Data provided by: Mr. hisham bahaa-aldeen (firstname.lastname@example.org)
SINEWAVE UPS TESTING AND FAULT FINDING
Construct the card
thereby confirming each and every wiring, this includes LED connectivity, ON/OFF
switch, feedback via inverter transformer, 6-volt mains sense to CN5, -VE
of battery to card, +VE of battery to large heatsink.
not plug the transformer primary to the pair of small heat sinks.
battery +ve wire to PCB via MCB and 50-amp ammeter.
Prior to proceeding
for the recommended testings be sure to check the +VCC voltage at the pins of
U1 - U5 in the following sequence.
U1:pin#8 and 9: +5V, pin#3: +12V, pin#6: +12V,
U2:pin#8 and 9: +5V, pin#3: +12V, pin6: +12V,
U3: pin14: +5V, U4: pin20: +5V, pin1:+5V, U5: pin4:+5V.
1) Power Up the
battery MCB and check the ammeter and also be certain it doesn't jump beyond
1-amp. If the ampere shoots then remove U1 and U2 briefly and switch ON the MCB
2) Power ON by toggling the given ON/OFF switch of the inverter and check whether or not the relay clicks ON, illuminating the "INV" LED. If it doesn't then check the voltage at pin#18 of the PIC which is supposed to be 5V. If this is absent check components R37 and Q5, one of this may be faulty or incorrectly connected. If you find the "INV" LED not switching ON, check if the voltage at pin#25 of the PIC is 5V or not.
If the above situation is seen to be normally executing, go to the next step as described below.
3) Using an oscilloscope test pin#13 of the PIC by alternately switching ON/OFF the inverter switch, you can expect to see a well modulated PWM signal appearing at this pinout each time the inverter mains input is switched OFF, if not then you can assume the PIC to be faulty, coding not implemented correctly or the IC is badly soldered or inserted in its socket.
If you succeed in getting the expected modified PWM feed over this pin, go to pin#12/in#14 of the IC and check the availability of 50Hz frequency on these pins, if not would indicate some fault in the PIC configuration, remove and replace it. If you are to get affirmative response on these pins, go to the next step as explained below.
4) The next step would be to test pin#10/pin#12 of the IC U3 (CD4081) for the modulated PWMs which are finally integrated with the mosfet driver stages U1 and U2. Additionally you would be also required to check the potential differences at pin#9/pin#12 which is supposed to be at 3.4V approximately, and at pin#8/pin#13 may be verified to be at 2.5V. Similarly verify pin#10/11 to be at 1.68V.
In case you fail to identify the modulated PWM across the CD4081 output pins, then you would want to verify the tracks terminating to the relevant pins of the IC CD4081 from the PIC, which could be broken or somehow the obstructing the PWMs from the reaching U3.
If all is fine, let's move to the next level.
5) Next, attach the CRO with U1 gate, toggle the inverter ON/OFF and as done above verify the PWMs on this spot which are M1 and M4, and also the gates M9, M12, however don't be surprised if the PWM switching are seen out of phase M9/M12 as compared to M1/M4, that's normal.
If the PWMs are entirely absent on these gates, then you can check pin#11 of U1 which is expected to be low, and if found high would indicate that U1 may be running in the shut-down mode. To confirm this situation check voltage at pin#2 of U5 which could be at 2.5V, and identically pin#3 of U5 could be at 0V or under 1V, if it's detected to be below 1V, then proceed and check R47/R48, but if the voltage is found to be above 2.5V then check D11, D9, along with mosfets M9, M12 and the relevant components around it to troubleshoot the persisting issue, until corrected satisfactorily..
In case where the pin#11 of U1 is detected low and still you are unable to find the PWMs from pin#1, and pin#7 of U1, then it's time to replace IC U1, which would possibly rectify the issue, which will prompt us to move to the next level below.
6) Now repeat the procedures exactly as done above for the gates of the mosfet array M5/M18 and M13/M16, the troubleshooting would be exactly as explained but with reference to U2 and the other complimentary stages which may be associated with these mosfets
7) After the above testing and confirmation are completed, now it's finally time to hook up the transformer primary with the mosfet heatsinks as indicated in the sinewave UPS circuit diagram. Once this is configured, switch ON the inverter switch, adjust preset VR1 to hopefully access the required 220V regulated, constant sinewave AC across the output terminal of the inverter.
