BJT behaviour at ridiculously low current levels...

On 5/14/2022 1:03 PM, piglet wrote:
On 14/05/2022 07:04, Hul Tytus wrote:
Piotr  you might look at the hc cmos series. They will operate in
the 1 volt range with the expected lower current. One maker spoke
of characterizing their devices at one something volts.
    I experimented with those supply voltages for devices
working in the 100k cps area and all seemed pheasable.

Hul


Yes, I actually experienced HC cmos retaining stored latch states down
into tens of millivolts region. However to make any kind of oscillator
means passing slowly through a linear region and current consumption
will shoot up. To make an oscillator a discrete design with high value
series resistors might be better, this draws an average 10nA to make
narrow pulses every 8-9 secs...

Version 4
[Snip SPICE listing]
piglet
Click to expand...

Keep everything clean while building!
 
jlarkin@highlandsniptechnology.com wrote:
On Tue, 24 May 2022 10:43:08 -0400, Phil Hobbs
pcdhSpamMeSenseless@electrooptical.net> wrote:

John Walliker wrote:
On Tuesday, 24 May 2022 at 09:08:21 UTC+1, whit3rd wrote:
On Monday, May 23, 2022 at 3:29:10 PM UTC-7, John Larkin wrote:
On Mon, 23 May 2022 11:06:00 -0700 (PDT), whit3rd <whi...@gmail.com
wrote:
On Monday, May 23, 2022 at 10:19:30 AM UTC-7, John Larkin wrote:

I want an LED blinker to show that a high-voltage power supply is
still dangerous. It should discharge the supply at some low current
and blink from, say, 1400 volts down to maybe 40.

So, a series string of fifteen current-limit diodes running a capacitor-
diac-LED blinkielight? S-272 seems suitable

https://www.mouser.com/datasheet/2/362/P22-23-CRD-1729293.pdf
The Supertex LND150 is a 500 volt depletion fet. A source resistor
will program the current, and they can apparently be strung in series
without problems.
Yeah, and a \'current-limit diode\' is also a depletion FET. With a zener
limit specified (I\'m not sure about LND150). Multiples only share
the voltage with such a matching zener specification.

A resistor would work, too, but the alarm blinkie would change frequency
dramatically.

Does anyone actually sell surface mount LND150s at the moment? All
the usual suppliers seem to be out of stock for at least several months.

John

TO-92s are good for the soul.

Rarely! But we\'re using a TO-92 LM35 to measure the temp of a CPU
cooler that we\'re using as a mosfet heat sink. We\'ll glue it to the
side of the cooler and plug it into three little pin sockets.


Cheers

Phil Hobbs

(Who just designed one into a supply-dictated board spin.) :(

Are TO-92s easier to get than surfmount parts? At least it\'s an
option.
Click to expand...

Some of them.

Cheers

Phil Hobbs


--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC / Hobbs ElectroOptics
Optics, Electro-optics, Photonics, Analog Electronics
Briarcliff Manor NY 10510

http://electrooptical.net
http://hobbs-eo.com
 
jlarkin@highlandsniptechnology.com wrote:
On Tue, 24 May 2022 10:08:29 -0400, Phil Hobbs
pcdhSpamMeSenseless@electrooptical.net> wrote:

John Larkin wrote:
On Mon, 23 May 2022 11:06:00 -0700 (PDT), whit3rd <whit3rd@gmail.com
wrote:

On Monday, May 23, 2022 at 10:19:30 AM UTC-7, John Larkin wrote:

I want an LED blinker to show that a high-voltage power supply is
still dangerous. It should discharge the supply at some low current
and blink from, say, 1400 volts down to maybe 40.

So, a series string of fifteen current-limit diodes running a capacitor-
diac-LED blinkielight? S-272 seems suitable

https://www.mouser.com/datasheet/2/362/P22-23-CRD-1729293.pdf

The Supertex LND150 is a 500 volt depletion fet. A source resistor
will program the current, and they can apparently be strung in series
without problems.

Three of them would do. They are around 30 cents each.

A spin on Piglet\'s circuit might to the blinker.


I\'d probably be happier with a gate bias string to force the voltages to
equalize. Of course the resistors would have to be physically quite
large to take the voltage.



