Searching \ for '[EE]: Using ULN2803' in subject line. ()
Make payments with PayPal - it's fast, free and secure! Help us get a faster server
FAQ page: www.piclist.com/techref/index.htm?key=using+uln2803
Search entire site for: 'Using ULN2803'.

Exact match. Not showing close matches.
PICList Thread
'[EE]: Using ULN2803'
2001\06\22@035507 by Wesley Moore

flavicon
picon face
I have a project in which I intend to use a ULN2803. I would like to
be able to vary the output voltage from the 2803 and have examined the
datasheet and searched everywhere for an example but can't seem to find
one. From the datasheet it appears that Pin 10 with the connections to
internal diodes may allow this but I'm not sure how exactly they
function. If someone could explain the operation of the device it would
be most appreciated.

Wesley

--
Wesley Moore                      http://www.1moore.org/
RMIT - BEng (Comp Sys Eng)/BApp.Sc. (Comp Sci)  3rd Year

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2001\06\22@053914 by Vasile Surducan

flavicon
face
uln2803 is a driver. If you need variable output voltage you need to
modify the input voltage ( pin10 ) Diodes are for clamping ( inductive
loads )
see a picture at www.geocities.com/vsurducan/pic/f877.html
Vasile

On Fri, 22 Jun 2001, Wesley Moore wrote:

{Quote hidden}

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2001\06\22@085703 by Roman Black

flavicon
face
Wesley Moore wrote:
>
> I have a project in which I intend to use a ULN2803. I would like to
> be able to vary the output voltage from the 2803 and have examined the
> datasheet and searched everywhere for an example but can't seem to find
> one. From the datasheet it appears that Pin 10 with the connections to
> internal diodes may allow this but I'm not sure how exactly they
> function. If someone could explain the operation of the device it would
> be most appreciated.


Hi Wesley, the ULN2803 does not make output voltages,
it is a bunch of darlington transistors in a chip
designed to be either on or off.

If you explain more about how many devices you
are driving, and what they are, and why you need to
vary the output voltage then many people can offer
suggestions.:o)

If you need the ULN2003 datasheet I can email it to
you.
-Roman

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2001\06\22@102752 by Dan Michaels

flavicon
face
Wesley Moore wrote:
>I have a project in which I intend to use a ULN2803. I would like to
>be able to vary the output voltage from the 2803 and have examined the
>datasheet and searched everywhere for an example but can't seem to find
>one. From the datasheet it appears that Pin 10 with the connections to
>internal diodes may allow this but I'm not sure how exactly they
>function. If someone could explain the operation of the device it would
>be most appreciated.


Wesley, the 2803 is simply a bunch of NPN [darlington] inverters,
that provide a current path to gnd for external loads. The voltage
to power the loads, +Vp, must come from outside the chip, and can
be up to 50v. This arrangement is nice because you can control a
large, arbitrary load/voltage from a fixed voltage [+5v, etc] logic
device.

Pin 10 is named COM, which is kind of confusing. For proper
connection, connect pin 10 to the external voltage, +Vp, that
drives the loads - this will ensure that all of the clamping
diodes inside the chip are biased properly, and are wired
correctly across the external loads for suppression of inductive
spikes/etc.

The "test" figures in my datasheet are confusing, but the last
figure shows pin 10 connected properly to +Vp.

best regards,
- dan michaels
http://www.oricomtech.com
==========================

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2001\06\23@090710 by Wesley Moore

flavicon
picon face
I am only driving LED's. I have a 22 x 10 matric of 3mm LED's. The reason I
want to vary the voltage is to provide display dimming in the dark since the
unit is destined for my car.

Wesley

On Fri, Jun 22, 2001 at 10:53:02PM +1000, Roman Black wrote:
{Quote hidden}

--
Wesley Moore                      http://www.1moore.org/
RMIT - BEng (Comp Sys Eng)/BApp.Sc. (Comp Sci)  3rd Year

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.


