Re: ULN2803 current limit


David Harmon wrote:
> ULN2803A: The datasheet says the 500 mA limit applies to "one device
> in the package". Does that really mean that the whole package can
> pass 4 Amps at once using 8 pins? Or do you only get the full 500mA
> while using only one device at a time.

Hi, Dave. Great question. Here are two data sheets for the 2803:

http://www.onsemi.com/pub/Collateral/ULN2803-D.PDF
http://www.allegromicro.com/datafile/2801.pdf

The first data sheet is remarkably opaque on this really basic
question. I would guess that's because it's kind of a "gumball" part
-- it's been around for decades.

Both data sheets tell you that each output is rated for 500mA.
However, the Vce of the outut darlington is typically between 1.0V and
1.5V for high output currents. Assuming worst case, eight outputs of
500mA with 1.5V across each output, you've got 6 watts being dissipated
by the IC package. Not happening in this world. The IC will glow red
for sure.

The second data sheet is more descriptive. It has a graph that says if
you've got a 100% duty cycle (the output is always on), you can't even
run two outputs at full current. This is because of power dissipation
limitations. The maximum output per pin goes from 500 mA (one output
on at one time) to about 470mA with two on, down to about 160mA with
all eight outputs on at 100% duty cycle. Look at the graph on page 6.

Actually, since these are maximums, you should derate the outputs on
the data sheet accordingly. I would think good engineering practice
would suggest no more than about 1 watt power dissipation maximum for
the IC package. That would mean 500mA for 1 output, 330mA for two
outputs, 220mA for 3 outputs, 167mA for 4, and so on. But how
conservative you want to be is up to you. Don't exceed the maximums
shown on the graph, though.

If you need 500mA from more than one output, use TO-220 NPN
darlingtons, and discrete resistors to give you 1mA of drive current at
the base of the transistor. Almost any CMOS or TTL output can drive
1mA. And the 500mA collector current is small enough so you won't need
heat sinks for the transistors.

Good luck
Chris

.

Relevant Pages

  • Re: N-channel MOSFET Safe Operating Area anomaly?
    ... It is not the expected power dissipation from the package in standard operating conditions as that depends on too many factors for a simple number to describe. ... Now when we want to go calculate the total dissipation in our case we can do so because the thermal resistances will add as there will not be any co-dependence between the different thermal cases. ... Another way to think about it is that in the ideal case, that is complete power absorbtion and removal from the device package, then the device package can dissipate it's maximum power dissipation withthe junction temperature being it's maximum. ...
    (sci.electronics.design)
  • RE: SUMIF across Columns
    ... Assumes headings in Data sheet are in row 1, starting column E and "var" is ... On Graph: result for NW is in B2 ... evaluate the division code there and sum each var going down the ...
    (microsoft.public.excel.worksheet.functions)
  • Re: Plot Widget for CPAN
    ... dictionary and I see that "graph" is a much better word. ... general name which might be well used by a more general package. ... interfaces into plot packages. ...
    (comp.lang.perl.tk)
  • Re: Power Output vs. Power Dissipation of ULN2803A
    ... I found a data sheet that lists the IC's ... > power dissipation as 2.25W for the whole package. ... > when each darlington pair has a load current rating of 500 mA ...
    (sci.electronics.design)
  • Re: format title column in data sheet of charts
    ... I am also displaying the data table for Graph. ... The query I used got the result. ... Thought formaing that column as text in the data sheet ...
    (microsoft.public.access.reports)