Somebody told me (was it you?) that they are OK in series. One or two
may avalanche somewhere above 500 volts and one will actually regulate
the current. I\'m talking 50 or 100 uA in my case.



I\'ve never used them above about 250V.
Click to expand...

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC / Hobbs ElectroOptics
Optics, Electro-optics, Photonics, Analog Electronics
Briarcliff Manor NY 10510

http://electrooptical.net
http://hobbs-eo.com
 
On Tue, 24 May 2022 16:47:30 -0400, Phil Hobbs
<pcdhSpamMeSenseless@electrooptical.net> wrote:

jlarkin@highlandsniptechnology.com wrote:
On Tue, 24 May 2022 10:08:29 -0400, Phil Hobbs
pcdhSpamMeSenseless@electrooptical.net> wrote:

John Larkin wrote:
On Mon, 23 May 2022 11:06:00 -0700 (PDT), whit3rd <whit3rd@gmail.com
wrote:

On Monday, May 23, 2022 at 10:19:30 AM UTC-7, John Larkin wrote:

I want an LED blinker to show that a high-voltage power supply is
still dangerous. It should discharge the supply at some low current
and blink from, say, 1400 volts down to maybe 40.

So, a series string of fifteen current-limit diodes running a capacitor-
diac-LED blinkielight? S-272 seems suitable

https://www.mouser.com/datasheet/2/362/P22-23-CRD-1729293.pdf

The Supertex LND150 is a 500 volt depletion fet. A source resistor
will program the current, and they can apparently be strung in series
without problems.

Three of them would do. They are around 30 cents each.

A spin on Piglet\'s circuit might to the blinker.


I\'d probably be happier with a gate bias string to force the voltages to
equalize. Of course the resistors would have to be physically quite
large to take the voltage.



Somebody told me (was it you?) that they are OK in series. One or two
may avalanche somewhere above 500 volts and one will actually regulate
the current. I\'m talking 50 or 100 uA in my case.



I\'ve never used them above about 250V.

Cheers

Phil Hobbs
Click to expand...

I tested one. It avalanches at 620 volts, no damage at low current.

The gate zeners at about +-40, again no damage.



--

Anybody can count to one.

- Robert Widlar
 
On 22/05/2022 7:32 pm, Piotr Wyderski wrote:
piglet wrote:

To make an oscillator a discrete design with high value series
resistors might be better, this draws an average 10nA to make narrow
pulses every 8-9 secs...

After some prototyping I can confirm, your UJT-emulating circuit wins
hands down. I removed R6, R2, Q3 and R5, changed C1 and C2 to 470p and
coupled to the C2 discharge energy via a small transformer, getting
beautiful 1us spikes @28Hz. Thank you, Piglet!

    Best regards, Piotr
Click to expand...

You are very welcome, always pleased when a dirt cheap descrete circuit
still beats the world. It is not my design, the complementary series
multivibrator has been around since the 1960s.

piglet
 
On 23/05/2022 11:28 pm, John Larkin wrote:
On Mon, 23 May 2022 11:06:00 -0700 (PDT), whit3rd <whit3rd@gmail.com
wrote:

On Monday, May 23, 2022 at 10:19:30 AM UTC-7, John Larkin wrote:

I want an LED blinker to show that a high-voltage power supply is
still dangerous. It should discharge the supply at some low current
and blink from, say, 1400 volts down to maybe 40.

So, a series string of fifteen current-limit diodes running a capacitor-
diac-LED blinkielight? S-272 seems suitable

https://www.mouser.com/datasheet/2/362/P22-23-CRD-1729293.pdf

The Supertex LND150 is a 500 volt depletion fet. A source resistor
will program the current, and they can apparently be strung in series
without problems.

Three of them would do. They are around 30 cents each.

A spin on Piglet\'s circuit might to the blinker.
Click to expand...

My preference would be to skip the depletion fets and just use dropper
resistors - cheaper, more reliable, and best of all the non constant
blink rate gives a good visual confirmation that the HV capacitor is
actually discharging.

Blinking could be as simple as a diac or bjt used as negistor. Spice
won\'t model E-B breakdown and reverse beta so can\'t simulate negistor
operation but breadboarding would be very quick.

piglet
 
piglet wrote:

You are very welcome, always pleased when a dirt cheap descrete circuit
still beats the world. It is not my design, the complementary series
multivibrator has been around since the 1960s.
Click to expand...

This is how it looks now. L1 and L3 pump the flyback core, L3 boosts the
voltage of the core reset pulse to 600mV+, even at 1.2V_IN, R5 steals
that pulse (or not), Q3 detects if there was the pulse, C3/Q4 is a
zero-power monostable (assuming there are no pulses from Q3) and M1
makes that information accessible to the external world. It is shocking
how simple it is and how well it works.

Best regards, Piotr

Version 4
SHEET 1 2136 1236
WIRE 416 -176 176 -176
WIRE 720 -176 416 -176
WIRE 1264 -176 720 -176
WIRE 1648 -176 1264 -176
WIRE 720 -128 720 -176
WIRE 416 -80 416 -176
WIRE 720 48 720 -48
WIRE 960 48 720 48
WIRE 960 64 960 48
WIRE 720 96 720 48
WIRE 416 144 416 0
WIRE 656 144 416 144
WIRE 960 160 960 144
WIRE 416 224 416 144
WIRE 720 272 720 192
WIRE 960 320 960 224
WIRE 960 320 784 320
WIRE 960 384 960 320
WIRE 176 528 176 -176
WIRE 416 576 416 288
WIRE 528 576 416 576
WIRE 2080 608 1904 608
WIRE 416 624 416 576
WIRE 528 624 528 576
WIRE 720 640 720 368
WIRE 2080 656 2080 608
WIRE 1904 736 1904 608
WIRE 176 752 176 608
WIRE 416 752 416 704
WIRE 416 752 176 752
WIRE 720 752 720 720
WIRE 720 752 416 752
WIRE 960 752 960 464
WIRE 960 752 720 752
WIRE 1264 768 1264 -176
WIRE 2080 768 2080 736
WIRE 176 784 176 752
WIRE 1648 816 1648 -176
WIRE 1264 864 1264 848
WIRE 1392 864 1392 784
WIRE 1392 864 1264 864
WIRE 1472 864 1392 864
WIRE 1584 864 1552 864
WIRE 1264 880 1264 864
WIRE 416 896 272 896
WIRE 1392 896 1392 864
WIRE 272 928 272 896
WIRE 528 928 528 704
WIRE 1056 928 528 928
WIRE 1200 928 1136 928
WIRE 1648 928 1648 912
WIRE 1776 928 1648 928
WIRE 1904 928 1904 816
WIRE 1648 944 1648 928
WIRE 1392 976 1392 960
WIRE 1072 992 1072 976
WIRE 416 1008 416 976
WIRE 1776 1008 1776 928
WIRE 1856 1008 1776 1008
WIRE 1264 1024 1264 976
WIRE 1120 1040 1120 976
WIRE 1648 1040 1648 1024
WIRE 1904 1040 1904 1024
WIRE 1120 1136 1120 1120
FLAG 176 784 0
FLAG 416 1008 0
FLAG 272 1040 0
FLAG 1264 1024 0
FLAG 1392 784 V_OUT
FLAG 1648 1040 0
FLAG 1392 976 0
FLAG 1072 992 0
FLAG 1120 1136 0
FLAG 2080 768 0
FLAG 1904 1040 0
SYMBOL voltage 176 512 R0
WINDOW 123 0 0 Left 0
WINDOW 39 0 0 Left 0
SYMATTR InstName V1
SYMATTR Value 3
SYMBOL pnp 784 368 R180
SYMATTR InstName Q1
SYMATTR Value BC557C
SYMBOL npn 656 96 R0
SYMATTR InstName Q2
SYMATTR Value BC547C
SYMBOL res 704 -144 R0
SYMATTR InstName R1
SYMATTR Value 10Meg
SYMBOL res 944 368 R0
SYMATTR InstName R2
SYMATTR Value {R_OSC}
SYMBOL res 400 -96 R0
