I'm trying to repair a "Flexineb E2" veterinary nebuliser for somebody, it holds a charge and works plausibly well if the battery connection is briefly interrupted but will not power-on if the power button is pressed.

PCB is 4-layer with multiple boost PSUs, and I'm very much trying to avoid tracing the circuit :-)

The power on/off button is momentary and looks like it has a FET across it to maintain power once the unit's on. There's a PIC33FJ12, which appears to do little else other than monitor the button for turn-off and presumably to detect that all medication has been dispensed.

If I short the PIC MCLR/ signal to ground the unit comes on, so an easy fix would be an external reset button.

MCLR/ has a 1k0 (SMD resistor marked 1001) pullup to 3v3 (labelled testpoint) with no nearby cap that I can see.

The datasheet etc. suggests a 10k pullup. Is it plausible that the problem as observed is due to the 1k0 pullup being vastly too strong?

Datasheet at https://ww1.microchip.com/downloads/en/devicedoc/70265b.pdf

  • \$\begingroup\$ Please link the datasheet. And yeah the pull-up could be too strong, but there's no telling without digging through the datasheet. \$\endgroup\$
    – Lundin
    Commented Mar 1 at 13:29
  • \$\begingroup\$ @Lundin Done. The reason that I'm not just swapping Rs is that there's a thin layer of conformal. \$\endgroup\$ Commented Mar 1 at 13:57
  • \$\begingroup\$ Have you measured the push button already? Maybe it has high resistance and is not pulling down the line correctly anymore. While 1k PU ist quite strong, I doubt the manufacturer would place it there if the product does not work correctly with it. Before changing resistor value I would make sure the obvious things (like push button) work correctly. \$\endgroup\$
    – jusaca
    Commented Mar 1 at 14:02
  • \$\begingroup\$ In a different processor, power supply sequencing caused similar problems. Troubleshooting it wasn't easy, and the ultimate fix counterintuitive. You may need to trace circuits. \$\endgroup\$
    – glen_geek
    Commented Mar 1 at 14:04
  • \$\begingroup\$ @jusaca Button seems fine (apart from anything else always turns the unit off OK), and is wired into the power circuit a long way from the PIC. \$\endgroup\$ Commented Mar 1 at 14:05

1 Answer 1


The nebuliser was needed fairly urgently, so in the interest of getting a marginally-useful fix (for what might be a common failure mode) into the record.

Provided that:

  • With the head connected and the charger and battery disconnected, plugging in the charger activates the nebuliser.

  • With the head connected and the charger and battery disconnected, connecting the battery activates the nebuliser.

  • In both the above cases, pressing the button deactivates the nebuliser.

  • With the head and battery connected, shorting the PIC reset (pin 1) to the marked ground activates the nebuliser.

An acceptable hack is to desolder the "on/off" button and to connect it instead between the PIC reset (pin 1) and the marked ground. With this done, the button turns the nebuliser on (green LED comes on) and it may be turned off by disconnecting the atomizer head (red LED comes on, shuts down after a few seconds).

The downside of this is that the PIC is permanently running, so the nebuliser benefits from being kept on the charger. Apart from that, since I believe there was no automatic cutoff once the medication had been dispensed the overall capabilities are very similar to the "as shipped" state.


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.