max756 dc-dc giving wrong output voltage? 3.7V

[ecerulm]

I connected the MAX756 as stated in the datasheet http://datasheets.maxim-ic.com/en/ds/MAX756-MAX757.pdf


L1 is a coilcraft 22uH
D1 is a 1N5817
C4 and C3 are 100uF

I'm applying 1.6V from a AA battery cell. When I measure the Vout with a multimeter it gives a reading of 3.7V instead of the expected 3.3V . What can be happening?

In the other hand if I connect the 3/~5 pin to GND. I get 5.12V in Vout, (expected 5.0V). So it seems that the regulator is working fine for 5V but not for 3.3V. (?)

[Zener]

Do you have a PCB or is this on a breadboard?

The data sheet says this:

"PC Layout and Grounding

The MAX756/MAX757 high peak currents and high-frequency
operation make PC layout important for minimizing
ground bounce and noise. The distance
between the MAX756/MAX757’s GND pin and the
ground leads of C1 and C2 in Figure 1 must be kept to
less than 0.2" (5mm). All connections to the FB and LX
pins should also be kept as short as possible. To
obtain maximum output power and efficiency and minimum
output ripple voltage, use a ground plane and
solder the MAX756/MAX757 GND (pin 7) directly to the
ground plane."

You could also try putting a little load on it, assuming it has none now, even though the data sheet says it is OK for 0 load. If it is not on a good PCB then anything could be happening.

[adafruit]

pin 2 should go to battery input +

[ecerulm]

Connecting pin 2 to battery input Vin doesn change the output still 3.7V, the evaluation kit for max756 seems to connect pin2 to Vout. http://www.datasheetcatalog.org/datashe ... 6EVKIT.pdf.

It's a breadboard setup, and I have a led + 220 ohm resistor load.

Is it possible that the output voltage ripple is so big to account for that +0.4V in the output?

I also found that changing the battery source from 1.6V AA to a coin cell 2.97V makes a big difference. With the coin cell the output is 3.31V (no load) and 3.6V with load (led + 220ohm) .

In the datasheet they say that 1.8V is guaranteed startup voltage, mine it's 1.6V, so I guess that could be it.

I'm not sure about how my cheap digital multimeter BS1901W actually makes the voltage measure. Could it be that the ripples in the output is confusing the multimeter into giving the wrong reading?

Also when I switched back to the 1.6V cell and measured the output voltage without load I can see that the reading in the multimeter never stabilizes. it goes from 4.12V to 3.35V and 4.12V again (over a period of maybe 10 seconds). But when I apply load (led+220ohm) I get a stable 3.7V.

I think I'll setup a voltage divider and see what happens when Vin (battery voltage) is 2.0V, 1.8V and 1.6V at startup. I'll post my findings here.

[ecerulm]

I´ve been reading a bit about SMPS and people stress the fact that the ESR of the output capacitor is very important to minimize output voltage ripple . I'm using a regular electrolytic capacitor 100uF. Do you think this would improve with a better lower ESR capacitor?

[Zener]

I am sticking with my original answer:

"High peak currents and high-frequency operation make PC layout important for minimizing ground bounce and noise."

Try moving your black probe of your meter to different ground points and see if you always get the same voltage.

What is the voltage measured between pins 7 and 6?

[ecerulm]

I tried several ground points and I get the same readings no matter where I put the probe.

The voltage between pins 6 and 7 is 2.2V

Also I want to inform you that I mounted the same setup in another breadboard with better results. Now I get 3.3V output with the 1.6V battery cell as input as expected.

I don't know what's happening in the other breadboard, though.

[JohnOC]

I tried several ground points and I get the same readings no matter where I put the probe.

The voltage between pins 6 and 7 is 2.2V

Also I want to inform you that I mounted the same setup in another breadboard with better results. Now I get 3.3V output with the 1.6V battery cell as input as expected.

I don't know what's happening in the other breadboard, though.

Possibly a dirty or loose set of contacts in the breadboard is adding extra resistance.