If you find the output to be exceeding this value or below this value, and void of the expected regulation, you may look for the following issues:
If the output is much higher, check voltage at pin#3 of the PIC which is supposed to be at 2.5V, if not then verify the feedback signal derived from the inverter transformer to connector CN4, further check voltage across C40, and confirm the correctness of the components R58, VR1 etc. until the issue is rectified.
8) After this attach an appropriate load to the inverter, and check the regulation, a 2 to 3 percent falter can eb considered normal, if still you fail a regulation, then check diodes D23----D26, you can expect one of these to be faulty or you may also try replacing C39, C40 for correcting the issue.
9) Once the above procedures are successfully completed, you can carry on by checking the LOW-BATT functioning. To visualize this try short circuiting R54 with the help of a pair of tweezers from the component side, which should instantly prompt the LOW-Batt LED to illuminate and the buzzer to beep for a period of around 9 seconds at the rate of a beep per second approximately.
In case the above does not happen, you may check pin#4 of the PIC, which should be normally at above 2.5V, and anything lower than this triggers the low batt warning indication. If an irrelevant voltage level is detected here check whether or not R55 and R54 are in a correct working order.
10) Next up it would be the overload tripping feature which would need to be confirmed. For testing you can select a 400 Wait incandescent bulb as the load and connect it with the inverter output. Adjusting VR2 the overload tripping should initiate at some point on the preset rotation.
To be precise, check the voltage at pin#7 of the PIC where under correct load conditions the voltage will be over 2V, and anything above this level will trigger overload cut-off action.
With a sample 400 watt, try varying the preset and try forcing an overload cut -off to initiate, if this does not happen, verify voltage at pin#14 of U5 (LM324) which is supposed to be higher than 2.2V, if not then check R48, R49, R50 and also R33 any of these could be malfunctioning, if everything's correct here simply replace U5 with a new IC and check the response.
Alternatively you can also try increasing the R48 value to around 470K or 560k or 680K etc and check if it helps solving the issue.
11) When the assessment of inverter processing is finished, experiment with the
mains changeover.Keep the mode switch in inverter mode (keep CN1 open) switch-ON
the inverter, hook up the mains wire to the variac, step up the variac voltage
to 140V AC and check the inv to mains changeover triggering occurs or not. If
you find no changeover in that case confirm the voltage at pin2 of
microcontroller, it needs to be > 1.24V, in case the voltage is smaller than
1.24V then inspect the sensing transformer voltage (6V AC at its secondary) or
take a look at the components R57,R56.
Now that the changeover
shows up scale down the variac voltage to below 90V and examine the
mains-to-inverter changeover action is established or not. The changeover
ought to happen since now the voltage at pin2 of microcontroller is less than
12) Soon after the above assessment is completed, experiment
with the mains-changeover in the UPS mode. Enabling the mode-switch in the UPS
mode (keep CN1 shorted) start the inverter, link up the mains wire to the
variac, increment the variac voltage to around 190V AC and observe the
UPS-to-mains changeover strikes or not. Should there be no changeover action
then simply take a look at the voltage at pin2 of microcontroller, it needs to
be over 1.66V, as long as the voltage is lower than 1.66V then simply confirm
the sensing transformer voltage (6V AC at its secondary) or perhaps inspect the
Right after the changeover pops up, scale back the
variac voltage to 180V and find out whether the mains-to-UPS changeover comes
about or not. The changeover ought to strike since now the voltage at pin2 of
microcontroller could be witnessed to be over 1.5V.
Eventually take a look at the customized charging of the attached battery. Hold the mode
switch in the inverter-mode, administer mains and step up the variac voltage to
230V AC, and determine the charging current which should rise smoothly in
Fiddle with the charging current by varying VR3, so
that the current variation could be witnessed varying in the middle of around
5-amp to 12/15-amp. Just in case the charging current is seen to be much
higher and not in a position to be scaled down at preferred level then you may
try increasing the value of R51 to 100k and/or if still that does not
improve the charging current to expected level then perhaps you can try
decreasing the value of R51 to 22K, please bear in mind that once the
sensed equivalent voltage at pin5 of microcontroller becomes at 2.5V
the microcontroller may be expected to regulate the PWM and consequently
the charging current.
In the course of the charging mode remember
that, precisely the lower branch of MOSFETs (M6 -M12 / M13 - M16) are switching
@8kHZ while the upper branch of MOSFETs are OFF.
14) Additionally you can
inspect the operation of the FAN, FAN is ON each time the inverter is ON, and
FAN could be seen switched OFF whenever the inverter is OFF. In a similar manner
FAN is ON as soon as Charging is ON and FAN will be OFF when charging is OFF