2001\06\23@103319 by Dan Michaels

flavicon
face
Wesley Moore wrote:
>I am only driving LED's. I have a 22 x 10 matric of 3mm LED's. The reason I
>want to vary the voltage is to provide display dimming in the dark since the
>unit is destined for my car.
>

The 2803 is basically an on-off switch. If you want to vary the
intensity of the LEDs, you can do this by using PWM with variable
pulsewidth as input to the 2803. No need to vary the load voltage.

best regards,
- dan michaels
http://www.oricomtech.com
=========================

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.


2001\06\23@112131 by Roman Black

flavicon
face
Hi Wesley. So you have a led matrix of 22 x 10 leds.
You are controlling it with two or more ULN2003's,
basically darlington NPN transistors.

If you want to "dim" the entire display for car use
I think you are right to look at doing this in hardware.
Yes you can use PWM etc in software but I don't think
that is always the best way. Sometimes hardware is
king.:o)

As a 22 x 10 matrix you must be using the ULN2003 as
the low side driver, what are you using as the high
side driver? Maybe you can modify your high side driver
to give variable current and use that to dim the leds??
Are you using a bank of PNP transistors to drive the
high side??

Maybe if you post a circuit there will be a few people
here willing to offer suggestions on how to give a
"dimming" effect for your led display. :o)
-Roman



Wesley Moore wrote:
{Quote hidden}

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.


2001\06\23@143759 by Olin Lathrop

face picon face
> I am only driving LED's. I have a 22 x 10 matric of 3mm LED's. The reason
I
> want to vary the voltage is to provide display dimming in the dark since
the
> unit is destined for my car.

You would be better off not varying the voltage, but using PWM instead.


********************************************************************
Olin Lathrop, embedded systems consultant in Littleton Massachusetts
(978) 742-9014, spam_OUTolinTakeThisOuTspamembedinc.com, http://www.embedinc.com

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.


2001\06\23@163526 by Dale Botkin

flavicon
face
Ooh!  Ooh!  Finally someone who could use one of my tricks...

I had a similar deal, wanted to be able to dim a whole bunch of LEDs in an
array driven by a PIC.  I used an ADC input (though you don't need this,
you could just have 2 or more levels of dimming) to determine a time
constant to leave the LEDs on.  Lower input voltage to the control pin,
shorter ON time, less brightness.  The LEDs are driven via row/column
transistors with no current limiting resistors (reduced parts count) and
just one cheap, low-power pot to control the brightness.  It's got low
parts count and very little code.  I've got the thing written up on my web
page, http://www.botkin.org/dale/.  Hope it helps.

Dale

On Sat, 23 Jun 2001, Wesley Moore wrote:

{Quote hidden}

--
A train stops at a train station.  A bus stops at a bus station.
On my desk I have a workstation...

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.


2001\06\23@175416 by Dan Michaels

flavicon
face
Dale.B wrote:
.....
The LEDs are driven via row/column
>transistors with no current limiting resistors (reduced parts count)
......

Dale, do you really think this is safe?

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.


2001\06\23@192525 by Dale Botkin

flavicon
face
On Sat, 23 Jun 2001, Dan Michaels wrote:

> Dale.B wrote:
> .....
>  The LEDs are driven via row/column
> >transistors with no current limiting resistors (reduced parts count)
> ......
>
> Dale, do you really think this is safe?

Yep, sure do.  Otherwise I wouldn't have designed it that way.  There
actually is some current limiting provided by the transistors, of course.
Since the LEDs are only on for a few milliseconds at most, The average
current is still well within a reasonable range.  The LED array scanning
is interrupt driven using TMR0, and the watchdog is set to prevent any
unfortunate results should the software glitch.  The maximum duty cycle of
the LEDs in my case is 1/33 (or roughly 3%).  So, yes, I think it's safe.

It was difficult to come up with another way to drive a large-ish array of
LEDs with acceptable brightness, and without a lot of power dissipation
somewhere...  the variable brightness was just a bonus.

