FreeLists / si-list / [SI-LIST] problem adjustable low dropout regulator

["Jan Vercammen" <jan.vercammen1@xxxxxxxx>]

Hello si-list,
first of all my best whishes for 2004.

I need your help on a problem with an adjustable low dropout regulator or 
LDO for short.
We have spent considerable effort on this and we are stuck. It concerns a 
National LP3964-ADJ
800mA LDO in a SOT-223 package.
I am not stating that this component is a bad device, only that in our 
application we are 
experiencing some unexpected problems.

Let me first decribe the LDO. It is a CMOS chip with a P-MOS pass 
transitor. A resistor feedback 
divider (with one capacitor) is compared with an internal reference 
voltage of 1.215V. The circuit
is as described in the data sheets and is very simple.
The input voltage is 3.3V, the output voltage is 2.5V. Both voltages are 
supplied by internal planes.
The 2.5V is suplied to the core of a Motorola MPC107 bridge and to the 
core of an Altera EP10K30 
FPGA. We have measured a 2.5V current of 230mA. These LDO circuits are 
used on several
printed circuit boards (most PCBs are 12 layers, one is 10 layer).

We have experienced two problems, they are not necessarily related, but 
likey are. 
- First problem: the 2.5V drops exponentially by about 300-400mV over 
50-100us and then ramps up
very quickly in about 2-3us. The drop starts - we think - when the 107 
bridge and FPGA enter a low
activity faze. The ramp - we think - starts when the bridge and FPGA 
become active again. The
voltage drop of 400mV on the bridge and FPGA core voltage does not create 
any functional problems!
- Second problem: there are problems with the target value of the 2.5V. On 
most boards we have 
a problem of obtaining 2.5V, we only get 2.25V to 2.4V. We also 
experimented with a setting of 2.7V, which
we need for a future design. If we adjust the resistors for 2.8V, we only 
obtain 2.5V. We have measured the 
voltage after the resitive divider an found about 1.15V and not  - what 
you would expect - the reference
voltage of 1.215V. It seems that the internal voltage reference has 
changed in value or that the 
internal amplifier(s) is plagued by DC offsets.

Here is what we did to improve some of the problems:
-1- a non-adjustable LDO (with internal feedback network) does not have 
the DC-offset or the voltage drops!
-2- we used Kelvin connections for the feedback and bulk (tantalum) 
decoupling, but this does not seem to help
-3- a 1.5Amp LDO version seems to behave somewhat better with respect to 
DC-offset and voltage drops
-4- using 1 ohm series resistors on the tantalums  (2x22uF) helps to 
remove the voltage drops
-5- a bleeder resistor (25-100mA) helps to remove some of the voltage 
drops
-6- mounting the feedback circuit to the top of the PCB does not make a 
difference
-7- using a ferrite bead in series with the input does not seem to help
-8- the voltage drops are very sensitive to the output decoupling, but 
this depends on the ESR of the bulk (tantaal)
      decoupling and not on the ceramic capacitors (about 12x 100nF on 
2.5V)
-9- extra resistance and/or inductance on the output removes the voltage 
drops - e.g. a long fat track (of about 16mOhm)
      or a power supply track grid 
-10-  a higher input voltage (of 5V) does not help
-11- removing the FPGA from the 2.5V does not help

So here are our questions:
-1- can anyone explain why a fixed LDO (using an internal feedback 
network) does NOT have the problems
      of a adjustable LDO. A fixed LDO does not need the measures of the 
list above - it is always OK!!
-2- is it possible that the SOT-223 package is the problem? Has anyone 
experience in this respect? Is another
      package better (e.g. TO-263 or TO220)??
-3- can anyone explain the shift of the inernal voltage reference? Or are 
the internal OPamps experiencing
      an out-of-band response of RF noise on the 3.3V or 2.5V??
-4- Can you use an LDO with a 3.3V+/-5%  input voltage and 2.7V+/-5% 
output voltage??

We are stuck here. Currently one scenario is to consider another component 
from - possibly - another
manufacturer. But that may not be a good idea as we may end up in the same 
situation.


Any hint or help is appreciated.

Kind regards,

Jan Vercammen





Kind regards,

Jan Vercammen
Agfa-Gevaert NV
Mortsel, Belgium



------------------------------------------------------------------
To unsubscribe from si-list:
si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field

or to administer your membership from a web page, go to:
http://www.freelists.org/webpage/si-list

For help:
si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field

List technical documents are available at:
                http://www.si-list.org

List archives are viewable at:     
                http://www.freelists.org/archives/si-list
or at our remote archives:
                http://groups.yahoo.com/group/si-list/messages
Old (prior to June 6, 2001) list archives are viewable at:
                http://www.qsl.net/wb6tpu