what's the market for newly produced tunnel diodes?


 

Hi all,

there's a few companies producing clones of modern chips for music
synthesis, including CoolAudio (aka Uli Behringer) in China, ALFA RPAR
in Latvia, and SSI. The first two have their own foundries and SSI
supposedly uses Diodes Corp. Would approaching these people to get a
run of tunnel diodes done be a good idea? ALFA specifically seem to be
a smaller company running a tighter ship, but Uli is into a lot of
vintage electronics so he might be interested as well. Obviously for
every company there's some sort of price cutoff point.

Which means one has to ask, how many tunnel diodes could the world use
within the next 10 years? What do you all think?

Cheers


Stephen
 

I think that would be a very good idea.
I read often that many a unit gets trashed or forgotten because a tunnel diode failure. That’s sad IMHO.


 

In a similar vein, I'm wondering if it is possible to replace the tunnel diodes in a 465/475/475A with something other than a tunnel diode. I thought that there was a recent thread that mentioned replacing tunnel diodes in another instrument (a sampling plug-in?) with Schottky diodes, which got me thinking about this.

-- Jeff Dutky


Bill E
 

My opinion, close to zero for the traditional tunnel diode. While they were useful in their time, they just can't do what modern equivalents can do. Not capable of producing much power, and even for tunneling behavior, there is newer tech that is far better, e.g. RTDs (resonant tunneling diodes).
But, while I'm an EE, haven't had any relevant experience with TDs for, hmm, a long time. Which might in itself validate my comment. ::)


Chuck Harris <cfharris@...>
 

Conservatively, I would estimate 100.

Look at how many of "us" exist, than notice
that we don't all have things that need their
TD's replaced.

Commercially, the requirement would be essentially
zero. They don't exist in TV's radios, or video
games. They really never took off in that sphere.

The Russians were big in TD's, because they were useful,
and something they could do with their foundries in
the 1960's. TD's didn't demand the cleanliness, or
purity of materials that high speed silicon devices
do.

If you want a foundry to start making TD's again, you
are going to have to create some game altering need
for them to exist... in quantity.

Many universities have mini foundries that they use
to teach practical semiconductor manufacturing techniques.

One such application, I think, is power switching in
solar arrays... or so I am led to believe from the papers
on high power TD's coming from China.

I know that my local university was letting grad students
make LED's, and other crude diodes for educational purposes.

To make TD's, you don't need much more than a chemical hood,
some really nasty chemicals, GE wafers, a way to cut them,
a vacuum deposition chamber... AKA: sputterer, and some stereo
microscopes...

They could have also made TD's.

-Chuck Harris


On Fri, 2 Apr 2021 18:08:35 +0200 "cheater cheater"
<cheater00social@gmail.com> wrote:
Hi all,

there's a few companies producing clones of modern chips for music
synthesis, including CoolAudio (aka Uli Behringer) in China, ALFA RPAR
in Latvia, and SSI. The first two have their own foundries and SSI
supposedly uses Diodes Corp. Would approaching these people to get a
run of tunnel diodes done be a good idea? ALFA specifically seem to be
a smaller company running a tighter ship, but Uli is into a lot of
vintage electronics so he might be interested as well. Obviously for
every company there's some sort of price cutoff point.

Which means one has to ask, how many tunnel diodes could the world use
within the next 10 years? What do you all think?

Cheers


Bill E
 

And, of course, if you want them to fix old stuff (which we all do), realize that's really a non-market. Setting up fab for a few hundred units that only crazy people like us would buy isn't economically viable.


Bill E
 

Chuck, TD for power switching in solar arrays? Why? Sorry, didn't bother to track down the links, but what's the scenario that isn't handled by modern stuff? (Just to be clear, not a bash, just wondering what China thinks).


Stephen
 

I understand and agree with all your valid points.
But since you seem to say that there are far better modern alternatives, why can’t they be used to repair let’s say, a Type 284 for example?
PS: I’m not an EE, and know nothing about this. I’m just wondering why people are looking for TD’s at ridiculous prices, if there are better modern alternatives... 🤷‍♂️


Eric
 

The issue the D180 in the 284 is its speed. It only had 1.5pf of capacitance to get he 70ps edge speed so in terms of TD it is on the exotic side and the hardest to make. Last time I looked there was a company you could buy from but they were a special run and $280 each diode. However you could buy them in lots of 1 so that is a + I guess

Eric

-----Original Message-----
From: TekScopes@groups.io <TekScopes@groups.io> On Behalf Of Stephen
Sent: Friday, April 2, 2021 12:46 PM
To: TekScopes@groups.io
Subject: Re: [TekScopes] what's the market for newly produced tunnel diodes?