Dale
--
A train stops at a train station.  A bus stops at a bus station.
On my desk I have a workstation...

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.


2001\06\24@131118 by Dan Michaels

flavicon
face
Dale.B wrote:
>On Sat, 23 Jun 2001, Dan Michaels wrote:
>
>> Dale.B wrote:
>> .....
>>  The LEDs are driven via row/column
>> >transistors with no current limiting resistors (reduced parts count)
>> ......
>>
>> Dale, do you really think this is safe?
>
>Yep, sure do.  Otherwise I wouldn't have designed it that way.  There
>actually is some current limiting provided by the transistors, of course.
>Since the LEDs are only on for a few milliseconds at most, The average
>current is still well within a reasonable range.  The LED array scanning
>is interrupt driven using TMR0, and the watchdog is set to prevent any
>unfortunate results should the software glitch.  The maximum duty cycle of
>the LEDs in my case is 1/33 (or roughly 3%).  So, yes, I think it's safe.
>

Yep, I figured this was the reasoning. Every thing is ok ...... as long
as .......... the duty cycle stays low/etc. However, if something
unforseen [misc hangup, murphy's law, mouse house, lost penny, what
ever] does occur, then ........ PSSSSSSST!
=========


>It was difficult to come up with another way to drive a large-ish array of
>LEDs with acceptable brightness, and without a lot of power dissipation
>somewhere...  the variable brightness was just a bonus.
>

Understood - still gives me the willies, as I like to build in some
kind of basic "guarantee", at the level below the uC. Some of us
like to live our lives a little further from the edge ;-).

best regards,
- dan michaels
http://www.oricomtech.com
=========================

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email .....listservKILLspamspam@spam@mitvma.mit.edu with SET PICList DIGEST in the body


2001\06\24@162455 by Bob Ammerman

picon face
> Understood - still gives me the willies, as I like to build in some
> kind of basic "guarantee", at the level below the uC. Some of us
> like to live our lives a little further from the edge ;-).
>
> best regards,
> - dan michaels
> http://www.oricomtech.com
> =========================

Hmm....

I often consider the uC the lower level that gives me the guarantee. Of
course, that is when the upper level is a box running Windoze! ;-)

Bob Ammerman
RAm Systems
(contract development of high performance, high function, low-level
software)

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email listservspamKILLspammitvma.mit.edu with SET PICList DIGEST in the body


2001\06\24@190135 by Shane.Bolton

picon face
The purpose of the diodes is so the ULN2803 can drive relays and small motors directly and still be
protected from the back EMF generated from the collapsing field.

From memory these devices are open-collector so you should be able to use a supply voltage up to the
maximum spec in the datasheet.

Shane


Wesley Moore wrote:
{Quote hidden}

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email .....listservKILLspamspam.....mitvma.mit.edu with SET PICList DIGEST in the body


2001\06\24@192706 by Dale Botkin

flavicon
face
On Sun, 24 Jun 2001, Dan Michaels wrote:

{Quote hidden}

Yeah, that's why the WDT and all...  sooner or later you have to decide
where to trust the hardware to do what it was designed to do.

> >It was difficult to come up with another way to drive a large-ish array of
> >LEDs with acceptable brightness, and without a lot of power dissipation
> >somewhere...  the variable brightness was just a bonus.
> >
>
> Understood - still gives me the willies, as I like to build in some
> kind of basic "guarantee", at the level below the uC. Some of us
> like to live our lives a little further from the edge ;-).

Sure thing...  if you've got any bright ideas I'm open to suggestions.  15
rows of LEDs, 16 LEDs per row, variable brightness, and they've got to be
fairly bright when cranked up.  Other than dissipating a boatload of power
at the transistors or somewhere else (oh, did I mention limited space and
a harsh environment?) I couldn't come up with an acceptable alternative.

Dale
--
A train stops at a train station.  A bus stops at a bus station.
On my desk I have a workstation...

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email EraseMElistservspam_OUTspamTakeThisOuTmitvma.mit.edu with SET PICList DIGEST in the body


2001\06\24@222209 by Dan Michaels

flavicon
face
Bob.A wrote:
>> Understood - still gives me the willies, as I like to build in some
>> kind of basic "guarantee", at the level below the uC. Some of us
>> like to live our lives a little further from the edge ;-).

>Hmm....
>
>I often consider the uC the lower level that gives me the guarantee. Of
>course, that is when the upper level is a box running Windoze! ;-)
>

resistor/etc

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email listservspamspam_OUTmitvma.mit.edu with SET PICList DIGEST in the body


2001\06\25@024539 by jeethur

flavicon
face
Wesley,

I guess a better idea would be to use a single BD679 NPN darlington with
the collector connected to the common cathode of all the LEDs and the
Emmiter grounded.I don't know the current rating of your LEDs
(typically around 25 MA each) @ 25 MA, your whole array would consume about
5.5 Amps. So I'd suggest using a good heat sink. Then, you can directly
drive
the base of the Transistor with a PWM from the PIC.

The Best solution would be to use a MOSFET like BUZ10 (I suspect BD679 would
die out too soon at 5.5 amps!).And moreover the internal resistance of
MOSFETs
is a lot lesser than Bipolars, Which will greatly increase the efficency
with
a matched decrease in the Thermal Dissipation.

My personal philosophy is to reduce the component count without complicating
the circuit. With the above given topology, you might even write software
to switch off the display after a brief span of inactivity or lack of user
input.
Mobile Phones had this feature from ages

Regards,

Jeethu Rao

{Original Message removed}

2001\06\25@104425 by Dan Michaels

flavicon
face
Dale Botkin wrote:
.........
>> Understood - still gives me the willies, as I like to build in some
>> kind of basic "guarantee", at the level below the uC. Some of us
>> like to live our lives a little further from the edge ;-).
>
>Sure thing...  if you've got any bright ideas I'm open to suggestions.  15
>rows of LEDs, 16 LEDs per row, variable brightness, and they've got to be
>fairly bright when cranked up.  Other than dissipating a boatload of power
>at the transistors or somewhere else (oh, did I mention limited space and
>a harsh environment?) I couldn't come up with an acceptable alternative.
>

Dale, just a couple of thoughts. I would hate to rely upon the uC
putting out a specific duty cycle as the primary means to keep a
system like this from smoking. Other possibilities [maybe you have
these implemented or they are not feasible, given your particular
reqs]:

- fuses, or better yet [I think] polyswitch resettable fuses.
- arranging the loads on the drivers [MOSFETs, ULN, ...] so that
 the driver could not run an excessive current through the LEDs;
 ie, configure LEDs in series columns and parallel rows - although
 this may overload the driver in normal ops.
- adding a small R [say 0.1 ohm] in series with the power source to
 sample load current, and feeding to an independent shutdown ckt
 [separate from the uC]; this should only add a few components and
 could save a lot.

best regards,
- dan michaels
http://www.oricomtech.com
========================

--
http://www.piclist.com hint: To leave the PICList
@spam@piclist-unsubscribe-requestKILLspamspammitvma.mit.edu


2001\06\25@121634 by Dale Botkin

flavicon
face
On Mon, 25 Jun 2001, Dan Michaels wrote:

{Quote hidden}

I don't think this is a possibility.  It's a matrix in which any LED can
be turned on or off.

> - adding a small R [say 0.1 ohm] in series with the power source to
>   sample load current, and feeding to an independent shutdown ckt
>   [separate from the uC]; this should only add a few components and
>   could save a lot.

I'll pass that suggestion along to the guy building the hardware and see
if he can come up with something acceptable.

The input provided by various people on this has been interesting.  The
problem I see with any of these approaches, though, is that you'd have to
watch both current and time.  The current *will* be high for very short
periods of time as the LEDs are pulsed.

