2-terminal negative resistance circuits

On Thursday, July 4, 2019 at 5:07:55 PM UTC-4, John Larkin wrote:
On 4 Jul 2019 11:23:58 -0700, Winfield Hill
wrote:

Bill Sloman wrote...

[ snip ]

Bill, you're getting awfully snarky in your old age.

He should design something. He might feel better.

Did he ever finish that world shaking low distortion oscillator he started about a decade ago?
 
On Thursday, July 4, 2019 at 4:16:39 PM UTC-7, John Larkin wrote:
ood for testing crystals.
How would you make a 2-terminal negative resistor with an LM13700?

Positive input to terminal 1, negative input to terminal 2, and the
output is a current source proportional to the voltage difference; wire
that to terminal 1. Either midpoint/ground terminal 2, or (if you want to get
fancy and have more compliance range) use the second section
amplifier with inverted input polarities and wire its output to terminal 2.

inputs won't like +/-9V range, but maybe you can power it with a couple of lithium
cells and get by. One or two resistors, to bias the amplifier(s) is
the rest of the part count. One chip, two resistors, two batteries.
An Altoids box and some black paint and voila!
 
On Friday, July 5, 2019 at 3:09:24 AM UTC+2, amdx wrote:
On 7/4/2019 5:12 PM, Bill Sloman wrote:
On Thursday, July 4, 2019 at 11:07:55 PM UTC+2, John Larkin wrote:
On 4 Jul 2019 11:23:58 -0700, Winfield Hill <winfieldhill@yahoo.com
wrote:

Bill Sloman wrote...

[ snip ]

Bill, you're getting awfully snarky in your old age.

He should design something. He might feel better.

I feel fine. I've even designed some stuff recently - though the NSW IEEE July Newsletter is not the kind of electronic design that John Larkin would like us to think that he is thinking of.

Coming home from an excellent dinner in a three star restaurant is an even better way of getting to feel fine, though it doesn't seem to make me any more tolerant of half-baked opinions.


Maybe you did, but I see by your responses to John's posts, you are
very jealous of what he does.

I asdmire his capacity to get paid for it, but it is bit irritating to listen to him congratulate himself for "innovative" stuff that we were doing back in the late 1980's.

He's got faster components to play with, so in one sense he's doing it better, but we worked out a few tricks that he doesn't seem to have taken on board.

He has got himself a fairly sweet gig - for him. It's not one that would suit me, so I'm probably not actually jealous.

--
Bill Sloman, Sydney
 
On Friday, July 5, 2019 at 4:25:31 AM UTC+2, Michael Terrell wrote:
On Thursday, July 4, 2019 at 5:07:55 PM UTC-4, John Larkin wrote:
On 4 Jul 2019 11:23:58 -0700, Winfield Hill
wrote:

Bill Sloman wrote...

[ snip ]

Bill, you're getting awfully snarky in your old age.

He should design something. He might feel better.

Did he ever finish that world shaking low distortion oscillator he started about a decade ago?

If the low distortion oscillator had been world-shakng, it might have shaken loose a few potential customers.

The schematics are complete enough, and I started putting them into KiCad a few years ago, with the idea of turning them into a printed circuit layout which I could get made and stuffed, and test, but in the total absence of any potential customers I've never got on with it.

--
Bill Sloman, Sydney
 
On Friday, July 5, 2019 at 1:03:50 AM UTC+2, k...@notreal.com wrote:
On Thu, 04 Jul 2019 14:07:44 -0700, John Larkin
jjlarkin@highlandtechnology.com> wrote:

On 4 Jul 2019 11:23:58 -0700, Winfield Hill <winfieldhill@yahoo.com
wrote:

Bill Sloman wrote...

[ snip ]

Bill, you're getting awfully snarky in your old age.

He should design something. He might feel better.

Not possible. Either sentence.

Krw's grasp of possibilities is as feeble as the rest of his grasp of reality.

--
Bill Sloman, Sydney
 
On 04/07/2019 11:29 am, Robert Baer wrote:
  2-terminal negative resistance circuits anyone?

  Most especially "Z" types.

  Thanks.