I understand and agree with all your valid points.
But since you seem to say that there are far better modern alternatives, why can’t they be used to repair let’s say, a Type 284 for example?
PS: I’m not an EE, and know nothing about this. I’m just wondering why people are looking for TD’s at ridiculous prices, if there are better modern alternatives... 🤷‍♂️


Bill E
 

That's the key. Circuits were designed in the day for the characteristics of the components available. Change parameters without understanding the rest of the circuit, fail.


Tom Lee
 

Classic tunnel diodes can't be replaced by Schottky didoes. The latter lack the negative resistance region that the former have.

There are so-called "backward diodes", though, which are tunnel diodes that have been doped just heavily enough to eliminate the forward-current peak. That gives you, essentially, a Zener diode with a reverse breakdown voltage that is smaller than the forward drop. Exchanging anode and cathode then produces a diode with a lower "forward" drop than a conventional diode made of the same base semiconductor. That's the only type of tunnel diode for which a Schottky might be suitable. Schottkys will not sub for the classic tunnel diodes whose negative resistance is used for realizing fast, low-voltage flip-flops of trigger circuits.

--
Prof. Thomas H. Lee
Allen Ctr., Rm. 205
350 Jane Stanford Way
Stanford University
Stanford, CA 94305-4070
http://www-smirc.stanford.edu

On 4/2/2021 09:16, Jeff Dutky wrote:
In a similar vein, I'm wondering if it is possible to replace the tunnel diodes in a 465/475/475A with something other than a tunnel diode. I thought that there was a recent thread that mentioned replacing tunnel diodes in another instrument (a sampling plug-in?) with Schottky diodes, which got me thinking about this.

-- Jeff Dutky




Bill E
 

Yes, the negative resistance is what made TDs useful, nothing else, the TD reverse breakdown wasn't anything that made them commercially viable.


Tom Lee
 

Backward diodes found some use in rf detectors and energy-harvesting circuits, where a low "forward" drop seems to be attractive (it's not quite the advantage that is assumed, for quite fundamental physics reasons). They've never completely gone away for that reason.

I agree with the prevailing sentiment here that restarting (or asking) a fab to make tunnel diodes makes no business sense. A good fraction of the global market for TDs seems to be well served by the Cold War leftovers available on ebay.

Tom

--
Prof. Thomas H. Lee
Allen Ctr., Rm. 205
350 Jane Stanford Way
Stanford University
Stanford, CA 94305-4070
http://www-smirc.stanford.edu

On 4/2/2021 10:23, Bill E wrote:
Yes, the negative resistance is what made TDs useful, nothing else, the TD reverse breakdown wasn't anything that made them commercially viable.




Keith Erickson
 

I have 12 NIB 1N3713 GE TD in original packaging.
Email for photos.
Keith Erickson
Wayzata, MN
Kognw at gte.net

On Apr 2, 2021, at 12:36 PM, Tom Lee <tomlee@ee.stanford.edu> wrote:

Backward diodes found some use in rf detectors and energy-harvesting circuits, where a low "forward" drop seems to be attractive (it's not quite the advantage that is assumed, for quite fundamental physics reasons). They've never completely gone away for that reason.

I agree with the prevailing sentiment here that restarting (or asking) a fab to make tunnel diodes makes no business sense. A good fraction of the global market for TDs seems to be well served by the Cold War leftovers available on ebay.

Tom

--
Prof. Thomas H. Lee
Allen Ctr., Rm. 205
350 Jane Stanford Way
Stanford University
Stanford, CA 94305-4070
http://www-smirc.stanford.edu

On 4/2/2021 10:23, Bill E wrote:
Yes, the negative resistance is what made TDs useful, nothing else, the TD reverse breakdown wasn't anything that made them commercially viable.








Chuck Harris <cfharris@...>
 

Only takes a minute to look things up these days...

One paper from '77 indicates their use as a thin film
solar energy conversion array. They are claiming that
an array using Al-SiO2-Si structure, is actually a tunnel
diode. They call it a tunnel MIS solar cell.

In another paper from '79, they describe an AlGaAs tunnel
diode with a band gap of 1.6eV that can withstand very high
solar exposure without degradation.

Another, from 2013, about a solar array "rectenna", that uses
a host of microscopic tunnel diodes in the conversion of solar
energy into power by rectifying the E-M wave of IR light
directly into DC.

So, tunnel diodes are seeing some new applications... Doesn't
seem to help us, though...

