The Cyber-Spy.Com Usenet Archive Feeds Directly
From The Open And Publicly Available Newsgroup
This Group And Thousands Of Others Are Available
On Most IS NNTP News Servers On Port 119.
Cyber-Spy.Com Is NOT Responsible For Any Topic,
Opinions Or Content Posted To This Or Any Other
Newsgroup. This Web Archive Of The Newsgroup And
Posts Are For Informational Purposes Only.
From: "Frontline Electronics"
Subject: Re: Driving High current with PWM
X-Newsreader: Microsoft Outlook Express 4.72.3110.1
Date: Sat, 12 Oct 2002 01:34:50 GMT
NNTP-Posting-Date: Sat, 12 Oct 2002 01:34:50 GMT
Organization: AT&T Worldnet
John Popelish wrote in message <3DA2148F.B25BFB85@rica.net>...
>Richard Soennichsen wrote:
>> I need some help with the analog portion of a simple design. I need to
>> drive high current loads (~5A) with a micro driving PWM outputs. The
>> digital side is a snap for me but the analog, which seems like it should
>> simple is giving me fits.
>> The micro drives four outputs of PWM. (0-5V transitions) I need
>> to switch the HIGH side of the loads (12V lamps). I have experimented
>> PNP power transistors which functionally work but get VERY hot, hot
>> to de-solder themselves. Obviously the design is not sound. I am
>> the base of the PNP power with an NPN BJT. The base of the PNP power BJT
>> pulled up to 12V through a 5K resistor, this keeps it off unless the NPN
>> is on.
>> Also, it sure would be nice to use NPN power BJTS as then I could
>> them to the chassis as a heatsink but unfortunately the tabs are at 12v
>> well, the chassis is not.
>To get low on state voltage, you need to have plenty of gain in the
>PNP switch, which implies that it is rated for more than 5 amps. A 10
>to 15 amp rated transistor should do. Here are some rough ideas for
>The base drive should be a pretty fair fraction of the output current,
>say, 1/20 ( 5 amps /20 = 250 ma.) (If you have a high gain PNP/ you
>may get by with only 1/50 of output current for the base.) And the
>turn off current should be a fair fraction of that, say 1/5 or 50 ma.
>This means that the base to emitter shunt resistor should pass about
>50 ma when the base is forward biased at about .7 volts, so the base
>to emitter resistor would be .7/.05=15 ohms. Using a higher value of
>base resistor to save drive current will slow the turn off and produce
>more heat in the PNP during this transition, so a tradeoff is
>possible.) So the NPN drive transistor has to deliver both the base
>drive and the turn off current or about 300 ma. This implies that
>about .3 amps * 12 volts = 3.6 watts of heat is going somewhere when
>the NPN driver is on. If the NPN regulates the 300 ma, then it will
>have ot dump this power. If the NPN saturates, then a series resistor
>of about 12/.3=39 ohms must handle that power.
>All this shows why people like mosfet power transistors as switches,
>since they require no DC to keep them on, but only need current to
>turn them on or off. If you used a P channel mosfet as the output
>switch, you will need a low power 12 volt driving signal for its
>gate. A CD4049 or CD4050 hex inverter or buffer with all stages
>paralleled can probably provide the current needed to swing the gate
>capacitance. An open collector NPN transistor with a pull up resistor
>can convert your micro output to a 12 volt swing to drive the CMOS
Yes, and nice explaination. We use Mosfets to drive a 0.45 0hm load
to about 25W @ less than 4A and after Hrs of running the fets are
warm with no heatsink and they are smaller than your fingernail!
PWM is cool.
Go Back To The Cyber-Spy.Com
Usenet Web Archive Index Of
The sci.electronics.design Newsgroup