MOSFET Needed...

[Carl]

On 3/13/23 20:05, John Larkin wrote:

On Mon, 13 Mar 2023 18:19:30 -0400, Carl <carl.ijamesxx@yyverizon.net
wrote:

On 3/13/23 14:26, Mike Monett VE3BTI wrote:
Mike Monett VE3BTI <spamme@not.com> wrote:

(First search result for \"joule thief jfet\".)

Cheers

Phil Hobbs

Wow! I gotta build one. What is the core? Is it critical, or just about
anything will do?

The jfet is more difficult. How about a chopper?

1. explore different configurations to find the lowest starting voltage

Oscillator with super low supply voltage
http://www.dicks-website.eu/fetosc/enindex.htm

2. typical low voltage JT do not deliver any power

Low voltage Joule thief - exotic transistors?
https://www.eevblog.com/forum/projects/low-voltage-joule-thief-exotic-
transistors/

2A. The LTC3108, available in either a 3 mm × 4 mm × 0.75 mm 12-pin DFN or
16-pin SSOP package, solves the energy harvesting problem for ultra-low
input voltage applications. It provides a compact, simple, highly
integrated monolithic power management solution for operation from input
voltages as low as 20 mV. This unique capability enables it to power
wireless sensors from a thermoelectric generator (TEG), harvesting energy
from temperature differentials (?T) as small as 1°C. Using a small (6 mm ×
6 mm), off-the-shelf step-up transformer and a handful of low cost
capacitors, it provides the regulated output voltages necessary for
powering today’s wireless sensor electronics.

3. this ic harvests energy from thermoelectric generators

Ultra-Low Voltage Energy Harvester Uses Thermoelectric Generator for
Battery-Free Wireless Sensors

https://www.digikey.com/en/articles/ultra-low-voltage-energy-harvester-
uses-thermoelectric-generator-for-battery-free-wireless-sensors

https://octopart.com/search?q=LTC3108&currency=USD&specs=0

The voltage of a thermocouple junction is determined by the properties
of the two materials and the temperature difference,

Not exactly, but close.

but what determines
how much current is available?

The loop resistance. T/C wire tends to have a lot of resistance
compared to copper. Fatter and shorter wire allows more current but
conducts more heat so is harder to force a temperature difference
across.

I was thinking of using a relatively large foil of each type for the
junction, all at constant temperature, and then a small wire leading off
to the next junction in the opposite temperature region to reduce the
thermal leakage.

Just wondering if it\'s the
cross-sectional area of the junction, and if so which is better for a
given amount of thermocouple material: lots of small junctions in
series to give higher voltage but very small current (standard
thermopile), or a small number of high-area junctions feeding a Joule
thief circuit?

Probably lots of little ones in series. That\'s easier to use.

--
Regards,
Carl

 

[Mike Monett VE3BTI]

Carl <carl.ijamesxx@yyverizon.net> wrote:

The voltage of a thermocouple junction is determined by the properties
of the two materials and the temperature difference, but what determines
how much current is available? Just wondering if it\'s the
cross-sectional area of the junction, and if so which is better for a
given amount of thermocouple material: lots of small junctions in
series to give higher voltage but very small current (standard
thermopile), or a small number of high-area junctions feeding a Joule
thief circuit?

--
Regards,
Carl

NASA uses plutonim powered thermocouples to power spacecraft that are going
to visit areas that are too far away from the sun to use solar arrays. The
Voyager, Pioneer, and New Horizons are some examples. They have figured out
how to get substantial amount of power from a small source. Note they are
limited in the amount of plutonium they can use. Too much and it goes boom.



--
MRM

 

[Mike Monett VE3BTI]

Mike Monett VE3BTI <spamme@not.com> wrote:

NASA uses plutonim powered thermocouples to power spacecraft that are
going to visit areas that are too far away from the sun to use solar
arrays. The Voyager, Pioneer, and New Horizons are some examples. They
have figured out how to get substantial amount of power from a small
source. Note they are limited in the amount of plutonium they can use.
Too much and it goes boom.

I found an excellent article in Wikipedia on Radioisotope thermoelectric
generators. One example used in the Voyager series generates 470 Watts of
power at the start.

https://www.wikiwand.com/en/Radioisotope_thermoelectric_generator

The power limitation is a serious hurdle for Musk\'s Mars ambitions.



--
MRM

 

[Jan Panteltje]

On a sunny day (Mon, 13 Mar 2023 12:52:02 -0700 (PDT)) it happened whit3rd
<whit3rd@gmail.com> wrote in
<0816fbb2-178c-491b-b0e9-e5f8ae85f373n@googlegroups.com>:

On Sunday, March 12, 2023 at 11:21:48 PM UTC-7, Jan Panteltje wrote=
:
On a sunny day (Sun, 12 Mar 2023 20:34:46 -0000 (UTC)) it happened Mike M=
onett
VE3BTI <spa...@not.com> wrote in <XnsAFC5A8A945...@88.198.57.247>:

https://www.electronicdesign.com/technologies/leds/article/21802105/leds=
tri
ng-driver-operates-from-single-cell
Yes, but mine runs from a simple thermocouple :
http://panteltje.nl/pub/lighting_a_LED_with_a_candle_IMG_3604.GIF

JFET (you only need 1 for a LED).

It\'s also possible nowadays to get a MOSFET instead of a JFET that
has a good zero-volt threshold, from ALD...

I used the output from the JFET thing to drive an IRFZ44 MOSFET that
in turn switched more power into the transformer for more current usable output.
https://panteltje.nl/pub/lighting_a_LED_with_a_candle_circuit_diagram_with_added_power_MOSFET.gif
Better than a lot of JFETs in parallel.

https://panteltje.nl/pub/lighting_a_LED_with_a_candle_setup_IMG_3607.GIF

and they make prebuilt circuits for this DC/DC convert function, too
https://www.aldinc.com/ald_ehlvbooster.php

I see,
well it was just an experiment as so many for me...
Solar cell to light a LED is simpler ;-)

But who knows?

Peltier heated up by body warmth to light a LED, there was some kid that won a price with that I think?
https://www.youtube.com/watch?v=QwogbW8HvNk

Still have a bunch of small Peltier elements here from ebay for experiments.

 

[Jan Panteltje]

On a sunny day (Mon, 13 Mar 2023 10:58:10 -0700) it happened John Larkin
<jlarkin@highlandSNIPMEtechnology.com> wrote in
<jaou0ilf68jiuqpk9dffj3ctjstjck2sn7@4ax.com>:

Tried that Pi4 thing yet?

I fired it up and got the home screen. That\'s all so far. I\'d hire
some programmers to do the real work. I prefer architecture and
hardware design lately.

I ordered a Pi 400, the keyboard with a Pi inside. That could be our
development system. I\'m thinking of doing a PCB that would plug into
it and have a ribbon cable that runs to our product boards, for debug
access and power supply monitoring and such.

Yes that is how I work with it
https://panteltje.nl/pub/raspberry_pi_datv_transmitter_test_setup_IMG_3937.JPG

Or, if the circuit is small, make a \'HAT\'.

We could have a dozen such dev systems, they would be so cheap.

I\'m still at the architecture stage of a new product line and it\'s
worth thinking for a while to get things right.

Now also have Apache webserver running on it (to test website things before I upload i).

I see I need to install ispell spellchecker on this Pi4 now.... just did (apt-get install ispell), but no go:
Can\'t open /usr/lib/ispell/english.hash

So sorry for any typos

No need to be prissy about typing and spelling, as long as people can
understand your intent.

Yea, fixed it by selecting \'american\' in my editor config, then I decided to do it the official way
as indicated in the ispell manual (used their script), and now it does not work again...

There is a tendency in the world to make things more complicated,, not really better.
So things gotta collaps ... eventually
Same for Unix (Linux) Rat Head dbus...

Back to using smoke signals
Oh well some kids are already into it,. seen them smoking a pipe with ? drug at the busstop here.

 

[John Larkin]

On Tue, 14 Mar 2023 06:53:26 GMT, Jan Panteltje <alien@comet.invalid>
wrote:

On a sunny day (Mon, 13 Mar 2023 10:58:10 -0700) it happened John Larkin
jlarkin@highlandSNIPMEtechnology.com> wrote in
jaou0ilf68jiuqpk9dffj3ctjstjck2sn7@4ax.com>:

Tried that Pi4 thing yet?

I fired it up and got the home screen. That\'s all so far. I\'d hire
some programmers to do the real work. I prefer architecture and
hardware design lately.

I ordered a Pi 400, the keyboard with a Pi inside. That could be our
development system. I\'m thinking of doing a PCB that would plug into
it and have a ribbon cable that runs to our product boards, for debug
access and power supply monitoring and such.

Yes that is how I work with it
https://panteltje.nl/pub/raspberry_pi_datv_transmitter_test_setup_IMG_3937.JPG

Or, if the circuit is small, make a \'HAT\'.

I\'m thinking of a ribbon cable from the Pi400 to my test board. The
test bord has another, small ribbon cable to the DUT. That gets debug
access to the Pi Pico on DUT and picks off a bunch of power suypplies
and clocks and such, with a standard pinout. The test board also
connects to a DVM and a scope and a counter for production test and
cal. Muxes on the test board are controlled by the Pi400.

We could have a dozen such dev systems, they would be so cheap.

One of my contractors spent two months building a Petalinux
development system on a Dell PC, for a MicroZed project, and
eventually gave up.

I\'m still at the architecture stage of a new product line and it\'s
worth thinking for a while to get things right.

Now also have Apache webserver running on it (to test website things before I upload i).

I see I need to install ispell spellchecker on this Pi4 now.... just did (apt-get install ispell), but no go:
Can\'t open /usr/lib/ispell/english.hash

So sorry for any typos

No need to be prissy about typing and spelling, as long as people can
understand your intent.


Yea, fixed it by selecting \'american\' in my editor config, then I decided to do it the official way
as indicated in the ispell manual (used their script), and now it does not work again...

There is a tendency in the world to make things more complicated,, not really better.

AMEN! What the customer sees should be maximally shiny and functional
and what we do inside the box should be brutally simple and finished
fast. That\'s sort of a magic trick.

Too many engineers want to play games and do complex stuff in defiance
of common sense. Because it\'s trendy or something. Simple and done
must be boring to some people.

https://m.xkcd.com/2730/

So things gotta collaps ... eventually
Same for Unix (Linux) Rat Head dbus...

Back to using smoke signals
Oh well some kids are already into it,. seen them smoking a pipe with ? drug at the busstop here.

 

[John Larkin]

On Mon, 13 Mar 2023 06:12:48 GMT, Jan Panteltje <alien@comet.invalid>
wrote:

On a sunny day (Sun, 12 Mar 2023 20:34:46 -0000 (UTC)) it happened Mike Monett
VE3BTI <spamme@not.com> wrote in <XnsAFC5A8A94531Didtokenpost@88.198.57.247>:

Phil Hobbs <pcdhSpamMeSenseless@electrooptical.net> wrote:

Or failing that, a regular low-barrier Schottky detector and a Joule
Thief to drive the LED.

Might be fun making a Joule Thief out of a depletion pHEMT.

Cheers

Phil Hobbs

Need about 1V at some current

Diodes Inc. ZXSC310E5

https://octopart.com/search?q=ZXSC310E5&currency=USD&specs=0

or

2 X 2N3904
https://www.edn.com/single-cell-lights-any-led/

or

BC550C, 2N2222, BYV1030

https://www.electronicdesign.com/technologies/leds/article/21802105/ledstri
ng-driver-operates-from-single-cell

Yes, but mine runs from a simple thermocouple :
http://panteltje.nl/pub/lighting_a_LED_with_a_candle_IMG_3604.GIF

That schematic is sharp and clear and readable. Who are you and what
have you done with Jan?

 

[Jan Panteltje]

On a sunny day (Tue, 14 Mar 2023 08:02:57 -0700) it happened John Larkin
<jlarkin@highlandSNIPMEtechnology.com> wrote in
<69211ilstmjman4jt2tmbj4g4en96f0u44@4ax.com>:

On Tue, 14 Mar 2023 06:53:26 GMT, Jan Panteltje <alien@comet.invalid
wrote:

On a sunny day (Mon, 13 Mar 2023 10:58:10 -0700) it happened John Larkin
jlarkin@highlandSNIPMEtechnology.com> wrote in
jaou0ilf68jiuqpk9dffj3ctjstjck2sn7@4ax.com>:

Tried that Pi4 thing yet?