-Chuck Harris


On Fri, 02 Apr 2021 09:43:51 -0700 "Bill E" <solartron@quackers.net>
wrote:
Chuck, TD for power switching in solar arrays? Why? Sorry, didn't
bother to track down the links, but what's the scenario that isn't
handled by modern stuff? (Just to be clear, not a bash, just
wondering what China thinks).





 

OK so here's an application that I think would be interesting, and
could probably use tunnel diodes similar to what's being used in tek
scopes: music circuits. Stomp boxes, filters, distortion, EQ,
oscillators, etc. Tunnel diodes are most likely going to sound very
different. There are few items out there with actual negative
resistance characteristics and this could easily be something that
would find a large market. If you want to reissue old electronics, I'm
fairly certain that would be useful. I think a diode ladder filter
could be very interesting. Does anyone with (any) tunnel diodes want
to set up a simple circuit like this to see if it'll work out? If
there's an interesting sound to this, people will jump on it and the
market will suddenly be much larger. Note that the schematic has
transistors configured as diodes.

http://www.florian-anwander.de/roland_filters/TB303.jpg

Are there any other resistant filters (low pass, band pass, high pass,
band reject) with voltage controlled cutoff or resistance that
uniquely use negative resistance?

Cheers,
Damian

On Fri, Apr 2, 2021 at 9:01 PM Chuck Harris <cfharris@erols.com> wrote:

Only takes a minute to look things up these days...

One paper from '77 indicates their use as a thin film
solar energy conversion array. They are claiming that
an array using Al-SiO2-Si structure, is actually a tunnel
diode. They call it a tunnel MIS solar cell.

In another paper from '79, they describe an AlGaAs tunnel
diode with a band gap of 1.6eV that can withstand very high
solar exposure without degradation.

Another, from 2013, about a solar array "rectenna", that uses
a host of microscopic tunnel diodes in the conversion of solar
energy into power by rectifying the E-M wave of IR light
directly into DC.

So, tunnel diodes are seeing some new applications... Doesn't
seem to help us, though...

-Chuck Harris


On Fri, 02 Apr 2021 09:43:51 -0700 "Bill E" <solartron@quackers.net>
wrote:
Chuck, TD for power switching in solar arrays? Why? Sorry, didn't
bother to track down the links, but what's the scenario that isn't
handled by modern stuff? (Just to be clear, not a bash, just
wondering what China thinks).









 

Tom Lee wrote:

Schottkys will not sub for the classic tunnel diodes whose negative resistance is
used for realizing fast, low-voltage flip-flops of trigger circuits.
I have to admit that I have completely failed to understand the operation of the trigger circuits in either the 475/A or my newer, 2200-series scopes (which do not appear to use tunnel diodes) on the first few read-throughs. I have gone back to the Theory of Operation section of my 2236 service manual, and now it seems like it makes more sense to me.

First, I might have gotten Schmitt triggers and Schottky diodes confused. The 2236 manual discusses a Schmitt trigger as part of the trigger circuit.

Second, I notice that the trigger circuit on the 2236 contains a pair of D FFs (not even anything exotic, just a 74LS74A). I'm not convinced that the 74LS74A can be the part that is handling the trigger off of 100 MHz signals, as it's rated maximum clock frequency is only 25 MHz (according to the TI data sheet).

That leaves the Schmitt trigger (and associated circuitry). Reading further I see that some of the ICs related to the Schmitt trigger circuit are ECL devices (U480 - ECL NOR, U502 - ECL monostable multivib., U506 - ECL MS FF, and U532 - ECL NOR), so they may be able to operate at 100 MHz. I've been looking into ECL logic for a couple of months as 1) it looks like ECL devices are needed in order to do GHz frequency counting, and 2) I considered ECL to be an even higher form of magic than regular transistor circuits, and I was tired of being an ignorant savage.

Now I will need to read through the 475 service manual more carefully and see how the trigger circuits compare to those in the 2236.

-- Jeff Dutky


 

The nice thing about an arrangement where the same parts are used for
audio and for testing equipment is that when selecting parts for
scopes you'll be looking for something different than the people using
them for audio, so there will be a nice selection for either of those
uses.

Another question. You can configure a FET as a diode. But is it
possible to have tunnel FETs? If we had those, then there might be way
more uses even. For example, microphones often need a first amplifier
that has very, very low capacity. That is why they often use tubes. It
improves high frequency response as well as some other parameters. A
very low capacity FET would be desirable in such a scenario. This
could easily raise the market as well. If we could use such a fet as a
diode then that would be great.