Please what is meant by "Z type" ?

piglet
 
On 2019-07-05 09:09, piglet wrote:
On 04/07/2019 10:43 pm, Dave Platt wrote:

Lambda diode (e.g. two JFETs).

Or indeed one fet and one bjt.

piglet

Like https://w3jdr.wordpress.com/lambda-diode-experiments/

Nice circuit! Why isn't that in the AoE?
(at least not in the index)

Arie
 
Arie de Muynck wrote...
On 2019-07-05 09:09, piglet wrote:
On 04/07/2019 10:43 pm, Dave Platt wrote:

Lambda diode (e.g. two JFETs).

Or indeed one fet and one bjt.

Like https://w3jdr.wordpress.com/lambda-diode-experiments/

Nice circuit! Why isn't that in the AoE?
(at least not in the index)

Nice idea, Arie. What can you do with it?

We like depletion devices, whether JFETs or MOSFETs.
But the -Vgs curve isn't specified or very dependable.


--
Thanks,
- Win
 
On a sunny day (5 Jul 2019 02:30:02 -0700) it happened Winfield Hill
<winfieldhill@yahoo.com> wrote in <qfn5aq0g8p@drn.newsguy.com>:

Arie de Muynck wrote...

On 2019-07-05 09:09, piglet wrote:
On 04/07/2019 10:43 pm, Dave Platt wrote:

Lambda diode (e.g. two JFETs).

Or indeed one fet and one bjt.

Like https://w3jdr.wordpress.com/lambda-diode-experiments/

Nice circuit! Why isn't that in the AoE?
(at least not in the index)

Nice idea, Arie. What can you do with it?

We like depletion devices, whether JFETs or MOSFETs.
But the -Vgs curve isn't specified or very dependable.


One trannny:
0 V +
|
___ |________
| | |
R1 | |
| |/ c |
|---- -| |
| |\|e R3
--- | |
\ / > |----
--- LED R2
| |
--------
|
0 V -

Z1 1.5V LED
R1 10k
R2 1k
R3 10k



At low voltages circuit conducts a lot via transistor and R2,
tranny basically shorts R3.

When V goes high at some point (about 10 V) emitter is lifted bove VZ1 - Vbe,
Ic drops to zero, only current left via R1 and LED plus current through R2 + R3.
 
On 2019-07-05 12:06, Jan Panteltje wrote:
On a sunny day (5 Jul 2019 02:30:02 -0700) it happened Winfield Hill
winfieldhill@yahoo.com> wrote in <qfn5aq0g8p@drn.newsguy.com>:

Arie de Muynck wrote...

On 2019-07-05 09:09, piglet wrote:
On 04/07/2019 10:43 pm, Dave Platt wrote:

Lambda diode (e.g. two JFETs).

Or indeed one fet and one bjt.

Like https://w3jdr.wordpress.com/lambda-diode-experiments/

Nice circuit! Why isn't that in the AoE?
(at least not in the index)

Nice idea, Arie. What can you do with it?

We like depletion devices, whether JFETs or MOSFETs.
But the -Vgs curve isn't specified or very dependable.


One trannny:
0 V +
|
___ |________
| | |
R1 | |
| |/ c |
|---- -| |
| |\|e R3
--- | |
\ / > |----
--- LED R2
| |
--------
|
0 V -

Z1 1.5V LED
R1 10k
R2 1k
R3 10k



At low voltages circuit conducts a lot via transistor and R2,
tranny basically shorts R3.

When V goes high at some point (about 10 V) emitter is lifted bove VZ1 - Vbe,
Ic drops to zero, only current left via R1 and LED plus current through R2 + R3.

Very nice idea too. Reminds me of the SOA foldback in PSU and audio
output amplifier circuits. Problem for me would be the high remaining
current through R3, there is no current gain in that path like with the
BJE + JFET circuit. I need a high (inverse) current ratio between lo and
high voltage load.

BTW - running out of pencils?

Arie
 
On Friday, 5 July 2019 00:16:39 UTC+1, John Larkin wrote:
On Thu, 4 Jul 2019 15:51:29 -0700 (PDT), whit3rd <whit3rd@gmail.com
wrote:
On Thursday, July 4, 2019 at 7:35:04 AM UTC-7, John Larkin wrote:

As a college ee project, I built a 2-terminal negative resistance box,
with an opamp and a couple of 9 volt batteries. If you plug in a
negative value for R into the common RC and RLC circuits, the math
still works and the waveforms are radical.

Parallel such a box with a variable resistor, and make a
quartz crystal tester (which starts from white noise when the 'gain' reaches
Barkhausen's criterion).

I'd be thinking of using a LM13700 instead of op amp, though. Transconductance
is a the property you want, it just takes a bias resistor or two.

I used a slow opamp, 709 or maybe 741 by then, so it was only good to
100 KHz or so. Not so good for testing crystals.

Could do watch crystals. I also have a glass valve type encased crystal that's 5 or 6 kHz.


NT
 
On 2019-07-05 11:30, Winfield Hill wrote:
Arie de Muynck wrote...

On 2019-07-05 09:09, piglet wrote:
On 04/07/2019 10:43 pm, Dave Platt wrote:

Lambda diode (e.g. two JFETs).

Or indeed one fet and one bjt.

Like https://w3jdr.wordpress.com/lambda-diode-experiments/

Nice circuit! Why isn't that in the AoE?
(at least not in the index)

Nice idea, Arie. What can you do with it?

We like depletion devices, whether JFETs or MOSFETs.
But the -Vgs curve isn't specified or very dependable.

Dynamic load for power supplies. I had a problem with leaking schottky's
in a dual input PSU (PoE and 24Vac input circuits). When the sun burned
on the equipment (80C inside), the leakage from the isolation diode to
the PoE circuits caused lockup, PoE would not start in some
circumstances. Solved it with a more complex circuit, this negative
resistor would have worked fine. The Vgs spread would have been
tolerable but maybe the JFET + BJE circuit would have been more
reproducable. Time for some LTspicing...

Arie
 
On 4/7/19 12:29, Robert Baer wrote:
  2-terminal negative resistance circuits anyone?

  Most especially "Z" types.

  Thanks.

Cross-coupled transistor pair.

Pere
 
On Friday, 5 July 2019 10:30:23 UTC+1, Winfield Hill wrote:
Arie de Muynck wrote...
On 2019-07-05 09:09, piglet wrote:
On 04/07/2019 10:43 pm, Dave Platt wrote:

Lambda diode (e.g. two JFETs).

Or indeed one fet and one bjt.

Like https://w3jdr.wordpress.com/lambda-diode-experiments/

Nice circuit! Why isn't that in the AoE?
(at least not in the index)

Nice idea, Arie. What can you do with it?

make stable things unstable. Improve Q.

NT

We like depletion devices, whether JFETs or MOSFETs.
But the -Vgs curve isn't specified or very dependable.
 
On a sunny day (Fri, 5 Jul 2019 13:22:44 +0200) it happened Arie de Muynck
<no.spam@no.spam.org> wrote in <5d1f3304$0$22349$e4fe514c@news.xs4all.nl>:

On 2019-07-05 12:06, Jan Panteltje wrote:
On a sunny day (5 Jul 2019 02:30:02 -0700) it happened Winfield Hill
winfieldhill@yahoo.com> wrote in <qfn5aq0g8p@drn.newsguy.com>:

Arie de Muynck wrote...

On 2019-07-05 09:09, piglet wrote:
On 04/07/2019 10:43 pm, Dave Platt wrote:

Lambda diode (e.g. two JFETs).

Or indeed one fet and one bjt.

Like https://w3jdr.wordpress.com/lambda-diode-experiments/

Nice circuit! Why isn't that in the AoE?
(at least not in the index)

Nice idea, Arie. What can you do with it?

We like depletion devices, whether JFETs or MOSFETs.
But the -Vgs curve isn't specified or very dependable.


One trannny:
0 V +
|
___ |________
| | |
R1 | |
| |/ c |
|---- -| |
| |\|e R3
--- | |
\ / > |----
--- LED R2
| |
--------
|
0 V -

Z1 1.5V LED
R1 10k
R2 1k
R3 10k



At low voltages circuit conducts a lot via transistor and R2,
tranny basically shorts R3.