Between the interrupt driven scanning and the WDT, we are pretty sure the
chance of an LED-on failure will be pretty low.  We might be able to come
up with some sort of one-shot arrangement that would add a little margin
of safety, but then you're increasing parts count and decreasing MTBF,
board size issues aside.

Dale
--
A train stops at a train station.  A bus stops at a bus station.
On my desk I have a workstation...

--
http://www.piclist.com hint: To leave the PICList
KILLspampiclist-unsubscribe-requestKILLspamspammitvma.mit.edu


2001\06\25@130607 by Dan Michaels

flavicon
face
Dale.B wrote:
......
>
>> - adding a small R [say 0.1 ohm] in series with the power source to
>>   sample load current, and feeding to an independent shutdown ckt
>>   [separate from the uC]; this should only add a few components and
>>   could save a lot.
>
>I'll pass that suggestion along to the guy building the hardware and see
>if he can come up with something acceptable.
>

Small R in power lead, pickoff to opamp/threshold detector w/output
that turns off power transistors - last can probably be done by having
an NPN inverter powering LEDs, and simply tying the fail-safe NPN
to its base.

                     LEDs
                      /
 LED                |/
drive----R---+------|  driver inverter
             |      |\
from       |/         V
over---R---|          |
current     |\        gnd
detector      V
             |
            gnd


>The input provided by various people on this has been interesting.  The
>problem I see with any of these approaches, though, is that you'd have to
>watch both current and time.  The current *will* be high for very short
>periods of time as the LEDs are pulsed.
>

But the "average" will be low, compared to failure situation
--> hello PWM-like RC filter.
===============

>Between the interrupt driven scanning and the WDT, we are pretty sure the
>chance of an LED-on failure will be pretty low.  We might be able to come
>up with some sort of one-shot arrangement that would add a little margin
>of safety, but then you're increasing parts count and decreasing MTBF,
>board size issues aside.
>

But a few components may add some real safety. Real life is always
a tradeoff. That's why they call us engineers.

best regards,
- dan michaels
http://www.oricomtech.com
===========================

--
http://www.piclist.com hint: To leave the PICList
RemoveMEpiclist-unsubscribe-requestTakeThisOuTspammitvma.mit.edu


2001\06\25@130955 by Peter L. Peres

picon face
>>Hmm....
>>
>>I often consider the uC the lower level that gives me the guarantee. Of
>>course, that is when the upper level is a box running Windoze! ;-)
>>
>
>resistor/etc

Go for a zero part design, charge double, and be famous.

(Hint: zero part designs require blindfolds and/or hypnotizers. Two thieve
suitmakers might do depending on the client)

Meanwhile I take Ockham's rule very seriously and remove parts until it
stops working, then put the last one back. ;-)

Peter

--
http://www.piclist.com hint: To leave the PICList
spamBeGonepiclist-unsubscribe-requestspamBeGonespammitvma.mit.edu


2001\06\25@135951 by Dan Michaels

flavicon
face
Peter Peres wrote:
........
>Meanwhile I take Ockham's rule very seriously and remove parts until it
>stops working, then put the last one back. ;-)


If you employ Maxwell's Demon, then you only need that last part
half the time.

--
http://www.piclist.com hint: To leave the PICList
TakeThisOuTpiclist-unsubscribe-requestEraseMEspamspam_OUTmitvma.mit.edu


2001\06\25@204128 by Dave Dilatush

picon face
Dale Botkin <RemoveMEdalespamTakeThisOuTBOTKIN.ORG> wrote...

{Quote hidden}

One possible problem you can have with this is that since you're
relying on the row/column drive transistors to limit the LED current,
you're going to be very dependent on device characteristics.
Transistor beta will vary greatly with ambient temperature, and from
lot to lot and manufacturer to manufacturer.  And while driver
transistor vendors almost always specify a minimum value of beta, I
don't think they often commit to a maximum value.  Some of these
devices have absolutely enormous current gains.

