Re: A Cure for Hot Snubbers (fwd)

[<sroys@xxxxxxxxxxxxxxxxx>]

---------- Forwarded message ----------
Date: Mon, 18 Oct 2004 17:57:33 +1300
From: Malcolm Watts <m.j.watts@xxxxxxxxxxxx>
To: High Voltage list <hvlist@xxxxxxxxxx>
Subject: Re: A Cure for Hot Snubbers (fwd)

Hi Matthew,

On 17 Oct 2004, at 22:45, High Voltage list wrote:

> Original poster: <sroys@xxxxxxxxxxxxxxxxx>
> 
> 
> 
> ---------- Forwarded message ----------
> Date: Fri, 15 Oct 2004 07:23:58 +0930
> From: Matthew Smith <matt@xxxxxxxxxx>
> To: High Voltage list <hvlist@xxxxxxxxxx>
> Subject: A Cure for Hot Snubbers
> 
> Hi All
> 
> I'm working on some very low powered (NiMH battery supply) high
> voltage circuits.
> 
> One thing that concerns me is snubbers getting hot.  I'm using an
> integrated switcher which current limits itself, so at least I've got
> rid of the heat (wasted energy) there.
> 
> Now, am I getting this right?
> * The snubber exists to prevent the switching device from getting
> cooked.

Any core energy not delivered to the load ends up being dissipated in 
the resistor in the snubber network. The component values are chaosen 
so that the capacitance initially stores whatever energy is present 
in the core when the transistor switches off. It must be sized so 
that Vsw = SQRT(2E/C)  where Vsw is the maximum voltage you want to 
be impressed across the now-off transistor (i.e. preferably Vceo - 
10%), C is the snubber capacitor and E is the residual core energy; 
and the RC time constant ensures that energy is dissipated in 
resistor before the transistor is switched back on.

 * All energy that isn't transferred to the secondary goes
> through the snubber. * All I really want to cut is voltages greater
> than the switch can handle.
> 
> So...
> 
> If I use a fast diode in opposition to another (Zener?) diode as I
> have seen in some circuits that will only pass voltages over a little
> below the switch's maximum voltage, there will be considerably less
> energy wasted.

Residual core energy won't change, just the way you get rid of it. 
Best is to have the load absorb the lot if things can be arranged 
that way. The energy is strongly a function of the leakage inductance 
between the primary and secondary windings assuming there is a load 
present and this can be minimized by having the windings in as close 
proximity to each other as possible - i.e. wound on top of one 
another, not on opposite core legs.

Malcolm

> 
> Is this correct or am I barking up the wrong tree?
> 
> Cheers
> 
> M
> 
> -- 
> Matthew Smith
> Kadina Business Consultancy
> South Australia
> 
> Outside temperature:  18.2 degrees Celcius.
> Wind due WNW,  6.1km/hr gusting to  8.0km/hr
> Temperatures today (avg/max/min):   7.8/ 20.5/  5.2 degrees Celcius.
> Max wind speed today  8.0km/hr at 07:05 Precipitation today: 0.000mm
> 
> 
>