an introductory "Stock Market 101 tutorial" for dummies, the other day, simply because I know nothing about the market. The first graph discussed in that SM 101 tutorial was called a "cup and handle" which sort-of looks like the TD Fig.1, in the Rogers Tek doc, - but the SM graph is reversed along the horizontal axis, compared to the TD graph. Which brings me to a question that must be obvious, and studied: -- Have TD's been modelled mathematically in efforts to better understand TD's non-linear behaviour and to produce better units ? Presumably there are (many ?) EE theses in this area. and -- Why have TD's become less of a produced and explored area of semiconductors ? Is there something that better replaces them now ? Rik |
|
My understanding is that by the 1970's or so, TD's were obsolete because most of the practical circuits that used them could be done cheaper with modern transistors and IC's.
TD's could do certain things with a bare minimum parts count, but that wasn't important enough, with the parts count (inside of IC's) exploding in other areas. Also note that many of the interesting apps for TD's involve inductors, but inductors are not cheap. I could be wrong... Pete |
|
Tunnel diodes excited a lot of interest for about a decade after their commercial introduction. In fact, TDs and the first IC from Fairchild (a four-transistor flip-flop) competed for attention at what is now the ISSCC conference, in 1961. The press largely ignored the IC. Articles from that time were obsessed with the tunnel diode instead. The future was going to be powered by TDs. Supercomputers, satellites, Dick Tracy wristwatch TVs...all were Coming Real Soon (tm), thanks to TDs. Then reality hit: Yes, they were much faster than contemporary transistors, but they were two-terminal devices. Making a chain of amplifiers was difficult because a change in load /here/, rippled all the way back to the input /there/. The same problem afflicted networks of logic gates. So scaling up to large systems seemed unlikely. Various desperate, complicated arrangements were devised in an effort to fix that and other practical problems, but the added complexity nullified the putative advantages of the TD. After a few turns of the Moore's law crank, it was clear that the IC was the future, and the TD gradually became the answer only to trivia questions.
toggle quoted message
Show quoted text
--Tom -- Prof. Thomas H. Lee Allen Ctr., Rm. 205 350 Jane Stanford Way Stanford University Stanford, CA 94305-4070 On 2/18/2021 13:54, saipan59 (Pete) wrote:
My understanding is that by the 1970's or so, TD's were obsolete because most of the practical circuits that used them could be done cheaper with modern transistors and IC's. |
|
(And yes, I know that we already had satellites by 1961. But not powered by TDs)
toggle quoted message
Show quoted text
-- Prof. Thomas H. Lee Allen Ctr., Rm. 205 350 Jane Stanford Way Stanford University Stanford, CA 94305-4070 On 2/18/2021 14:14, Tom Lee wrote:
Tunnel diodes excited a lot of interest for about a decade after their commercial introduction. In fact, TDs and the first IC from Fairchild (a four-transistor flip-flop) competed for attention at what is now the ISSCC conference, in 1961. The press largely ignored the IC. Articles from that time were obsessed with the tunnel diode instead. The future was going to be powered by TDs. Supercomputers, satellites, Dick Tracy wristwatch TVs...all were Coming Real Soon (tm), thanks to TDs. Then reality hit: Yes, they were much faster than contemporary transistors, but they were two-terminal devices. Making a chain of amplifiers was difficult because a change in load /here/, rippled all the way back to the input /there/. The same problem afflicted networks of logic gates. So scaling up to large systems seemed unlikely. Various desperate, complicated arrangements were devised in an effort to fix that and other practical problems, but the added complexity nullified the putative advantages of the TD. After a few turns of the Moore's law crank, it was clear that the IC was the future, and the TD gradually became the answer only to trivia questions. |
|
Tom Lee wrote on 2/18/2021 5:15 PM:
(And yes, I know that we already had satellites by 1961. But not powered by TDs)Hello,? Tom and the group-- I dimly recall a mention of 100-ampere tunnel diodes, likely lab curiosities that never saw production in any meaningful quantities. One possible application might have served as switches in a DC-to-DC convertor fed by power from low-voltage sources (thermopiles?). 73-- Brad? AA1IP |
|
Hmmm... Might there be a way to combine the TD and the transistor?
