Crowbar? (was: Re: Testing Power Supplies!)

Tony Duell ard at p850ug1.demon.co.uk
Mon Nov 15 22:56:44 CST 2004


> 
> 
> Tony Duell wrote of crowbar protection circuits thus:
> 
> > I'd rather not trust a protection device if I don't have to. Sure I like 
> > crowbar circuits, and they have saved my chips a few times, but I don't 
> > like to depend on them.
> 
> Ah, twice in one week I get to expose my wanton ignorance.  :)  What 
> typically makes up a crowbar circuit? How are they triggered? If I'm 
> understanding this correctly, a crowbar quickly kills the PSU and then 
> drains power from the system rapidly and nondestructively... but how it 
> accomplishes this, or knows it _should_, I don't yet understand.
> 
> O'course, one could accomplish this with an actual crowbar applied 
> across the PSU output... but I suspect this would not be described as 
> "nondestructive".  :)

CLose :-). The idea is to be non-destructive to the rest of the machine, 
even if it ends up destroying parts of the defective PSU. You want to 
short-circuit the PSU outputs (normally only the +5V output) if it goes 
overvoltage. Typically the circuit is something like : 


  -------------+-----------------------+--------------------o  +5V
               |                       |
             -----                     |
  5V1 zener    ^  |                    |
             / | \                     |
               |                       |
               |                     \ | /
               +--------------------\  V      SCR
               |                    --------      
               |                       |
               /                       |
               \                       |
   100 Ohm     /                       |
               \                       |
               |                       |
 --------------+-----------------------+-------------------o  oV

If the 5V rail goes overvoltage, the zener breaks down and turns on the 
SCR which short-circuits the 5V rail (and carries on shorting it until 
the PSU output current drops to close to zero).

-tony



More information about the cctalk mailing list