Hi!
I've made a power source(out 1.3V & from Vin(3.3V)) with LP3965.
I think everything is o.k. in conditions.
But if load current is big, output voltage drops like a ramp.
Because the pass element doesn't work at this time, the ramp wave form must be made by load current.
1) Another charactoristics of my circuits is as follows ;
DC 24V --> Switching Reg.(Buck, 3.3V,3A) ---> LDO input and Another components(DRAM )
2) Wave forms
Yellow line : Vin=3.3V, Red line Vout=1.3V
Which package type is this (or, exact part #), and just how much current is 'big' ( "... if load current is big ...") ?
I presume this is -ADJ version, what are your values for R1, R2, and CF ?
If the waveform included a line for load current, that would be helpful.
Thanks for your promptly response.
It's part name is LP3965EMP-ADJ and the package is sot 223-5.
My external parts values are R1 = 10k , R2 = 690 Cf=100pF.
During A down ramp, the LDO's output current is zero.
The wave form is attached below.
Yellow Line : 3.3V as the LDO input.
Red Line : 1.3V as the LDO output.
Green Line : the LDO output current ( between LDO output and Capacitor )
Based on what I see, I can only think of two things that would cause output current to go to zero : Thermal Shutdown, and Shutdown pin voltage.
The current spikes are from the output turning back on and recharging the output capacitor.
Thermal shutdown seems unlikely. Even with minimum copper area (see "Figure 6, Scenario A" in the data sheet) and dissipating ~300mW, the internal temperature rise would be ~45°C above the ambient temperature. Unless you are operating in a hot environment (i.e. ambient temperature > 115°C ) this not likely a probable cause.
That leaves the voltage on the "/SD" pin as the remaining probable cause.
The data sheet calls for the /SD pin to be less than 0.2V for Output=Off, or within 0.3V of Vin for Output=On. These are most likely idealized On/Off values, with the real On/Off threshold voltage(s) being unknown.
How are you biasing the /SD pin? Could you measure the voltage at the /SD pin relative to the GND tab?
Thank you for promftly response again.
1. Thermal shutdown is not likely the real cause as your opinion.
When I touch the element with my fingers, It is not hot at all.
2. The /SD pin is pull-up with input(R=10k, 3.3V).
And the wavwform folllows. (yellwo line is LDO output 1.3V, Red line is the /SD pin=3.3V)
3. There are Two Pehenomena I experienced as follows
1) When I change the LDO's input from 3.3V to 5.0V, the LDO does work very well.
( The 3.3V power is provided as LDO input and DRAMs in the PCB )
( Also they probably work at same time as if cpu core access to DRAM )
2) When I change the LDO's input to independant enternal 3.3V, the LDO does work exectly also.
Sounds like you are likely getting noise spikes from the 3.3V through the 10k resistor into the /SD pin.
If the voltage spikes go low enough the LP3965 will be shutdown momentarily. To confirm this I would ask you to either power the /SD pin from another (noise free) supply, or add a low value capacitor (0.1uF) from the /SD pin to Gnd and keep the 10k resistor in place to create an RC averaging filter.
I have attached a plot of the Vsdt On/Off thresholds vs. Vin. This is typical only, not guaranteed, and for room temperature only. It does show that the On/Off thresholds have a stong dependence on the input voltage.
Some Test Result..
Thanks for your prompt response always.
As you suggest, I‘ve measured the voltage of the /SD pin at the LDO.
1) In case of adding 0.1uF capacitor
=> It's same to previous wave form with voltage drop.
2) in case of using the independant external 3.3V input
3) In case of adding /SD and Vin Short
4) In case of adding /SD Pull Up 10K
Because I could not find the cause of the LDO's voltage drop, I tried to test three different LDOs with same circuits.
1) Substituting the LM11117(LDO adj, National) for LP3965.
==> 1.3V output is O.K. ( there is no voltage drop )
2) Substituting the MCP1727(LDO adj, MICROCHIP ) for LP3965.
3) Substituting the LTC3406(switching) for LP3965.
Besides,
When providing the LDO(LP3965) with external 3.3V , the LDO's output is O.K.
As follows;
At this point I have run out of ideas. The behavior you have described here is certainly not consistent with LP3965 performance.
It might be that the LP3965 is somehow damaged or defective.
Hi Jones
Really I have got the strange result i never seen before.
I cut the pad line between LDO's input (3.3V) pin and Input capacitor.
And I connect the two points with about 5cm wire. ( The first picture)
The result is very good. (The second picture)
But, when I use a very short wire between them, output voltage drop appears in disp
What makes me be confused?
I'd like to get your bright explanation.
After bring in my first question into this forum, I always appreciate your kindness.
I could not understand that the cause of output voltage drop on LP3965-ADJ seems to be the level of /SD pin,
And I replace it with LP3855-ADJ in the PCB. As you know, they are pin-compatible.
The LDO(LP3855-adj :Vout=1.3V) does work very well.
The difference of them is the threshold voltage level on /SD pin.
I'd like to ask you some questions.
1) When the LDO(LP3965-ADJ)'s current flows above 300mA, Why is the LDO sensitive ?
And have you exprienced(or received) the familiar question about the LDO?
2) The LDO(LP3855-adj) i replaced works very well. what can you recommand me to do(or check) ?
I've already producted 4000sets with the LP3965-adj.
I had to explain to my boss that the LDO was changed with LDO(LP3855-adj).
Best Regards
Q1a : When the LDO(LP3965-ADJ)'s current flows above 300mA, Why is the LDO sensitive ?
A1a : There should be no measurable difference for the LP3965-ADJ On/Off thresholds due to output load current.
Q1b : And have you exprienced(or received) the familiar question about the LDO?
A1b: No.
Q2 : The LDO(LP3855-adj) i replaced works very well. what can you recommand me to do(or check) ?
A2 : This behavior is not normal for LP3965. I do suggest that you contact your vendor to:
a) arrange for warranty replacement;
b) to have some LP3965-ADJ samples returned to us for Failure Analysis; or,
c) both a and b
When vendor submits samples for Failure Analysis please ask them to include "Refer to E2E Thread ID: 155579" on the form in the failure description text, along with their description.