toggle quoted message
Show quoted text
Something compatible with the ginormous silicon CMOS infrastructure already in place? I no longer work for a semiconductor shop, so I don't have access to the brainpower there anymore, but people on this group must. Put your thinking caps on.... Jim Ford ------ Original Message ------
From: "Tom Lee" <tomlee@...> To: [email protected] Sent: 2/18/2021 2:14:43 PM Subject: Re: [TekScopes] tunnel diodes retrace lines in curve tracer Tunnel diodes excited a lot of interest for about a decade after their commercial introduction. In fact, TDs and the first IC from Fairchild (a four-transistor flip-flop) competed for attention at what is now the ISSCC conference, in 1961. The press largely ignored the IC. Articles from that time were obsessed with the tunnel diode instead. The future was going to be powered by TDs. Supercomputers, satellites, Dick Tracy wristwatch TVs...all were Coming Real Soon (tm), thanks to TDs. Then reality hit: Yes, they were much faster than contemporary transistors, but they were two-terminal devices. Making a chain of amplifiers was difficult because a change in load /here/, rippled all the way back to the input /there/. The same problem afflicted networks of logic gates. So scaling up to large systems seemed unlikely. Various desperate, complicated arrangements were devised in an effort to fix that and other practical problems, but the added complexity nullified the putative advantages of the TD. After a few turns of the Moore's law crank, it was clear that the IC was the future, and the TD gradually became the answer only to trivia questions. |
|
I spoke with EE's involved with tunnel diode in the early 60's.They told me of their struggles. TD's were going to be the great savior... except that is was hard to make a run of them that had consistent characteristics! The mix was variable...? Somewere used for low level microwave oscillators, I had a few boxes from NSA that used them as check sources.? Narrow band amplifiers were popular in the microwave world using circulators as signal separation devices, I have an Aertech 8.6-9.4 GHz Tunnel Diode Amplifier box. Even some high speed gates were made. Just couldnt make enough the same to do large volume production. The tunnel diode, or IIRC the Esaki diode, was largely a curiosity. Good I guess in some Tek circuits, but even they struggled, I think, to get consistency. I think I have some Tektronix "Diode Rise Time Fixtures" around that Tek made for testing and sorting diodes like this.
Jeff KruthIn a message dated 2/18/2021 4:55:02 PM Eastern Standard Time, saipan1959@... writes:?My understanding is that by the 1970's or so, TD's were obsolete because most of the practical circuits that used them could be done cheaper with modern transistors and IC's. TD's could do certain things with a bare minimum parts count, but that wasn't important enough, with the parts count (inside of IC's) exploding in other areas. Also note that many of the interesting apps for TD's involve inductors, but inductors are not cheap. I could be wrong... Pete |
|
Chuck Harris
Some of the paper abstracts I have seen when searching tunnel diodes
toggle quoted message
Show quoted text
seem to imply that they are being used in solar arrays for switching. -Chuck Harris Brad Thompson wrote: ... I dimly recall a mention of 100-ampere tunnel diodes, likely lab curiosities |
|
Hi Brad,
toggle quoted message
Show quoted text
RCA made them. They came in a 1" wide stud mounted case. The stud's threads were 3/8" - 24. The RCA 40079 had an Ip of 180 to 220 AMPS! The RCA 40070 was rated at 90 to 110 AMPS Ip. Dennis Tillman W7pF -----Original Message-----
