Here's the situation.

  • I'm possessed of an inverter purchased nearly a decade ago.
  • A few months ago it started to make a clicking noise of it's own volition
  • The repairman said he replaced the relay and then it worked fine ... until now
  • By trial-and-error I discovered that enabling the bypass switch (so the inverter is powered from the battery) would kill the tripping. Killing the bypass brought up the tripping again.
  • More importantly, the longer the duration the bypass was enabled, the longer the wait before it would start to trip again; slowly at first and picking up speed with each cycle.
  • The repairman who fixed it originally and the other ones I've contacted are not bothered and recommend replacing the inverter altogether!
  • I have some basic exposure to electronics; my impression is that the relay is good but there is some condensor out of the kilter.

The idea of replacing the inverter merely because it flips isn't fun, and stinks of an attempt to earn some commission on the sale. I'll replace if I must, but not without giving it a try myself!

What I would like to know from you

  • Does my impression about the condensor sound right?
  • How should I go about finding out the fault?
  • 2
    \$\begingroup\$ How long are the cycles? If left on battery for xxx how long does it stay on mains without the fault? Is the relay the mains to inverter changeover? // Highly speculative idea: Maybe the unit faults when the battery is charged. If so, running on bat discharges the bat - longer bat time = longer to fault. When the battery is fully charged the fault returns. Test this by running on battery for 30 seconds on no load and on largest possible load. Is the fault return time much larger after loading? Caps die with age. Check appearances. Check ESR . DISCHARGE WELL AND FULLY BEFORE ACCESSING !!! \$\endgroup\$ – Russell McMahon Aug 9 '11 at 11:06
  • \$\begingroup\$ a) When the inverter was on battery (in bypass) for a couple of hours, the initial cycle was a little over 30 seconds. Within a minute it was flipping too rapidly to count. b) The relay is indeed the mains/inverter changeover c) No, it also happened when the battery was discharged. I know this because the repairman took nearly a month after I called him to turn up. During this period, the battery was never charged and I used the battery itself for other stuff. d) Could you please elaborate on the 'Check appearances'? \$\endgroup\$ – Everyone Aug 9 '11 at 12:14
  • 1
    \$\begingroup\$ Electrolytic capacitors may swell with age or blow vents on top, or sometimes at board level. If they look bad they are probably dead. If they look good they are possibly dead :-). At 10 years old electrolytics will often be dead. \$\endgroup\$ – Russell McMahon Aug 9 '11 at 12:38
  • \$\begingroup\$ Got it! Thank you! (+: I'll check it out during daylight hours otmorrow \$\endgroup\$ – Everyone Aug 9 '11 at 13:02
  • 1
    \$\begingroup\$ Here is some discussion on repair questions: meta.electronics.stackexchange.com/questions/684/… essentially the root of it is, if the question is asked around learning how everything works then it can potentially be ok, but this question is based purely around trying to find the easiest way to fix a device from a consumers perspective. \$\endgroup\$ – Kellenjb Aug 9 '11 at 16:04

I seems your invertor is detecting a false brown out condition (low Vac) or a DC reference voltage is too high inside.


I agree with the others who pointed out as you did, that a capacitor may be the sypmtom of failure. When capacitors (condenser) are under-rated for temperature or surge current, they wear out faster than normal. A healthy margin for voltage, max current & temp. allow for normal long life of 25yrs. Otherwise these stress factors can reduce it to less than a year such as the huge PSU recalls from major PC companies that are no longer in business.

The normal diode charge surges with internal ESR creates heat. Heat accelerates leakage aging and also increases ESR aging. The same is true for car batteries. Some electrolytics standardize at 85'C and offer higher reliability types at 125'C. This would be a wise choice if considering one cap. Also consider replacing all big Caps.

When the electrolyte exceeds it's operating temperature by some margin, the pop-corn lid expands to offer safe containment of the nasty chemical. (read MSDS sheet)

An inverter consists of;

  • a battery charger and battery
  • an inverter and
  • a controller

The controller and user front panel options decide when to charge, when to invert and when to switch from inverter output to mains output.

Symptoms and Analysis

Flip Flop switching in Auto mode If the line voltage is detected as too low, it switches over at the Brown-Out(BO) threshold before a line failure occurs.

If the charger current is high and that loads the line voltage BOT then the time it will take longer for the battery to return to idle charge current and thus toggle rate will be slow. As the charger current reduces, the toggle rate will increase.

So it appears the battery charger could be overloading the BO threshold and the BO threshold detector may need re-calibration. If the BO threshold comes from the same transformer(XFMR) secondary from the AC line, that might explain some interaction, but not completely.

Something could be loading down the XMFR secondary low voltage out. Could it be the battery is weak and lossy overloading the charger? Test S.G. in each cell. Replace battery if necessary. Test Charger voltages not to exceed 14.2V on battery (if 12V)

If the battery was the root cause then it might explain the results, but it also indicates an under-sized charger or at the least a poor way of detecting BO trip point... and poor method of detecting battery failure.

The next possibility, is there was a change in BO threshold and/or your AC line levels due to some other component failure.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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