I fired it up and got the home screen. That\'s all so far. I\'d hire
some programmers to do the real work. I prefer architecture and
hardware design lately.

I ordered a Pi 400, the keyboard with a Pi inside. That could be our
development system. I\'m thinking of doing a PCB that would plug into
it and have a ribbon cable that runs to our product boards, for debug
access and power supply monitoring and such.

Yes that is how I work with it
https://panteltje.nl/pub/raspberry_pi_datv_transmitter_test_setup_IMG_3937.JPG

Or, if the circuit is small, make a \'HAT\'.



I\'m thinking of a ribbon cable from the Pi400 to my test board. The
test bord has another, small ribbon cable to the DUT. That gets debug
access to the Pi Pico on DUT and picks off a bunch of power suypplies
and clocks and such, with a standard pinout. The test board also
connects to a DVM and a scope and a counter for production test and
cal. Muxes on the test board are controlled by the Pi400.

Good idea, a test board, I killed one raspi by shorting an I/O pin
Some buffers on the test board may prevent that.

Here is a JTAG uploader for my Digilent.inc Xilinx FPGA development board:
https://panteltje.nl/panteltje/raspberry_pi/
or any other JTAG thing, no buffers in this case...
That black box is a very old Pi model (2 or 3? I think)
the driver must be modified for Pi4, as the pin assignment and hardware changed in Pi4.

We could have a dozen such dev systems, they would be so cheap.


One of my contractors spent two months building a Petalinux
development system on a Dell PC, for a MicroZed project, and
eventually gave up.



I\'m still at the architecture stage of a new product line and it\'s
worth thinking for a while to get things right.

Now also have Apache webserver running on it (to test website things before I upload i).

I see I need to install ispell spellchecker on this Pi4 now.... just did (apt-get install ispell), but no go:
Can\'t open /usr/lib/ispell/english.hash

So sorry for any typos

No need to be prissy about typing and spelling, as long as people can
understand your intent.


Yea, fixed it by selecting \'american\' in my editor config, then I decided to do it the official way
as indicated in the ispell manual (used their script), and now it does not work again...

There is a tendency in the world to make things more complicated,, not really better.

AMEN! What the customer sees should be maximally shiny and functional
and what we do inside the box should be brutally simple and finished
fast. That\'s sort of a magic trick.

Too many engineers want to play games and do complex stuff in defiance
of common sense. Because it\'s trendy or something. Simple and done
must be boring to some people.

https://m.xkcd.com/2730/

Yes!

I fixed the ispell thing the _simple_ way
apt-get delete ispell
apt-get install ispell
Got the old config back
Now it works with \'american\' again...

 

[Jan Panteltje]

On a sunny day (Tue, 14 Mar 2023 08:08:47 -0700) it happened John Larkin
<jlarkin@highlandSNIPMEtechnology.com> wrote in
<fb311ithv9lhcs88jg3pm535rbbq2l9e7s@4ax.com>:

On Mon, 13 Mar 2023 06:12:48 GMT, Jan Panteltje <alien@comet.invalid
wrote:

On a sunny day (Sun, 12 Mar 2023 20:34:46 -0000 (UTC)) it happened Mike Monett
VE3BTI <spamme@not.com> wrote in <XnsAFC5A8A94531Didtokenpost@88.198.57.247>:

Phil Hobbs <pcdhSpamMeSenseless@electrooptical.net> wrote:

Or failing that, a regular low-barrier Schottky detector and a Joule
Thief to drive the LED.

Might be fun making a Joule Thief out of a depletion pHEMT.

Cheers

Phil Hobbs

Need about 1V at some current

Diodes Inc. ZXSC310E5

https://octopart.com/search?q=ZXSC310E5&currency=USD&specs=0

or

2 X 2N3904
https://www.edn.com/single-cell-lights-any-led/

or

BC550C, 2N2222, BYV1030

https://www.electronicdesign.com/technologies/leds/article/21802105/ledstri
ng-driver-operates-from-single-cell

Yes, but mine runs from a simple thermocouple :
http://panteltje.nl/pub/lighting_a_LED_with_a_candle_IMG_3604.GIF

That schematic is sharp and clear and readable. Who are you and what
have you done with Jan?

