[time-nuts] HP5328A LEDS driver transistor

[David]

On Mon, 23 Jan 2012 17:27:32 -0500, "Charles P. Steinmetz"
<charles_steinmetz at lavabit.com> wrote:

>Stan wrote:
>
>>The 5328A manual I have shows a 3.5 V supply (looks like it is used just
>>for the display).  But with a bit lower Vce sat drops, I'd guess you 
>>get around
>>1 V across the 10 ohm segment resistors.  Or only about 900 mA when all
>>9 segments are on (but at a low duty cycle).  Did HP also use a 5 V version
>>of the driver in some 5328A counters?
>
>Interesting.  I have 3 different sets of schematics for the 
>5328A.  The most legible of these (thus the one I consult first) 
>shows a 5 V supply for the display.  The other two show 3.5 V, as 
>yours does.  I cannot recall ever measuring the display supply 
>voltage in a 5328.
>
>Even with 900 mA per digit, I'd want a transistor with >/=2 A maximum 
>collector current for the digit drivers -- so I still think the 
>ZTX949 may be the best readily-available choice.

I am surprised the LED current is so high.  I always found diminishing
returns above 20 mA but some LED types are better than others for
multiplexed applications where the peak to average current is high.  8
x 20 mA also conveniently stays below the bond wire limit if you drive
7 segments plus the decimal point directly with a single IC without
multiplexing.

I would use a 2 amp or greater transistor just to keep the gain high
although I know those Zetex transistors much better about that.  I
think Fairchild has a similar line of high beta, high current, low
dissipation transistors now as well.