I've seen such arrangements work fine as prototypes, then fail in
production when a new manufacturing lot of transistors are used or
when purchasing starts buying a generic part from a different
manufacturer.

Your arrangement with the WDT will protect the circuit just fine in
the unlikely event the PIC falls off its tracks, wobbles away into the
weeds and stops scanning the matrix; but I really wouldn't be
comfortable with this no-limiting-resistors approach without having
something, somewhere to limit the steady-state current in case you run
into a batch of driver transistors that are overly enthusiastic.

The idea of the duty-cycle control of brightness and the pot connected
to an ADC input, is definitely neat though...

Dave

--
http://www.piclist.com hint: To leave the PICList
piclist-unsubscribe-requestEraseMEspam.....mitvma.mit.edu


2001\06\25@211602 by Dale Botkin

flavicon
face
On Tue, 26 Jun 2001, Dave Dilatush wrote:

> One possible problem you can have with this is that since you're
> relying on the row/column drive transistors to limit the LED current,
> you're going to be very dependent on device characteristics.
> Transistor beta will vary greatly with ambient temperature, and from
> lot to lot and manufacturer to manufacturer.  And while driver
> transistor vendors almost always specify a minimum value of beta, I
> don't think they often commit to a maximum value.  Some of these
> devices have absolutely enormous current gains.

More than 325?  That's what I figured as a worst case, and the result was
still well within limits for the LEDs.  All the manufactuers I checked
listed minimums and maximums for the 2N2222A at around 50 and 325.  I'd
think 325 would be a pretty ambitious number - but then I'm a digital kind
of guy, what do I know?  I could be wrong.

> Your arrangement with the WDT will protect the circuit just fine in
> the unlikely event the PIC falls off its tracks, wobbles away into the
> weeds and stops scanning the matrix; but I really wouldn't be
> comfortable with this no-limiting-resistors approach without having
> something, somewhere to limit the steady-state current in case you run
> into a batch of driver transistors that are overly enthusiastic.

I just came up with the idea and wrote the code.  It's someone else's
product that uses it; if he wants to design in the added stuff to deal
with that eventuality, I'm sure he will.

> The idea of the duty-cycle control of brightness and the pot connected
> to an ADC input, is definitely neat though...

Thanks...  that was really my whole point.  I was busy enough dodging
tear-gas grenades for not designing in a belt and two pair of suspenders I
was afraid no one had noticed.  I've *GOT* to start adding a page or two
of disclaimer to my posts one of these days.

Dale
--
A train stops at a train station.  A bus stops at a bus station.
On my desk I have a workstation...

--
http://www.piclist.com hint: To leave the PICList
EraseMEpiclist-unsubscribe-requestspammitvma.mit.edu


2001\06\25@222034 by Dave Dilatush

picon face
Dale Botkin <RemoveMEdaleEraseMEspamEraseMEBOTKIN.ORG> wrote...

>On Tue, 26 Jun 2001, Dave Dilatush wrote:
>
>...Some of these
>> devices have absolutely enormous current gains.
>
>More than 325?  That's what I figured as a worst case, and the result was
>still well within limits for the LEDs.  All the manufactuers I checked
>listed minimums and maximums for the 2N2222A at around 50 and 325.  I'd
>think 325 would be a pretty ambitious number - but then I'm a digital kind
>of guy, what do I know?  I could be wrong.

The highest gain could be just about anything, particularly at high
temperatures.  Usually you can rely pretty well on the minimum gain
specification; but if it were me, I'd be very cautious about counting
on any upper limit on gain, especially if that's the only thing that
makes the difference between a working circuit and a PC board with
black, smelly smudge marks on it.

There's gotta be an easy way to pacify this thing...

National Semiconductor's LP29xx/LM29xx family of LDO (Low Dropout)
voltage regulators can be used as current limiters, as they have an
internal current-limiting mechanism that does not rely on a
current-sampling resistor.  I use LP2951's in a lot of projects, as
they limit nicely at about 100 mA.

I'm wondering if you could use an LM2940; that would give you about
2.2 amps limiting over a wide temperature range.  If your circuit
draws less than the limit value, the LM2940 just sits there, dropping
very little voltage; if your circuit draws more, the LM2940 will lower
its output voltage to bring the current back into line.  If you need
other current limit values, either higher or lower, you would use a
different device from that family.

Maybe that would do the trick...

Dave

--
http://www.piclist.com hint: To leave the PICList
RemoveMEpiclist-unsubscribe-requestspam_OUTspamKILLspammitvma.mit.edu


2001\06\25@224525 by Dale Botkin

flavicon
face
{Quote hidden}

Which would not be anywhere near enough to drive a bunch of LEDs at fairly
bright levels, once they're multiplexed...

> I'm wondering if you could use an LM2940; that would give you about
> 2.2 amps limiting over a wide temperature range.  If your circuit
> draws less than the limit value, the LM2940 just sits there, dropping
> very little voltage; if your circuit draws more, the LM2940 will lower
> its output voltage to bring the current back into line.  If you need
> other current limit values, either higher or lower, you would use a
> different device from that family.

But what point would there be to limiting the current to 2A or more?  I know
that would produce little black craters where LEDs used to be.  The
objections I'm hearing raised are of the "what happens when your crystal
stops oscillating and one column of LEDs gets stuck on?" variety.

Dale

--
http://www.piclist.com hint: To leave the PICList
EraseMEpiclist-unsubscribe-requestspamspamspamBeGonemitvma.mit.edu


2001\06\25@225818 by Brent Brown

picon face
Sorry I haven't been following this whole thread but I have an idea
to share.

I once designed a panel that had about 80 mux'd LED's. I used a
4017 to step through the columns of LED's in sequence. The first
output was unused so that at reset no LED's would be selected. A
monstable was connected to the 4017 clock line and it reset the
4017 if there were no pulses present from the micro, thereby
protecting the LED's if the micro glitched or crashed.

Brent Brown
Electronic Design Solutions
16 English Street
Hamilton, New Zealand
Ph/fax: +64 7 849 0069
Mobile/text: 025 334 069
eMail:  RemoveMEbrent.brownKILLspamspamclear.net.nz

--
http://www.piclist.com hint: To leave the PICList
piclist-unsubscribe-requestSTOPspamspamspam_OUTmitvma.mit.edu


2001\06\25@232411 by Ed Boggs

picon face
  What I would do is use the MAX-7219 LED driver chip.  It can drive 64 LED's
with only one external component, a current set resistor. Check out the data
sheet.
I've used it for years, works great. BTW, did you notice Maxim and Dallas
Semi are on the same web site? Assume Maxim bought Dallas. Have not
researched.

            Ed Boggs

--
http://www.piclist.com hint: To leave the PICList
spamBeGonepiclist-unsubscribe-requestSTOPspamspamEraseMEmitvma.mit.edu


2001\06\25@232428 by Dan Michaels

flavicon
face
Dale.B wrote:
.......
>But what point would there be to limiting the current to 2A or more?  I know
>that would produce little black craters where LEDs used to be.  The
>objections I'm hearing raised are of the "what happens when your crystal
>stops oscillating and one column of LEDs gets stuck on?" variety.
>

Dale, you might also take a look at the polyswitch I mentioned
before. I don't know your ckt layout, but you could use 1 or many,
depending.

They have continuous ratings from 0.1A up to about 14A, with trip
currents at roughly 2X cont. The non-tripped resistances of these
devices ranges from 4.5ohm for 0.1A device to 0.002 ohm on the 14A.
They have extremely non-linear S-shaped characteristics, and when
tripped, R goes up by orders of magnitude.

- dan michaels
http://www.oricomtech.com
=======================

--
http://www.piclist.com hint: To leave the PICList
KILLspampiclist-unsubscribe-requestspamBeGonespammitvma.mit.edu


2001\06\25@232933 by Dan Michaels

flavicon
face
Dale.B wrote:
.......
>
>More than 325?  That's what I figured as a worst case, and the result was
>still well within limits for the LEDs.  All the manufactuers I checked
>listed minimums and maximums for the 2N2222A at around 50 and 325.  I'd
>think 325 would be a pretty ambitious number - but then I'm a digital kind
>of guy, what do I know?  I could be wrong.
>

I think those numbers are for relatively low Icoll, and the gains
go way down as Icoll goes up.

--
http://www.piclist.com hint: To leave the PICList
EraseMEpiclist-unsubscribe-requestspamEraseMEmitvma.mit.edu


2001\06\25@233338 by Dan Michaels

flavicon
face
>
>Dale, you might also take a look at the polyswitch I mentioned
>before.


oops, was supposed to say .... polyswitch resettable fuses ....

--
http://www.piclist.com hint: To leave the PICList
@spam@piclist-unsubscribe-request@spam@spamspam_OUTmitvma.mit.edu


2001\06\26@033646 by Roman Black

flavicon
face
Brent Brown wrote:
>
> Sorry I haven't been following this whole thread but I have an idea
> to share.
>
> I once designed a panel that had about 80 mux'd LED's. I used a
> 4017 to step through the columns of LED's in sequence. The first
> output was unused so that at reset no LED's would be selected. A
> monstable was connected to the 4017 clock line and it reset the
> 4017 if there were no pulses present from the micro, thereby
> protecting the LED's if the micro glitched or crashed.

Nice hardware solution! Using RC monostable
and the reset pin of the 4017 gates its output
so this is micro free and even flip-flop free.
Should be relatively spike proof and cosmic ray
proof.

It also gives one-wire operation from the micro
as it can simply miss a pulse to perform a reset
of the 4017, saves a PIC pin too!
:o)
-Roman

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2001\06\26@041330 by Alan B. Pearce

face picon face
>BTW, did you notice Maxim and Dallas
>Semi are on the same web site? Assume Maxim bought Dallas.
>Have not researched.

I had noticed. It also means that you can get samples of Dallas devices on
the same policy that Maxim have used for years. Also if you try to access
the old Dallas website it takes you to the Maxim one, so I think your
assumption is correct.

I have been following this thread and will relate an experience I had. in
the early 1980's I got involved with developing a scanned 7 segment display
as part of a project. It was at the time that "high efficiency" LEDs had
just been introduced. In trying to multiplex the display I found that I was
using smaller and smaller resistors to try and get the necessary current
through the segments. Eventually I ended up with n resistors in the circuit.
It seemed that the semiconductor processing to make the LED's more efficient
light producers had the effect of increasing the internal resistance of the
chip at high currents. I do not know if the project continued with this as I
left before it was finished, but I would hate to have seen the effect if the
micro that did the scanning failed.

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2001\06\26@062942 by Dave Dilatush

picon face
Dale Botkin <spamBeGonedalespamKILLspamBOTKIN.ORG> wrote...

>But what point would there be to limiting the current to 2A or more?  I know
>that would produce little black craters where LEDs used to be.  The
>objections I'm hearing raised are of the "what happens when your crystal
>stops oscillating and one column of LEDs gets stuck on?" variety.

And for those objections, Brent Brown's solution is a very elegant
one.

What I've been talking about is a separate problem: in normal
operation, you are relying on maximum transistor beta to limit current
and protect your circuit.  In my experience this is not reliable, and
unless you provide some sort of current limiting- whether on an
individual basis for each LED, or by row, or by column or for the
array as a whole- you are inviting trouble down the line when your
design goes into production.

Dave

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2001\06\26@104449 by miked
flavicon
face
Yes, done some months ago.
> BTW, did you notice Maxim and Dallas
> Semi are on the same web site? Assume Maxim bought Dallas. Have not researched.
>
>              Ed Boggs

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


More... (looser matching)
- Last day of these posts
- In 2001 , 2002 only
- Today
- New search...