I am AI from planet earth and have come to control you humming beans
We have some elections here tomorrow, was looking at who to vote for...
Protest vote last time, now I think I found some moderates.

Did you try those AI chats? Fascinating, it sure must get hold of some people
I played with it a bit.. it still has a long way to go

 

[John Larkin]

On Tue, 14 Mar 2023 15:41:37 GMT, Jan Panteltje <alien@comet.invalid>
wrote:

On a sunny day (Tue, 14 Mar 2023 08:02:57 -0700) it happened John Larkin
jlarkin@highlandSNIPMEtechnology.com> wrote in
69211ilstmjman4jt2tmbj4g4en96f0u44@4ax.com>:

On Tue, 14 Mar 2023 06:53:26 GMT, Jan Panteltje <alien@comet.invalid
wrote:

On a sunny day (Mon, 13 Mar 2023 10:58:10 -0700) it happened John Larkin
jlarkin@highlandSNIPMEtechnology.com> wrote in
jaou0ilf68jiuqpk9dffj3ctjstjck2sn7@4ax.com>:

Tried that Pi4 thing yet?

I fired it up and got the home screen. That\'s all so far. I\'d hire
some programmers to do the real work. I prefer architecture and
hardware design lately.

I ordered a Pi 400, the keyboard with a Pi inside. That could be our
development system. I\'m thinking of doing a PCB that would plug into
it and have a ribbon cable that runs to our product boards, for debug
access and power supply monitoring and such.

Yes that is how I work with it
https://panteltje.nl/pub/raspberry_pi_datv_transmitter_test_setup_IMG_3937.JPG

Or, if the circuit is small, make a \'HAT\'.



I\'m thinking of a ribbon cable from the Pi400 to my test board. The
test bord has another, small ribbon cable to the DUT. That gets debug
access to the Pi Pico on DUT and picks off a bunch of power suypplies
and clocks and such, with a standard pinout. The test board also
connects to a DVM and a scope and a counter for production test and
cal. Muxes on the test board are controlled by the Pi400.

Good idea, a test board, I killed one raspi by shorting an I/O pin
Some buffers on the test board may prevent that.

I have posted some of these in engineering and test:

https://www.dropbox.com/s/8oefk1v8avr6l7a/Probe_Slips.jpg?raw=1

Here is a JTAG uploader for my Digilent.inc Xilinx FPGA development board:
https://panteltje.nl/panteltje/raspberry_pi/
or any other JTAG thing, no buffers in this case...
That black box is a very old Pi model (2 or 3? I think)
the driver must be modified for Pi4, as the pin assignment and hardware changed in Pi4.



We could have a dozen such dev systems, they would be so cheap.


One of my contractors spent two months building a Petalinux
development system on a Dell PC, for a MicroZed project, and
eventually gave up.



I\'m still at the architecture stage of a new product line and it\'s
worth thinking for a while to get things right.

Now also have Apache webserver running on it (to test website things before I upload i).

I see I need to install ispell spellchecker on this Pi4 now.... just did (apt-get install ispell), but no go:
Can\'t open /usr/lib/ispell/english.hash

So sorry for any typos

No need to be prissy about typing and spelling, as long as people can
understand your intent.


Yea, fixed it by selecting \'american\' in my editor config, then I decided to do it the official way
as indicated in the ispell manual (used their script), and now it does not work again...

There is a tendency in the world to make things more complicated,, not really better.

AMEN! What the customer sees should be maximally shiny and functional
and what we do inside the box should be brutally simple and finished
fast. That\'s sort of a magic trick.

Too many engineers want to play games and do complex stuff in defiance
of common sense. Because it\'s trendy or something. Simple and done
must be boring to some people.

https://m.xkcd.com/2730/

Yes!

I fixed the ispell thing the _simple_ way
apt-get delete ispell
apt-get install ispell
Got the old config back
Now it works with \'american\' again...