When V goes high at some point (about 10 V) emitter is lifted bove VZ1 - Vbe,
Ic drops to zero, only current left via R1 and LED plus current through R2 + R3.

Very nice idea too. Reminds me of the SOA foldback in PSU and audio
output amplifier circuits. Problem for me would be the high remaining
current through R3, there is no current gain in that path like with the
BJE + JFET circuit. I need a high (inverse) current ratio between lo and
high voltage load.

Yes, but then again an Esaki (tunnel) diode also has high remaining current,
it just decreases a bit at some point.
He did not ask for 'linear' :)

>BTW - running out of pencils?

No, but this circuit is so simple that ASCII art will do.
It also better preserved as web links may come and go as servers do.
 
On Fri, 5 Jul 2019 04:18:48 -0700 (PDT), tabbypurr@gmail.com wrote:

On Friday, 5 July 2019 00:16:39 UTC+1, John Larkin wrote:
On Thu, 4 Jul 2019 15:51:29 -0700 (PDT), whit3rd <whit3rd@gmail.com
wrote:
On Thursday, July 4, 2019 at 7:35:04 AM UTC-7, John Larkin wrote:

As a college ee project, I built a 2-terminal negative resistance box,
with an opamp and a couple of 9 volt batteries. If you plug in a
negative value for R into the common RC and RLC circuits, the math
still works and the waveforms are radical.

Parallel such a box with a variable resistor, and make a
quartz crystal tester (which starts from white noise when the 'gain' reaches
Barkhausen's criterion).

I'd be thinking of using a LM13700 instead of op amp, though. Transconductance
is a the property you want, it just takes a bias resistor or two.

I used a slow opamp, 709 or maybe 741 by then, so it was only good to
100 KHz or so. Not so good for testing crystals.

Could do watch crystals. I also have a glass valve type encased crystal that's 5 or 6 kHz.


NT

That's sure a collectors item. Big flex mode slab of quartz, likely. I
wonder if the motion would be visible.

I thought this was beautiful, a custom 40 MHz AT cut sealed in vacuum.

https://www.dropbox.com/s/ujiyn8svzee9er4/Lap-Tech.JPG?raw=1

The '60' on the bottom is the turning point temperature.

I used to design my own OCXOs but they have gotten very good and very
cheap lately.


--

John Larkin Highland Technology, Inc trk

jlarkin att highlandtechnology dott com
http://www.highlandtechnology.com
 
Jan Panteltje wrote:
On a sunny day (Thu, 4 Jul 2019 02:29:38 -0800) it happened Robert Baer
robertbaer@localnet.com> wrote in <ZFjTE.67465$GS1.17679@fx48.iad>:

2-terminal negative resistance circuits anyone?

Most especially "Z" types.

Thanks.

neon bulb
Neon bulb more like it; not a circuit but is (mostly) N-type.
Perhaps with some care, one can extract the negative resistance
avalanche-type characteristics.

In any case, is there a SPICE model?
 
tabbypurr@gmail.com wrote:
On Thursday, 4 July 2019 11:24:47 UTC+1, Jan Panteltje wrote:
On a sunny day (Thu, 4 Jul 2019 02:29:38 -0800) it happened Robert Baer
robertbaer@localnet.com> wrote in <ZFjTE.67465$GS1.17679@fx48.iad>:

2-terminal negative resistance circuits anyone?

Most especially "Z" types.

Thanks.

neon bulb

Carbon arc
Fluorescent tube
Pretty much any discharge lamp
Carbon pile with solenoid
Nernst lamp


NT

Not circuits,not SPICE models.
 
piglet wrote:
On 04/07/2019 11:29 am, Robert Baer wrote:
   2-terminal negative resistance circuits anyone?

   Most especially "Z" types.

   Thanks.

Please what is meant by "Z type" ?

piglet
The I/V plot looks like a "Z".
For a real example, take any greater-than 7V zener or any silicon PNP
or NPN low signal transistor EB junction; run curve from zero to 10mA
and you are bound to see the resulting trash.
Some are very noisy.

Thanks.
 

Welcome to EDABoard.com

Sponsor

Back
Top