Reply by Phil Hobbs May 31, 20162016-05-31
>Testing the last poly fuse thing, I found that
somewhere in V-I space, the poly/zener has to
>operate as a (crapy) voltage regulator...
And happily waste whatever power you ask.
>It'd be nice if poly fuses had a sharper knee... >Oh.. polyzens... I haven't tried those.
Do they make 24- 30 Volt ones? They've been widening the range, but last time I checked they only went up to 16V or so (hence the auxiliary zener). Still, way over half of the zener dissipation is in the Polyzen, so it really helps sharpen up the knee. Cheers Phil Hobbs  
Reply by George Herold May 31, 20162016-05-31
On Tuesday, May 31, 2016 at 7:42:16 PM UTC-4, Phil Hobbs wrote:
> +1 for Polyzens--they're magic. A 16V PolyZen with a 10V zener in series with the ground lead will do a great job. A series Schottky from there to the load will help protect againdt supply reversal and input shorts. > > Cheers > > Phil Hobbs
Huh, OK split load. I'll have to think more... Testing the last poly fuse thing, I found that somewhere in V-I space, the poly/zener has to operate as a (crapy) voltage regulator... And happily waste whatever power you ask. It'd be nice if poly fuses had a sharper knee... Oh.. polyzens... I haven't tried those. Do they make 24- 30 Volt ones? George H.
Reply by krw May 31, 20162016-05-31
On Tue, 31 May 2016 17:07:18 -0700, John Larkin
<jjlarkin@highlandtechnology.com> wrote:

>On Tue, 31 May 2016 16:42:12 -0700 (PDT), Phil Hobbs ><pcdhobbs@gmail.com> wrote: > >>+1 for Polyzens--they're magic. A 16V PolyZen with a 10V zener in series with the ground lead will do a great job. A series Schottky from there to the load will help protect againdt supply reversal and input shorts. >> >>Cheers >> >>Phil Hobbs > >We use polyfuses with gigantic (SMB package) unipolar transzorb >zeners. Seems to work.
SMB isn't gigantic. I have a DO218 (or equivalent) TVS diode on almost every board (not on daughterboards that get power from the motherboard).
> >The surface-mount polyfuses are pretty bad, so we generally use the >radials.
Surface mount fuses aren't much better, though I use them for prototypes.
Reply by John Larkin May 31, 20162016-05-31
On Tue, 31 May 2016 16:42:12 -0700 (PDT), Phil Hobbs
<pcdhobbs@gmail.com> wrote:

>+1 for Polyzens--they're magic. A 16V PolyZen with a 10V zener in series with the ground lead will do a great job. A series Schottky from there to the load will help protect againdt supply reversal and input shorts. > >Cheers > >Phil Hobbs
We use polyfuses with gigantic (SMB package) unipolar transzorb zeners. Seems to work. The surface-mount polyfuses are pretty bad, so we generally use the radials. -- John Larkin Highland Technology, Inc picosecond timing precision measurement jlarkin att highlandtechnology dott com http://www.highlandtechnology.com
Reply by Phil Hobbs May 31, 20162016-05-31
+1 for Polyzens--they're magic. A 16V PolyZen with a 10V zener in series with the ground lead will do a great job. A series Schottky from there to the load will help protect againdt supply reversal and input shorts. 