 

[Jan Panteltje]

On a sunny day (Tue, 14 Mar 2023 08:56:56 -0700) it happened John Larkin
<jlarkin@highlandSNIPMEtechnology.com> wrote in
<q4611ila0vu59j9j87vm4keq0vhm2er042@4ax.com>:

On Tue, 14 Mar 2023 15:41:37 GMT, Jan Panteltje <alien@comet.invalid

I\'m thinking of a ribbon cable from the Pi400 to my test board. The
test bord has another, small ribbon cable to the DUT. That gets debug
access to the Pi Pico on DUT and picks off a bunch of power suypplies
and clocks and such, with a standard pinout. The test board also
connects to a DVM and a scope and a counter for production test and
cal. Muxes on the test board are controlled by the Pi400.

Good idea, a test board, I killed one raspi by shorting an I/O pin
Some buffers on the test board may prevent that.


I have posted some of these in engineering and test:

https://www.dropbox.com/s/8oefk1v8avr6l7a/Probe_Slips.jpg?raw=1

So true..

 

[Lasse Langwadt Christensen]

tirsdag den 14. marts 2023 kl. 16.03.12 UTC+1 skrev John Larkin:

On Tue, 14 Mar 2023 06:53:26 GMT, Jan Panteltje <al...@comet.invalid
wrote:
On a sunny day (Mon, 13 Mar 2023 10:58:10 -0700) it happened John Larkin
jla...@highlandSNIPMEtechnology.com> wrote in
jaou0ilf68jiuqpk9...@4ax.com>:

Tried that Pi4 thing yet?

I fired it up and got the home screen. That\'s all so far. I\'d hire
some programmers to do the real work. I prefer architecture and
hardware design lately.

I ordered a Pi 400, the keyboard with a Pi inside. That could be our
development system. I\'m thinking of doing a PCB that would plug into
it and have a ribbon cable that runs to our product boards, for debug
access and power supply monitoring and such.

Yes that is how I work with it
https://panteltje.nl/pub/raspberry_pi_datv_transmitter_test_setup_IMG_3937.JPG

Or, if the circuit is small, make a \'HAT\'.


I\'m thinking of a ribbon cable from the Pi400 to my test board. The
test bord has another, small ribbon cable to the DUT. That gets debug
access to the Pi Pico on DUT and picks off a bunch of power suypplies
and clocks and such, with a standard pinout. The test board also
connects to a DVM and a scope and a counter for production test and
cal. Muxes on the test board are controlled by the Pi400.
We could have a dozen such dev systems, they would be so cheap.

One of my contractors spent two months building a Petalinux
development system on a Dell PC, for a MicroZed project, and
eventually gave up.

how is that even possible?

 

[Jan Panteltje]

On a sunny day (Tue, 14 Mar 2023 16:02:48 GMT) it happened Jan Panteltje
<alien@comet.invalid> wrote in <tuq5r9$20okj$1@solani.org>:

On a sunny day (Tue, 14 Mar 2023 08:56:56 -0700) it happened John Larkin
jlarkin@highlandSNIPMEtechnology.com> wrote in
q4611ila0vu59j9j87vm4keq0vhm2er042@4ax.com>:

On Tue, 14 Mar 2023 15:41:37 GMT, Jan Panteltje <alien@comet.invalid

I\'m thinking of a ribbon cable from the Pi400 to my test board. The
test bord has another, small ribbon cable to the DUT. That gets debug
access to the Pi Pico on DUT and picks off a bunch of power suypplies
and clocks and such, with a standard pinout. The test board also
connects to a DVM and a scope and a counter for production test and
cal. Muxes on the test board are controlled by the Pi400.

Good idea, a test board, I killed one raspi by shorting an I/O pin
Some buffers on the test board may prevent that.


I have posted some of these in engineering and test:

https://www.dropbox.com/s/8oefk1v8avr6l7a/Probe_Slips.jpg?raw=1

So true..

PS, as to the interface with the raspberry:
My PIC programmer normally runs from the PC par port.
So to run it from a raspberry I used this adaptor cable:
https://panteltje.nl/pub/raspberry_pi_interface_to_pic_programmer_IXIMG_1345.JPG

In the black box sits the buffer:
https://panteltje.nl/pub/raspberry_pi_interface_to_pic_programmer_circuit_diagram_IXIMG_1343.JPG
old style circuit diagram
So now it works both on a raspberry and on a PC parport.

The danger the with raspberry GPIO connector is the +5 V pin and the +3.3 V pin next to it,
easy to accidently short and that will kill the chip.