SYMATTR InstName R3
SYMATTR Value {R_OSC}
SYMBOL cap 400 224 R0
SYMATTR InstName C1
SYMATTR Value {C_OSC}
SYMBOL cap 944 160 R0
SYMATTR InstName C2
SYMATTR Value {C_OSC}
SYMBOL res 704 624 R0
SYMATTR InstName R4
SYMATTR Value 1m
SYMBOL ind2 400 608 R0
WINDOW 0 -42 23 Left 2
WINDOW 3 -102 64 Left 2
SYMATTR InstName L1
SYMATTR Value {L_PRI}
SYMATTR Type ind
SYMATTR SpiceLine Cpar=0
SYMBOL ind2 400 880 R0
WINDOW 3 -55 -7 Left 2
SYMATTR Value {L_SENSE}
SYMATTR InstName L2
SYMATTR Type ind
SYMBOL ind2 544 720 R180
WINDOW 0 -41 86 Left 2
WINDOW 3 -111 59 Left 2
SYMATTR InstName L3
SYMATTR Value {L_SEC}
SYMATTR Type ind
SYMATTR SpiceLine Cpar=0
SYMBOL res 256 912 R0
SYMATTR InstName R5
SYMATTR Value 33
SYMBOL ind2 944 48 R0
SYMATTR InstName L4
SYMATTR Value {L_PRI}
SYMATTR Type ind
SYMBOL npn 1200 880 R0
SYMATTR InstName Q3
SYMATTR Value BC547C
SYMBOL res 1248 752 R0
SYMATTR InstName R6
SYMATTR Value 10Meg
SYMBOL pnp 1584 912 M180
SYMATTR InstName Q4
SYMATTR Value BC557C
SYMBOL res 1568 848 R90
WINDOW 0 0 56 VBottom 2
WINDOW 3 32 56 VTop 2
SYMATTR InstName R7
SYMATTR Value 1Meg
SYMBOL res 1632 928 R0
WINDOW 0 -56 28 Left 2
WINDOW 3 -91 53 Left 2
SYMATTR InstName R8
SYMATTR Value 10Meg
SYMBOL cap 1376 896 R0
SYMATTR InstName C3
SYMATTR Value 33n
SYMBOL sw 1040 928 R270
SYMATTR InstName S1
SYMATTR Value MYSW
SYMBOL voltage 1120 1024 R0
WINDOW 123 0 0 Left 0
WINDOW 39 0 0 Left 0
SYMATTR InstName V2
SYMATTR Value PULSE(0 1 300m 10n 10n 300m 1)
SYMBOL res 1888 720 R0
SYMATTR InstName R10
SYMATTR Value 1000
SYMBOL voltage 2080 640 R0
WINDOW 123 0 0 Left 0
WINDOW 39 0 0 Left 0
SYMATTR InstName V3
SYMATTR Value 12
SYMBOL nmos 1856 928 R0
SYMATTR InstName M1
SYMATTR Value BSS123
TEXT 216 -368 Left 2 !.param C_OSC=470p
TEXT 216 -344 Left 2 !.param R_OSC=100Meg
TEXT 560 -424 Left 2 !.param AL=5000n
TEXT 560 -400 Left 2 !.param N_PRI=26
TEXT 560 -376 Left 2 !.param N_SEC={N_PRI*4}
TEXT 560 -352 Left 2 !.param N_SENSE=8
TEXT 560 -320 Left 2 !.param L_PRI={N_PRI*N_PRI*AL}
TEXT 560 -296 Left 2 !.param L_SEC={N_SEC*N_SEC*AL}
TEXT 560 -272 Left 2 !.param L_SENSE={N_SENSE*N_SENSE*AL}
TEXT 240 800 Left 2 !K L1 L2 L3 L4 0.995
TEXT 144 1048 Left 2 !.tran 1
TEXT 912 -376 Left 2 ;PRI/SEC: twisted 6x26 DNE0.2mm @ 10mm OD ring F938.
TEXT 912 -352 Left 2 ;SENSE: TIW!!!
TEXT 1376 1048 Left 2 !.ic V(V_OUT)=0
TEXT 1008 1192 Left 2 !.model MYSW SW(Ron=1m Roff=1G Vt=.5 Vh=-.4)
 
Piotr Wyderski wrote:

Just a supplement: the interesting part is IMO discharging the caps
through the flyback chokes. It makes the pulses at no additional energy
cost (beyond what has already been transferred through the 100M
resistors), so they are both signal and power at the same time.
Unfortunately, the amount of energy stored in that 470p capacitor
charged to 1V is so tiny that the autotransformer is not able to boost
the voltage (merely ~150mV at the low V_IN end) to open Q3. This is a
constant energy regime: if you add more turns to L3, the voltage reaches
some saturation point and the reset pulses are getting longer instead,
as L and C both go up. This saturation point is ~400mV, too low for an
NPN to notice (assuming zero power detection; you can\'t afford a proper
PNP long tailed pair). A 400mV EPAD would cut the mustard, but another
solution is to double the amount of energy deposited to the core. Hence
the dual-pumping idea utilising L1 and L4. With that it works even down
to 1.1V V_IN. Balancing these available picojoules is sort of obscene. :)

Best regards, Piotr
 
jlarkin@highlandsniptechnology.com wrote:
On Tue, 24 May 2022 16:47:30 -0400, Phil Hobbs
pcdhSpamMeSenseless@electrooptical.net> wrote:

jlarkin@highlandsniptechnology.com wrote:
On Tue, 24 May 2022 10:08:29 -0400, Phil Hobbs
pcdhSpamMeSenseless@electrooptical.net> wrote:

John Larkin wrote:
On Mon, 23 May 2022 11:06:00 -0700 (PDT), whit3rd <whit3rd@gmail.com
wrote:

On Monday, May 23, 2022 at 10:19:30 AM UTC-7, John Larkin wrote:

I want an LED blinker to show that a high-voltage power supply is
still dangerous. It should discharge the supply at some low current
and blink from, say, 1400 volts down to maybe 40.

So, a series string of fifteen current-limit diodes running a capacitor-
diac-LED blinkielight? S-272 seems suitable

https://www.mouser.com/datasheet/2/362/P22-23-CRD-1729293.pdf

The Supertex LND150 is a 500 volt depletion fet. A source resistor
will program the current, and they can apparently be strung in series
without problems.

Three of them would do. They are around 30 cents each.

A spin on Piglet\'s circuit might to the blinker.


I\'d probably be happier with a gate bias string to force the voltages to
equalize. Of course the resistors would have to be physically quite
large to take the voltage.



Somebody told me (was it you?) that they are OK in series. One or two
may avalanche somewhere above 500 volts and one will actually regulate
the current. I\'m talking 50 or 100 uA in my case.



I\'ve never used them above about 250V.

I tested one. It avalanches at 620 volts, no damage at low current.

The gate zeners at about +-40, again no damage.
Click to expand...

Yabbut, at 1400V at least one and probably two of them will be
continuously avalanching during normal operation, unless the drain
curves are squishy enough to take up the unit-to-unit spread in I_DSS.
(Maybe they are.) That would take a lot of testing to confirm as safe.

The avalanche voltage increases with temperature, so there\'d be no
automatic sharing of dissipation.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC / Hobbs ElectroOptics
Optics, Electro-optics, Photonics, Analog Electronics
Briarcliff Manor NY 10510

http://electrooptical.net
http://hobbs-eo.com
 
On Wed, 25 May 2022 09:40:48 -0400, Phil Hobbs
<pcdhSpamMeSenseless@electrooptical.net> wrote:

jlarkin@highlandsniptechnology.com wrote:
On Tue, 24 May 2022 16:47:30 -0400, Phil Hobbs
pcdhSpamMeSenseless@electrooptical.net> wrote:

jlarkin@highlandsniptechnology.com wrote:
On Tue, 24 May 2022 10:08:29 -0400, Phil Hobbs
pcdhSpamMeSenseless@electrooptical.net> wrote:

John Larkin wrote:
On Mon, 23 May 2022 11:06:00 -0700 (PDT), whit3rd <whit3rd@gmail.com
wrote:

On Monday, May 23, 2022 at 10:19:30 AM UTC-7, John Larkin wrote:

I want an LED blinker to show that a high-voltage power supply is
still dangerous. It should discharge the supply at some low current
and blink from, say, 1400 volts down to maybe 40.

So, a series string of fifteen current-limit diodes running a capacitor-
diac-LED blinkielight? S-272 seems suitable

https://www.mouser.com/datasheet/2/362/P22-23-CRD-1729293.pdf

The Supertex LND150 is a 500 volt depletion fet. A source resistor
will program the current, and they can apparently be strung in series
without problems.

