A small "treat" for hobbyists

[ivconic]

[ivconic]

I tested both stages in practice separately.
I had to additionally correct and adjust the soft latching, so that it would work flawlessly.
But I haven't tried both assemblies together.
Certainly that circuit will work flawlessly, I'm sure of it.
However, one thing catches the eye; "SW" switch is redundant.
Until it is checked whether that circuit will "leak" current if it is constantly connected to the battery.
Since it is tested separately; in the nature of things it didn't occur to me to check for "leakage" of current.
Because the "consumption" of that stage is totally insignificant when the circuit is correctly connected and when
the device on the other side is active.
As for the soft latching circuit; I picked up a simpler version from the internet, there is also a Youtube video
where a character very enthusiastically explains the circuit... which in reality does not work well.
It works quite unstable and random, what he presented in the video.
But with minimal adjustments it worked fantastically well.
So this is a schematic with two stages that have been separately tested in practice and work flawlessly.

[ivconic]

I don't know what your experience is... but my experiences with "buyers" and "users" of devices have been devastating.
If the detector or similar device runs on a 9v battery or a smaller 12v battery...or even some Lipo combination...
sooner or later there will be a "master" who will connect it to an external "lots of amps" battery
and sooner or later they will mistakenly connect the wrong battery polarity (by the rule all such self-taught masters sooner or later burn their detector that way)
and thus produce enormous amounts of white smoke!
So I was forced to think of ways and methods against such "experts" who have infinite courage and give themselves too much freedom in their
electronic expertise!
Those struggles gave birth to this assembly.

[ivconic]

[Skippy]

The 'reverse polarity' part WILL drain the battery. Your CAD drawing does NOT show the internal body diode of the Mosfet, which is an essential part of the operation. That diode always conducts, regardless of whether the fet is turned on, so ther is always 100k in series with a 9V1 zener placed on the 12V supply. This calculates as about 30 microamps drain. The zener provides 9V between gate and source of the fet, turning it on fully; it is in parallel with the body diode, so it will largely 'short it out'.
Obviously, the 100k resistor could be larger, but remember the zener will have a lower voltage drop as the current reduces. It's probably only dropping 7V when run at 30 microamps.

[ivconic]

Ha!
You noticed that well!



And I totally overlooked that!



Yes, the "ExpressSch" software is to blame because it has simple symbols and
I usually first notice the problem visually... no diode on the symbol = no problem!



30 microamps is tolerable... solves the problem.
But now I want better than that, I would prefer 0 microamps.
Without the "SW" switch, of course...
I'll continue with the "shoveling"!

[ivconic]

The 'reverse polarity' part WILL drain the battery. Your CAD drawing does NOT show the internal body diode of the Mosfet, which is an essential part of the operation. That diode always conducts, regardless of whether the fet is turned on, so ther is always 100k in series with a 9V1 zener placed on the 12V supply. This calculates as about 30 microamps drain. The zener provides 9V between gate and source of the fet, turning it on fully; it is in parallel with the body diode, so it will largely 'short it out'.
Obviously, the 100k resistor could be larger, but remember the zener will have a lower voltage drop as the current reduces. It's probably only dropping 7V when run at 30 microamps.

Now when I have diode present on mosfet symbol... ARE YOU SURE!?
The 'reverse polarity' part SHALL NOT drain the battery BUT Q2 will!



Just for the moment you got me pretty much confused!
But THANKS A LOT for that!
Because anyway I still have a problem without "SW" switch!
...
Wait a bit!
Now I'm definitely confused to no end!
Q2 does not drain the current at all!
In practice (on the video) it was verified.

[Skippy]

30 microamps would not be OK for a 9 volt PP3 powered circuit, like a pinpoint probe. I don't know how high the 100k resistor could go ... there is gate capacitance to charge up / discharge, so changing it to 10 M Ohm may be too much.
I have seen versions of this circuit ( using a bipolar transistor with low V_{ce sat} and a Schottky diode ) , and the transistor was switched on by part of the powered circuit. When the electronics is 'woken up' by a button-push, a microcontroller turned on the bjt to short out the diode. I recall the bjt was specifically chosen for low V_ce , a Ferranti / Zetex part. Perhaps you can do similar with this mosfet version - turn it on when the 'soft switch' is on.

[Skippy]

Regarding your last post, crappy fet symbols etc. You now have n-channel depletion mode fet symbols. You need p-channel enhancement type.

[ivconic]

Regarding your last post, crappy fet symbols etc. You now have n-channel depletion mode fet symbols. You need p-channel enhancement type.

Yes again you are right!
I will have to rearrange the symbols in ExpressSch software, make my own macros and stay away from using the symbols from the software!
Damn!

[ivconic]

Now this is correct:

[ivconic]

Same thing with N-channel attempt... no good!

[ivconic]

30 microamps would not be OK for a 9 volt PP3 powered circuit, like a pinpoint probe. I don't know how high the 100k resistor could go ... there is gate capacitance to charge up / discharge, so changing it to 10 M Ohm may be too much.
I have seen versions of this circuit ( using a bipolar transistor with low V_{ce sat} and a Schottky diode ) , and the transistor was switched on by part of the powered circuit. When the electronics is 'woken up' by a button-push, a microcontroller turned on the bjt to short out the diode. I recall the bjt was specifically chosen for low V_ce , a Ferranti / Zetex part. Perhaps you can do similar with this mosfet version - turn it on when the 'soft switch' is on.

Yes I agree.
I don't like "half solutions" at all.
This must be done better.
... keep shoveling deeper...

[ivconic]

https://www.monolithicpower.com/desi...circuit-part-i

[ivconic]

Skippy you are good in math, what would be the voltage drop in this case?
I bet is insignificant for the most of the applications we use here.
In case I am right; this is simpler solution: