Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

mc33063 current limiting

295 views
Skip to first unread message

Michael Robinson

unread,
Jan 5, 2010, 8:54:40 PM1/5/10
to
How does the current limiting work in the MC33063. The datasheet doesn't
explain the actual mechanism.
http://www.onsemi.com/pub_link/Collateral/MC34063A-D.PDF
Does it discharge the timing cap when it detects overcurrent, which would
restart the oscillator, or does it let the oscillator keep running but at a
lower duty cycle?


whit3rd

unread,
Jan 6, 2010, 1:56:48 AM1/6/10
to
On Jan 5, 5:54 pm, "Michael Robinson" <nos...@billburg.com> wrote:
> How does the current limiting work in the MC33063.   The datasheet doesn't
> explain the actual mechanism.http://www.onsemi.com/pub_link/Collateral/MC34063A-D.PDF

> Does it discharge the timing cap when it detects overcurrent, which would
> restart the oscillator, or does it let the oscillator keep running but at a
> lower duty cycle?

It truncates the duty cycle when the INPUT current reaches the trip
point; the
output current is apparently not sensed directly. It's only a limit
on the peak
value of the input current, not an average, but of course that value
and the
operating frequency sets a current output limit.

E

unread,
Jan 6, 2010, 3:26:07 AM1/6/10
to

"Michael Robinson" <nos...@billburg.com> kirjoitti
viestiss�:w3S0n.1953$ap2....@newsfe18.iad...

It charges timing capacitor to some high voltage if input current exceeds
threshold
=> Output disabled until discharged back to normal current.
-ek


Raveninghorde

unread,
Jan 6, 2010, 5:59:46 AM1/6/10
to

The other posters have expalined how it works.

The current limit is poorly defined. It can not be used to provide a
constant current supply. If you need that then you need extra
components.

Hammy

unread,
Jan 6, 2010, 6:39:03 AM1/6/10
to
On Tue, 5 Jan 2010 20:54:40 -0500, "Michael Robinson"
<nos...@billburg.com> wrote:

You could use the NCP3063 or MC34063 as a constant current driver but
why would you? You could program the output current by Iout =
Vref/Rout. But you're much better off using a constant current
controller with a much lower reference like the TPS40211 or Onsemis
NCP3065.

This application note describes an elaborate CC driver using the
NCP3063

350 mA CC Driver

http://www.eetasia.com/ARTICLES/2008JAN/PDF/EEOL_2008JAN31_OPT_CTRLD_AN.pdf?SOURCES=DOWNLOAD

700 mA CC driver

http://www.eetindia.co.in/STATIC/PDF/200809/EEIOL_2008SEP12_OPTO_AN_01.pdf?SOURCES=DOWNLOAD

For comparison here is the NCP3065 constant current controller. This
controller is almost identical to the 3063 hysteretic control etc but
with a 235mV reference.

http://www.onsemi.com/pub/Collateral/NCP3065.PDF

It has already been explained how the current limit works but here is
a more detailed application note. See page 3.

http://www.intusoft.com/onsemipdfs/AN920-D.pdf

Hammy

unread,
Jan 6, 2010, 6:51:47 AM1/6/10
to
On Wed, 06 Jan 2010 06:39:03 -0500, Hammy <sp...@spam.com> wrote:

How not to build a constant current driver with the NCP3063.

http://dev.emcelettronica.com/cheap-high-current-led-driver#comment-1782

Phil Allison

unread,
Jan 5, 2010, 9:59:23 PM1/5/10
to

"Michael Robinson"


** OnSemi do have "app notes" you know !!

http://www.onsemi.com/pub_link/Collateral/AN920-D.PDF

See page 3 of the pdf, just under figure 4.

Seems the current sense circuit ( Ipk Sense) INCREASES the current going
to Ct and hence makes the output stage on time less. So the second of your
hunches applies.

..... Phil


Jamie

unread,
Jan 6, 2010, 7:48:14 PM1/6/10
to
My guess it would have to be duty cycle..

Current limiting takes precedence with
the value of Rsc, (6) and (7) obviously are
some internal comparator that will gate the
duty cycle shorter when needed.

If this isn't enough to maintain V, the comp
for V just simply attempts a max duty cycle, which
will only be what the current limiting circuit allows.


Briefly looking at it, that is what I get from it..

0 new messages