Three of them would do. They are around 30 cents each.

A spin on Piglet\'s circuit might to the blinker.


I\'d probably be happier with a gate bias string to force the voltages to
equalize. Of course the resistors would have to be physically quite
large to take the voltage.



Somebody told me (was it you?) that they are OK in series. One or two
may avalanche somewhere above 500 volts and one will actually regulate
the current. I\'m talking 50 or 100 uA in my case.



I\'ve never used them above about 250V.

I tested one. It avalanches at 620 volts, no damage at low current.

The gate zeners at about +-40, again no damage.

Yabbut, at 1400V at least one and probably two of them will be
continuously avalanching during normal operation, unless the drain
curves are squishy enough to take up the unit-to-unit spread in I_DSS.
(Maybe they are.) That would take a lot of testing to confirm as safe.

The avalanche voltage increases with temperature, so there\'d be no
automatic sharing of dissipation.

Cheers

Phil Hobbs
Click to expand...

If two avalanche at about 650 volts each, that runs the third one as
the current limiter, at roughly 100 volts. Seems like it just works.

At 100 uA, the two that avalanche dissipate maybe 65 mW each. The
linear one is necessarily less.

Now I need a good blinker circuit.





--

Anybody can count to one.

- Robert Widlar
 
On Wed, 25 May 2022 11:21:22 +0100, piglet <erichpwagner@hotmail.com>
wrote:

On 23/05/2022 11:28 pm, John Larkin wrote:
On Mon, 23 May 2022 11:06:00 -0700 (PDT), whit3rd <whit3rd@gmail.com
wrote:

On Monday, May 23, 2022 at 10:19:30 AM UTC-7, John Larkin wrote:

I want an LED blinker to show that a high-voltage power supply is
still dangerous. It should discharge the supply at some low current
and blink from, say, 1400 volts down to maybe 40.

So, a series string of fifteen current-limit diodes running a capacitor-
diac-LED blinkielight? S-272 seems suitable

https://www.mouser.com/datasheet/2/362/P22-23-CRD-1729293.pdf

The Supertex LND150 is a 500 volt depletion fet. A source resistor
will program the current, and they can apparently be strung in series
without problems.

Three of them would do. They are around 30 cents each.

A spin on Piglet\'s circuit might to the blinker.


My preference would be to skip the depletion fets and just use dropper
resistors - cheaper, more reliable, and best of all the non constant
blink rate gives a good visual confirmation that the HV capacitor is
actually discharging.
Click to expand...

The discharge would be a slow exponential and the blink rate would
approach zero towards the end. Neither sounds good to me.

Blinking could be as simple as a diac or bjt used as negistor. Spice
won\'t model E-B breakdown and reverse beta so can\'t simulate negistor
operation but breadboarding would be very quick.

piglet
Click to expand...

--

Anybody can count to one.

- Robert Widlar
 
jlarkin@highlandsniptechnology.com wrote:
On Wed, 25 May 2022 09:40:48 -0400, Phil Hobbs
pcdhSpamMeSenseless@electrooptical.net> wrote:

jlarkin@highlandsniptechnology.com wrote:
On Tue, 24 May 2022 16:47:30 -0400, Phil Hobbs
pcdhSpamMeSenseless@electrooptical.net> wrote:

jlarkin@highlandsniptechnology.com wrote:
On Tue, 24 May 2022 10:08:29 -0400, Phil Hobbs
pcdhSpamMeSenseless@electrooptical.net> wrote:

John Larkin wrote:
On Mon, 23 May 2022 11:06:00 -0700 (PDT), whit3rd <whit3rd@gmail.com
wrote:

On Monday, May 23, 2022 at 10:19:30 AM UTC-7, John Larkin wrote:

I want an LED blinker to show that a high-voltage power supply is
still dangerous. It should discharge the supply at some low current
and blink from, say, 1400 volts down to maybe 40.

So, a series string of fifteen current-limit diodes running a capacitor-
diac-LED blinkielight? S-272 seems suitable

https://www.mouser.com/datasheet/2/362/P22-23-CRD-1729293.pdf

The Supertex LND150 is a 500 volt depletion fet. A source resistor
will program the current, and they can apparently be strung in series
without problems.

