I
Ibis Designs
This posting may be of interest to anyone continuing to use Enecsys micro inverters:
The Enecsys micro inverters use a DC to DC boost converter to provide the approximately 550V DC required for the output stage. Under normal operating conditions when the units are generating and the sun is shining the operating voltage is close to 550V and all is well.
The attached snip from the schematic I have drawn shows a simple bridge rectifier feeding the main bank of storage capacitors prior to the high voltage HF switching stage. Enecsys have sensibly provided two series connected high voltage transient suppression zener diodes connected across the main DC lines via R167 and D31. The combined breakdown voltage for the network is 550V so all well (just).
Now suppose for some reason the mains fails but the sun continues to shine? The load on the 550V rail drops to zero and the DC potential rises to around 570V. D59 and D60 begin to conduct and the dissipation in D60 raises the temperature beyond t max and D60 goes short circuit. A few milliseconds later D59 follows in the same way and this puts a huge voltage across R167 which quickly goes open circuit.
The momentary loss of the DC HV rail is detected by the 8 bit PIC microcontroller and the unit shuts down.
Eventually the owner power cycles the unit and there is a good chance that if both mains supply and PV are restored the unit will continue to function again and all will appear normal. However, the potential across the damaged R167 eventually breaks down the tiny break in the conductive material which momentarily restores the low resistance load across the HV and triggers the unit back into a shutdown.
From the users point of view the inverter simply becomes intermittent in operation so it is swapped out.
The simple solution to preventing the problem is not switch off the mains supply during daylight hours especially if the sun is strong. Mains supply failure however will continue to be a problem.
If it becomes necessary to work on an Enecsys installation I suggest the PV panel is covered with a large sheet of cardboard prior to disconnecting the mains supply.
Enecsys obviously discovered this problem near the end of their term and they changed the 250V Zener to 300V plus increased R167 to 120 Ohm.
My schematic was drawn by inspecting the layout of an Enecsys main PCB. No proprietary information was used in this investigation and I have no connection with Enecsys or any of its employees.
My interest stems from having several of these things on my roof and I am keen for them to last a good while longer.
The main source of data for this investigation comes from a friend who has 50 Enecsys inverters on his roof and for logistical reasons the mains supply was not connected for some time and consequently a large number of the units died in the manner described above.
The Enecsys micro inverters use a DC to DC boost converter to provide the approximately 550V DC required for the output stage. Under normal operating conditions when the units are generating and the sun is shining the operating voltage is close to 550V and all is well.
The attached snip from the schematic I have drawn shows a simple bridge rectifier feeding the main bank of storage capacitors prior to the high voltage HF switching stage. Enecsys have sensibly provided two series connected high voltage transient suppression zener diodes connected across the main DC lines via R167 and D31. The combined breakdown voltage for the network is 550V so all well (just).
Now suppose for some reason the mains fails but the sun continues to shine? The load on the 550V rail drops to zero and the DC potential rises to around 570V. D59 and D60 begin to conduct and the dissipation in D60 raises the temperature beyond t max and D60 goes short circuit. A few milliseconds later D59 follows in the same way and this puts a huge voltage across R167 which quickly goes open circuit.
The momentary loss of the DC HV rail is detected by the 8 bit PIC microcontroller and the unit shuts down.
Eventually the owner power cycles the unit and there is a good chance that if both mains supply and PV are restored the unit will continue to function again and all will appear normal. However, the potential across the damaged R167 eventually breaks down the tiny break in the conductive material which momentarily restores the low resistance load across the HV and triggers the unit back into a shutdown.
From the users point of view the inverter simply becomes intermittent in operation so it is swapped out.
The simple solution to preventing the problem is not switch off the mains supply during daylight hours especially if the sun is strong. Mains supply failure however will continue to be a problem.
If it becomes necessary to work on an Enecsys installation I suggest the PV panel is covered with a large sheet of cardboard prior to disconnecting the mains supply.
Enecsys obviously discovered this problem near the end of their term and they changed the 250V Zener to 300V plus increased R167 to 120 Ohm.
My schematic was drawn by inspecting the layout of an Enecsys main PCB. No proprietary information was used in this investigation and I have no connection with Enecsys or any of its employees.
My interest stems from having several of these things on my roof and I am keen for them to last a good while longer.
The main source of data for this investigation comes from a friend who has 50 Enecsys inverters on his roof and for logistical reasons the mains supply was not connected for some time and consequently a large number of the units died in the manner described above.