On Fri, Apr 2, 2021 at 10:19 PM cheater cheater
<cheater00social@gmail.com> wrote:

OK so here's an application that I think would be interesting, and
could probably use tunnel diodes similar to what's being used in tek
scopes: music circuits. Stomp boxes, filters, distortion, EQ,
oscillators, etc. Tunnel diodes are most likely going to sound very
different. There are few items out there with actual negative
resistance characteristics and this could easily be something that
would find a large market. If you want to reissue old electronics, I'm
fairly certain that would be useful. I think a diode ladder filter
could be very interesting. Does anyone with (any) tunnel diodes want
to set up a simple circuit like this to see if it'll work out? If
there's an interesting sound to this, people will jump on it and the
market will suddenly be much larger. Note that the schematic has
transistors configured as diodes.

http://www.florian-anwander.de/roland_filters/TB303.jpg

Are there any other resistant filters (low pass, band pass, high pass,
band reject) with voltage controlled cutoff or resistance that
uniquely use negative resistance?

Cheers,
Damian

On Fri, Apr 2, 2021 at 9:01 PM Chuck Harris <cfharris@erols.com> wrote:

Only takes a minute to look things up these days...

One paper from '77 indicates their use as a thin film
solar energy conversion array. They are claiming that
an array using Al-SiO2-Si structure, is actually a tunnel
diode. They call it a tunnel MIS solar cell.

In another paper from '79, they describe an AlGaAs tunnel
diode with a band gap of 1.6eV that can withstand very high
solar exposure without degradation.

Another, from 2013, about a solar array "rectenna", that uses
a host of microscopic tunnel diodes in the conversion of solar
energy into power by rectifying the E-M wave of IR light
directly into DC.

So, tunnel diodes are seeing some new applications... Doesn't
seem to help us, though...

-Chuck Harris


On Fri, 02 Apr 2021 09:43:51 -0700 "Bill E" <solartron@quackers.net>
wrote:
Chuck, TD for power switching in solar arrays? Why? Sorry, didn't
bother to track down the links, but what's the scenario that isn't
handled by modern stuff? (Just to be clear, not a bash, just
wondering what China thinks).









Tom Lee
 

ECL is actually analog masquerading as digital, so they're actually easier to understand than, say, a 7400 NAND. If a bipolar transistor is ever driven into saturation, it takes a long time to recover, so ECL simply forbids saturation. The tradeoff for the increased speed is high static power consumption.

Tek's pulse generators regularly used ECL gates as analog elements (e.g., NOR gates as high-frequency oscillators). Old ECL parts make for nice and fast analog circuits. Line receivers in particular are terrific amplifiers (and damn fast Schmitts), as they are little more than a diff pair with a follower. I use a fair amount of ECL, but almost never for logic.

Tom

--
Prof. Thomas H. Lee
Allen Ctr., Rm. 205
350 Jane Stanford Way
Stanford University
Stanford, CA 94305-4070
http://www-smirc.stanford.edu

On 4/2/2021 14:09, Jeff Dutky wrote:
Tom Lee wrote:
Schottkys will not sub for the classic tunnel diodes whose negative resistance is
used for realizing fast, low-voltage flip-flops of trigger circuits.
I have to admit that I have completely failed to understand the operation of the trigger circuits in either the 475/A or my newer, 2200-series scopes (which do not appear to use tunnel diodes) on the first few read-throughs. I have gone back to the Theory of Operation section of my 2236 service manual, and now it seems like it makes more sense to me.

First, I might have gotten Schmitt triggers and Schottky diodes confused. The 2236 manual discusses a Schmitt trigger as part of the trigger circuit.

Second, I notice that the trigger circuit on the 2236 contains a pair of D FFs (not even anything exotic, just a 74LS74A). I'm not convinced that the 74LS74A can be the part that is handling the trigger off of 100 MHz signals, as it's rated maximum clock frequency is only 25 MHz (according to the TI data sheet).

That leaves the Schmitt trigger (and associated circuitry). Reading further I see that some of the ICs related to the Schmitt trigger circuit are ECL devices (U480 - ECL NOR, U502 - ECL monostable multivib., U506 - ECL MS FF, and U532 - ECL NOR), so they may be able to operate at 100 MHz. I've been looking into ECL logic for a couple of months as 1) it looks like ECL devices are needed in order to do GHz frequency counting, and 2) I considered ECL to be an even higher form of magic than regular transistor circuits, and I was tired of being an ignorant savage.

Now I will need to read through the 475 service manual more carefully and see how the trigger circuits compare to those in the 2236.

-- Jeff Dutky




 

Tom,

Isn't everything just analog masquerading as digital?

(or maybe the other way around? Avogadro's number is as close to infinite as makes no odds for most folks)

-- Jeff Dutky