Three of them would do. They are around 30 cents each.

A spin on Piglet\'s circuit might to the blinker.


I\'d probably be happier with a gate bias string to force the voltages to
equalize. Of course the resistors would have to be physically quite
large to take the voltage.



Somebody told me (was it you?) that they are OK in series. One or two
may avalanche somewhere above 500 volts and one will actually regulate
the current. I\'m talking 50 or 100 uA in my case.



I\'ve never used them above about 250V.

I tested one. It avalanches at 620 volts, no damage at low current.

The gate zeners at about +-40, again no damage.

Yabbut, at 1400V at least one and probably two of them will be
continuously avalanching during normal operation, unless the drain
curves are squishy enough to take up the unit-to-unit spread in I_DSS.
(Maybe they are.) That would take a lot of testing to confirm as safe.

The avalanche voltage increases with temperature, so there\'d be no
automatic sharing of dissipation.


If two avalanche at about 650 volts each, that runs the third one as
the current limiter, at roughly 100 volts. Seems like it just works.

At 100 uA, the two that avalanche dissipate maybe 65 mW each. The
linear one is necessarily less.
Click to expand...

Provided that there\'s no long-term damage mechanism. It\'s going to be
in that state for a long long time.

> Now I need a good blinker circuit.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC / Hobbs ElectroOptics
Optics, Electro-optics, Photonics, Analog Electronics
Briarcliff Manor NY 10510

http://electrooptical.net
http://hobbs-eo.com
 
On Wednesday, 25 May 2022 at 06:41:00 UTC-7, Phil Hobbs wrote:
....
Somebody told me (was it you?) that they are OK in series. One or two
may avalanche somewhere above 500 volts and one will actually regulate
the current. I\'m talking 50 or 100 uA in my case.



I\'ve never used them above about 250V.
I tested one. It avalanches at 620 volts, no damage at low current.

The gate zeners at about +-40, again no damage.
Yabbut, at 1400V at least one and probably two of them will be
continuously avalanching during normal operation, unless the drain
curves are squishy enough to take up the unit-to-unit spread in I_DSS.
(Maybe they are.) That would take a lot of testing to confirm as safe.

The avalanche voltage increases with temperature, so there\'d be no
automatic sharing of dissipation.
....
Click to expand...

The Art of Electronics - The X Chapters. depicts that configuration and variants with a discussion in section 3.x.6.5

The simple configuration with just source resistors can have steps in the current as the voltage is varied as each device in turn avalanches.

kw
 
On Wed, 25 May 2022 09:22:48 -0700 (PDT), \"ke...@kjwdesigns.com\"
<keith@kjwdesigns.com> wrote:

On Wednesday, 25 May 2022 at 06:41:00 UTC-7, Phil Hobbs wrote:
...
Somebody told me (was it you?) that they are OK in series. One or two
may avalanche somewhere above 500 volts and one will actually regulate
the current. I\'m talking 50 or 100 uA in my case.



I\'ve never used them above about 250V.
I tested one. It avalanches at 620 volts, no damage at low current.

The gate zeners at about +-40, again no damage.
Yabbut, at 1400V at least one and probably two of them will be
continuously avalanching during normal operation, unless the drain
curves are squishy enough to take up the unit-to-unit spread in I_DSS.
(Maybe they are.) That would take a lot of testing to confirm as safe.

The avalanche voltage increases with temperature, so there\'d be no
automatic sharing of dissipation.
...

The Art of Electronics - The X Chapters. depicts that configuration and variants with a discussion in section 3.x.6.5

The simple configuration with just source resistors can have steps in the current as the voltage is varied as each device in turn avalanches.

kw
Click to expand...

The simple version, 3x57a with LND150s would be fine for a 1400 volt
LED blinker. And there are some diacs that would work for the periodic
cap discharge into the LED. 9 or so parts total.

--

If a man will begin with certainties, he shall end with doubts,
but if he will be content to begin with doubts he shall end in certainties.
Francis Bacon
 
John Larkin wrote:

The simple version, 3x57a with LND150s would be fine for a 1400 volt
LED blinker.
Click to expand...

Why wouldn\'t you just use a 1.7kV+ rated DMOS?

Best regards, Piotr
 
On Wed, 25 May 2022 20:31:12 +0200, Piotr Wyderski
<bombald@protonmail.com> wrote:

John Larkin wrote:

The simple version, 3x57a with LND150s would be fine for a 1400 volt
LED blinker.

Why wouldn\'t you just use a 1.7kV+ rated DMOS?

Best regards, Piotr
Click to expand...

Had a part in mind? If it\'s enhancement mode, it will need something
to bias it up.

We do have the LND150s in stock, SOT23 package.



--

Anybody can count to one.

- Robert Widlar
 
jlarkin@highlandsniptechnology.com wrote:

Had a part in mind? If it\'s enhancement mode, it will need something
to bias it up.

We do have the LND150s in stock, SOT23 package.
Click to expand...

Mouser has this IXTA1N170DHV. More expensive that LND150, but I am not
sure if a safety circuit is the best place for savings and IIRC your
business is of the high-added-value/low volume profile anyway. Just
saying, it\'s your decision.

I use the IXTA08N100D2HV for the same purpose (have a blinker too, and a
buzzer specifically tuned to emulate a fighter jet warning system:
https://youtu.be/fd50GtnXR7Y?t=1145, just couldn\'t resist), but my cap
voltage is under 900V, 650V typical.

Best regards, Piotr
 
On Thu, 26 May 2022 07:51:28 +0200, Piotr Wyderski
<bombald@protonmail.com> wrote:

jlarkin@highlandsniptechnology.com wrote:

Had a part in mind? If it\'s enhancement mode, it will need something
to bias it up.

We do have the LND150s in stock, SOT23 package.

Mouser has this IXTA1N170DHV. More expensive that LND150, but I am not
sure if a safety circuit is the best place for savings and IIRC your
business is of the high-added-value/low volume profile anyway. Just
saying, it\'s your decision.
Click to expand...

The 1400v supply is low current and doesn\'t store lethal energy, but
someone could damage parts by poking around while there\'s still
voltage. The linear discharge with blinker helps with that, and is
cool too. Sometimes we do things just because they are cool.

I use the IXTA08N100D2HV for the same purpose (have a blinker too, and a
buzzer specifically tuned to emulate a fighter jet warning system:
https://youtu.be/fd50GtnXR7Y?t=1145, just couldn\'t resist), but my cap
voltage is under 900V, 650V typical.

Best regards, Piotr
Click to expand...

That is an interesting part, but Mouser has four for $20 each. That\'s
another factor these days, availability and price.

We also use DN2530, a 300 v depl fet in SOT89. It good for discharging
bigger caps.

In a couple of boxes with gigantic caps, we used a depletion fet + LED
indicator and a discharge resistor + pushbutton inside, so our techs
are warned and didn\'t have to wait minutes to poke around.



--

Anybody can count to one.

- Robert Widlar
 
jlarkin@highlandsniptechnology.com wrote:

> The 1400v supply is low current and doesn\'t store lethal energy

Oh, in that case never mind. My comment was entirely safety-driven.

> Sometimes we do things just because they are cool.

So do I. :)

We also use DN2530, a 300 v depl fet in SOT89. It good for discharging
bigger caps.
Click to expand...

I use it as a resettable fuse. The short-circit current is limited and
the part just stops being hot when the short-circuit is removed. The
carbon-based PTCs wear off.

Best regards, Piotr
 
On Thu, 26 May 2022 18:06:43 +0200, Piotr Wyderski
<bombald@protonmail.com> wrote:

jlarkin@highlandsniptechnology.com wrote:

The 1400v supply is low current and doesn\'t store lethal energy

Oh, in that case never mind. My comment was entirely safety-driven.

Sometimes we do things just because they are cool.

So do I. :)

We also use DN2530, a 300 v depl fet in SOT89. It good for discharging
bigger caps.

I use it as a resettable fuse. The short-circit current is limited and
the part just stops being hot when the short-circuit is removed. The
carbon-based PTCs wear off.

Best regards, Piotr
Click to expand...

Ixys makes a self-protecting SSR, CPC1540. Off leakage is picoamps.

If you drive the input full-time, it becomes a fuse.



--

Anybody can count to one.

- Robert Widlar
 
You must log in or register to reply here.

Welcome to EDABoard.com

Sponsor

Back
Top