I am using a TI LM2776 Switched Capacitor Inverter (datasheet here) to obtain a -3.3v bias voltage for an LCD (Newhaven Display, NHD-12864MZ-FSW-GBW-L, datasheet here).

The LM2776 is connected exactly as shown in this digram from its datasheet:

LM2776 Typical Application

Vin is 3.3v, EN is connected to Vin, and I'm using the same capacitor values.


When I turn on the circuit with no load, the output of the LM2776 is about -0.9v. When I connect it to the LCD bias pin to load it, the output is about -0.7v. Why am I not seeing about -3.3v?

Other Notes

LCD Load

When I connect the LCD bias voltage pin to a DC power supply (at -3.3v) instead of the LM2776, the LCD draws about 20mA, so the LCD should be an appropriate load for the LM2776. According to the datasheet, I shouldn't be seeing any significant voltage drop at this current draw:

LM2776 Vout vs Iout


All capacitors are ceramic X7R radial capacitors.


The circuit is currently on a breadboard, and the LM2776 is connected to it through a SOT23-6 -> DIP adapter. I know these devices can be sensitive to layout considerations, but the same configuration worked a couple of weeks ago, and I have no reason to suspect there was damage of any kind done to it since then.

LM2776 Breadboard

  • \$\begingroup\$ Pictures of your circuit layout and details about the capacitors would be appropriate, as switching converters are sensitive to these things. \$\endgroup\$
    – uint128_t
    Mar 13, 2016 at 7:09
  • \$\begingroup\$ @uint128_t Added. \$\endgroup\$
    – skrrgwasme
    Mar 13, 2016 at 7:29
  • \$\begingroup\$ @uint128_t I know they're sensitive, but since I've gotten it to work before, I thought I was set. Can you think of any reason they may have worked before when they probably shouldn't have? \$\endgroup\$
    – skrrgwasme
    Mar 13, 2016 at 7:48
  • \$\begingroup\$ What kind of power supply are you using for the 3.3V? Can you confirm it fulfills the current supply specifications for LM2776 (EN=1) and for the LCD? \$\endgroup\$
    – Alf
    Mar 13, 2016 at 11:38
  • \$\begingroup\$ Can you get a scope plot at the C1+ pin (or anywhere on that net)? I would be suspicious that the device is not switching. \$\endgroup\$ Mar 13, 2016 at 13:38

2 Answers 2


The breadboard prototype of this circuit did work a couple of times, (maybe 3 out of 50 tries), but those may have just been flukes that mislead me into thinking that something must have changed to make it not work anymore.

When I assembled the circuit on a PCB with all surface-mount capacitors that adhere to the layout recommendations given in the LM2776 datasheet, it works 100% of the time.

I suspect uin28_t's comment was on the right track - the through-hole capacitor leads and DIP adaptor for the IC introduced enough impedance for the LM2776 to be unable to maintain the correct voltage, and I just got lucky the rare times it did work.

  • \$\begingroup\$ "When I assembled the circuit on a PCB with all surface-mount capacitors that adhere to the layout recommendations given in the LM2776 datasheet, it works 100% of the time." This, exactly. Just worked on a similar problem with the LM2776; having all three capacitors too far away lowered the output voltage to -2.0v when +5.0v was used as a supply and no load on output. Moving them directly next to the LM2776 using 0805 SMTs worked. \$\endgroup\$
    – dgreene
    Mar 12, 2020 at 17:34

The suggested layout at section 10.2 of the datasheet does not match the part pinout in section 5! I have been through much grief finding this out. http://www.ti.com/lit/ds/symlink/lm2776.pdf

  • \$\begingroup\$ Do you mean sections 5 and 8.2? I remember being confused by that, but 10.2 and 5 match. \$\endgroup\$
    – skrrgwasme
    Aug 23, 2018 at 22:48

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.