From: [email protected] [mailto:[email protected]] On Behalf Of Brad Thompson Sent: Thursday, February 18, 2021 2:22 PM I dimly recall a mention of 100-ampere tunnel diodes, likely lab curiosities that never saw production in any meaningful quantities. One possible application might have served as switches in a DC-to-DC convertor fed by power from low-voltage sources (thermopiles?). 73-- Brad AA1IP -- Dennis Tillman W7pF TekScopes Moderator |
|
Hi Rik,
toggle quoted message
Show quoted text
Tunnel (Esaki) diodes rely on quantum tunneling. That is probably beyond the expertise of most electronics engineers to model in the way I think you mean. On the other hand it is not that hard to come up with a polynomial equation that models the shape of the I/V curve of the tunnel diode. I have done this myself. Excel can even automatically calculate a halfway decent polynomial from a spreadsheet you fill with about 20 or 30 I versus V points you plot off of a curve tracer. A really good curve fit requires about 12 or 13 polynomial terms. The answer to your 2nd question has a lot to do with biasing. With only two leads you are stuck with very simple biasing schemes and options. That severely limits what you can do with a Tunnel Diode and achieve CONSISTENT results. Without consistency it is hard to make reliable products. When they were first developed their simplicity and their speed were seen as big plusses. When it turned out that simplicity and consistency were not compatible that left their speed as their most interesting features. Gradually over time the semiconductor industry has learned how to make faster and faster semiconductors and that makes TDs less valuable as a circuit component. Dennis Tillman W7pF -----Original Message-----
From: [email protected] [mailto:[email protected]] On Behalf Of garp66 Sent: Thursday, February 18, 2021 10:59 AM To: [email protected] Subject: Re: [TekScopes] tunnel diodes retrace lines in curve tracer OT: Fig. 1 of the Tektronix Jack Rogers doc.... Tunnel Diodes and the stock market ?? { apologies ... really } I was looking at Fig. 1 in the Jack Rogers doc that Dennis graciously posted, .. and was immediately struck, ...had just been doing an introductory "Stock Market 101 tutorial" for dummies, the other day, simply because I know nothing about the market. The first graph discussed in that SM 101 tutorial was called a "cup and handle" which sort-of looks like the TD Fig.1, in the Rogers Tek doc, - but the SM graph is reversed along the horizontal axis, compared to the TD graph. Which brings me to a question that must be obvious, and studied: -- Have TD's been modelled mathematically in efforts to better understand TD's non-linear behaviour and to produce better units ? Presumably there are (many ?) EE theses in this area. and -- Why have TD's become less of a produced and explored area of semiconductors ? Is there something that better replaces them now ? Rik -- Dennis Tillman W7pF TekScopes Moderator |
|
Dennis Tillman W7pF wrote on 2/18/2021 7:14 PM:
Hi Rik,<snip> Hello-- Does anyone recall "Laugh In" (a satirical TV series of the 1960s which featured certain repetitive routines-- one called "sock it to me" in particular). I give you the following: "It may be a tunnel diode to you, but it's an Esaki to me...Esaki to me...." <grin> 73-- Brad? AA1IP |
|
No, not grin, Brad, groan!
toggle quoted message
Show quoted text
I never did understand the "sock it to me" chorus in Aretha Franklin's "RESPECT". I guess I'm too young (born 1965) to get it.... Jim Ford ------ Original Message ------
From: "Brad Thompson" <brad.thompsonaa1ip@...> To: "[email protected]" <[email protected]> Sent: 2/18/2021 8:10:22 PM Subject: Re: [TekScopes] tunnel diodes retrace lines in curve tracer Dennis Tillman W7pF wrote on 2/18/2021 7:14 PM:Hi Rik,<snip> |
|
Brad Thompson wrote:
Dennis Tillman W7pF wrote on 2/18/2021 7:14 PM:Hi Rik,<snip> It was being shown on Decades TV, not long ago. It might still be. Two episodes, M-F from 7-9PM, EST. |
saipan59 (Pete)
Michael A. Terrell
to navigate to use esc to dismiss