Cheers

Phil Hobbs 
Reply by George Herold May 31, 20162016-05-31
On Friday, May 27, 2016 at 10:15:07 PM UTC-4, John Larkin wrote:
> On Fri, 27 May 2016 09:16:50 -0700 (PDT), George Herold > <gherold@teachspin.com> wrote: > > >A circuit will have 24V @ 1.5A (max) input from a wall wart. > >I'd like to design in reverse polarity and maybe over voltage > >input protection. (There are lots of threads about this.) > >I don't like the Zener/ polyfuse idea, cause there is going to > >be a lot of power in the Zener... I'm afraid it will smoke before the > >polyfuse lets go. So some FET circuit. > >Here's Jim T's offering. (from a past thread.) > >http://www.analog-innovations.com/SED/OverAndReverseVoltageProtection.pdf > > > >The gate source voltage on most of the P-fets I've looked at is > >+/- 20V. So the 24V is a bit of a concern. I can just put in some resistors > >to make a voltage divider.. so not a big deal. > > > >But do I need two P-fets? (Oh snap, never mind.. I had to draw the circuit.) > > > >Well I'll make it a question anyway... are then any P-fets that can handle more > >G-S voltage... (well more in absolute value.) > > > >TIA > >George H. > > Polyfuse+transzorb+cap works fine. A bunch of semiconductors and > passives (9 in Jim's case) will have hazards of their own. Like ESD > damage, for example.
Thanks, I worry that there won't be enough current to blow the poly fuse, and then the transorb/ Zener is sitting there dissipating ~10's of watts.. and won't last long. We're just going to hard wire the power supply in place, not as pretty but no fuss.
> > Most fets blow out around 75 volts Vgs. There are lots of fets that > have internal g-s ESD zeners, typically +-40 volts turf, and they can > be protected with just a series current-limit resistor.
I don't know FET's that well, but a quick sampling of spec sheets seems like they all list Vgs max as +/- 20V. I was wondering if this is mostly a "lazy" spec*. In that few people really care, and it would be too costly to spec it and measure. It seems like a 200V (Vsd) FET should have Vgs somewhere near 200 V. George H. Like the reverse bias voltage on LED's, all listed as 5V.
> > > -- > > John Larkin Highland Technology, Inc > > lunatic fringe electronics
Reply by May 28, 20162016-05-28
On Saturday, 28 May 2016 06:32:41 UTC+1, upsid...@downunder.com  wrote:
> On Fri, 27 May 2016 09:16:50 -0700 (PDT), George Herold > <gherold@teachspin.com> wrote: > > >A circuit will have 24V @ 1.5A (max) input from a wall wart. > >I'd like to design in reverse polarity and maybe over voltage > >input protection. (There are lots of threads about this.)
> If the source is floating, use a bridge rectifier, so the device works > regardless of polarity and even works on AC as a bonus.
you can often use that when things aren't floating too, If they're both grounded the ground connections just short out the 0v line diodes. NT
Reply by May 28, 20162016-05-28
On Fri, 27 May 2016 09:16:50 -0700 (PDT), George Herold
<gherold@teachspin.com> wrote:

>A circuit will have 24V @ 1.5A (max) input from a wall wart. >I'd like to design in reverse polarity and maybe over voltage >input protection. (There are lots of threads about this.) >I don't like the Zener/ polyfuse idea, cause there is going to >be a lot of power in the Zener... I'm afraid it will smoke before the >polyfuse lets go. So some FET circuit. >Here's Jim T's offering. (from a past thread.) >http://www.analog-innovations.com/SED/OverAndReverseVoltageProtection.pdf
24 V is quite high voltage so a series diode 0.6 V voltage drop doesn't too much harm. The diode solves the reverse polarity problem. Select a diode with a large reverse voltage rating, say 1 kV. If the source is floating, use a bridge rectifier, so the device works regardless of polarity and even works on AC as a bonus. For large positive peaks, use some series inductance to limit the peak current, helping any parallel protection component to do their job. However, if you expect constant over voltages, then some fusing component in series would be required.
Reply by Chris Jones May 28, 20162016-05-28
On 28/05/2016 02:16, George Herold wrote:
> A circuit will have 24V @ 1.5A (max) input from a wall wart. > I'd like to design in reverse polarity and maybe over voltage > input protection. (There are lots of threads about this.) > I don't like the Zener/ polyfuse idea, cause there is going to > be a lot of power in the Zener... I'm afraid it will smoke before the > polyfuse lets go.
It might be worth considering the Polyzen and any equivalents. It is a series polyfuse followed by a shunt Zener, the trick being that the Zener is thermally coupled to the polyfuse so that when the Zener starts cooking, it heats the polyfuse which makes it go high-impedance even if it was not conducting anywhere near its trip current. In that way the Zener is thermally protected. Availability might be a problem.
Reply by John Larkin May 27, 20162016-05-27
On Fri, 27 May 2016 09:16:50 -0700 (PDT), George Herold
<gherold@teachspin.com> wrote:

>A circuit will have 24V @ 1.5A (max) input from a wall wart. >I'd like to design in reverse polarity and maybe over voltage >input protection. (There are lots of threads about this.) >I don't like the Zener/ polyfuse idea, cause there is going to >be a lot of power in the Zener... I'm afraid it will smoke before the >polyfuse lets go. So some FET circuit. >Here's Jim T's offering. (from a past thread.) >http://www.analog-innovations.com/SED/OverAndReverseVoltageProtection.pdf > >The gate source voltage on most of the P-fets I've looked at is >+/- 20V. So the 24V is a bit of a concern. I can just put in some resistors >to make a voltage divider.. so not a big deal. > >But do I need two P-fets? (Oh snap, never mind.. I had to draw the circuit.) > >Well I'll make it a question anyway... are then any P-fets that can handle more >G-S voltage... (well more in absolute value.) > >TIA >George H.
Polyfuse+transzorb+cap works fine. A bunch of semiconductors and passives (9 in Jim's case) will have hazards of their own. Like ESD damage, for example. Most fets blow out around 75 volts Vgs. There are lots of fets that have internal g-s ESD zeners, typically +-40 volts turf, and they can be protected with just a series current-limit resistor. -- John Larkin Highland Technology, Inc lunatic fringe electronics