Hello Osama Arafa ,

FYI, using aluminum foil does not really work due to the presence of a durable but invisible oxide layer, Al2O3 (sapphire), on all exterior surfaces of the foil. A better alternative is copper foil backed with a conductive adhesive, check the EMC buyers guide, or copper braid. Am I correct in thinking that there are no sub-enclosures inside the main enclosure? Your photo of the complete system in the enclosure with the enclosure door open gives the impression that everything is inside the enclosure but without sub-enclosures. You may need sub-enclosures together with feedthrough capacitors, ferrite beads or cores, EMC filters, etc. for any wires, or cables penetrating the sub-enclosures. Even though you said that each subsysem had been tested individually, you did not indicate an EMI sniffer was utilized to check which circuits produce electrical noise; a crude EMI sniffer would be small battery powered AM radio tuned to a quiet region of the tuning dial. Or have you already done that?

I recommend the following books.

[1] László Tihanyi; Electromagnetic Compatibility in Power Electronics; IEEE Press; 1995.

[2] Tim Williams; EMC for Product Designers, 3rd Ed.; Newnes; 2001.

Regards,

Tom Cuff

Thanks Tom, I appreciate your support.I will study your recommendations in detail and I will try to give more clarifications on your inquiries.

You have to break every possible common mode current path in your system.

It makes sense that each unit you have works properly on it its own. This eliminates any design errors.

First make sure that you are feeding your system from isolated power supplies.

Second check your gate driver design for all switches and confirm that the isolation stage works properly

Third follow the signal from the microcontroller to the switch to trace the source of the voltage overshoot contamination in order to use a common mode choke to clean it up.

I hope this helps!

Good Luck

Thanks Ahmed for your valuable recommendations

Thomas Cuff yes Tomas, for this question (Am I correct in thinking that there are no sub-enclosures inside the main enclosure?) we don't use sub enclosures except for the two switch mode power supplies for the +/- 15, 5V for sensors feeding and control voltage (galvanically isolated at output side). For the EMI sniffer, none has been utilized until now. We are considering this important step based on your advice, but we could only identify the noise characteristics using a regular scope trace. It is running at the same fixed PWM frequency (10 kHz). I wish I could get an easy access to the problem solving from reviewing the mentioned books, thank you

Hi Osama Arafa ,

I reread the text under your question title, and I have a question. In the second paragraph, "However, when the whole system is integrated and placed into the enclosure", you say everything (boosters, inverter, safety interlocks, etc.) is in the enclosure. But in the last paragraph, "I attach here some photos of the inverter cabinet", you refer to the "inverter cabinet" as if only the inverter is housed in the enclosure. Does the enclosure contain everything, except, of course, the PV arrays?

Are the ±15 VDC and +5 VDC switching power supplies, which you mentioned were in sub-enclosures, located in the upper left corner of the "inverter enclosure" photo? The reason I am asking is because these two switching supplies look like they are on bent metal sheets instead of true wrap-around metal sub-enclosures. I have one other question. The ±15 VDC and +5 VDC switching power supplies, are they DC-to-DC converters or are they AC-to-DC converters, i.e., what is the source of their input power?

Do the two booster DC-to-DC converters (switchers), which are shown in your block diagram, also use 10 kHz as the switching frequency?

Here is some added literature that might be of some interest to you:

[1] Mark Waller; Surges, Sags, and Spikes; Prompt Publications; 1992.

[2] The Interference Technology Engineer's Master (ITEM), which is available from the following URL: https://learn.interferencetechnology.com/item-2021/ , has a catalog section for EMC components (EMI/RFI filters, common mode chokes, ferrite cores and beads, etc.) and equipment (EMI sniffers for use with oscilloscopes and spectrum analyzers).

Regards,

Tom

Hi @Thomas Cuff,

1) "Does the enclosure contain everything, except, of course, the PV arrays?" Yes, Tom, the inverter is the generic equipment name of all these components altogether!

2) "Are the ±15 VDC and +5 VDC switching power supplies, which you mentioned in sub-enclosures, located in the upper left corner of the "inverter enclosure" photo?" Yes they are! and yes their enclosure is a bent aluminum sheet that is surrounding the SM power supply from all directions but it has vents for cooling. These aluminum enclosures are firmly tied to the Cabinet chassis.

3) "The ±15 VDC and +5 VDC switching power supplies, are they DC-to-DC converters or are they AC-to-DC converters, i.e., what is the source of their input power?".. These are ready-made switch mode power supplies. They receive AC 220 Vrms input from one of the grid phases.(line to neutral voltage). They are internally equipped with a front-end diode rectifier and I think they are then based on a flyback DC-DC converter.

4) "Do the two booster DC-to-DC converters (switchers), which are shown in your block diagram, also use 10 kHz as the switching frequency?" yes both are switched at 10 kHz

Thanks, Tom for your support and follow-up

Hi Osama Arafa ,

Thanks for the clarifications. What I am thinking is the following. Place the two boosters, the inverter, and the filter inductors together in their own sub-enclosure. And for each wire or cable penetrating the sub-enclosure provide them with the appropriate EMI/RFI mitigation modality. For example, wires from the PV array can penetrate the sub-enclosure using feedthrough capacitors, which do not impede the DC input power, but present any AC EMI/RFI signal, originating from the electronics inside the sub-enclosure, with a low impedance pathway to ground. The output power phases from the filter inductors to the power grid need to pass through the walls of the sub-enclosure via one bandpass or low pass filter centered around 50-60 Hz per phase. Now, you also have input and output signals from the controller, which need to be conditioned as they pass through the wall of the sub-enclosure. I think that we can agree that the sources of the EMI/RFI noise arise from the two boosters, inverter, and the filter inductors. So it makes sense to isolate them from the control circuitry. Note, because of the heat generated by the circuits inside the aforementioned sub-enclosure, you will need a fine mesh copper screen bonded along its border to the sub-enclosure on the intake and outtake ports for the fans associated with the sub-enclosure; the fans will be mounted on the outside of the sub-enclosure. The lid of the sub-enclosure will need an EMI gasket so that the isolation is not degraded as a result of the lid being removed and replaced, over and over. I am sure you already thought of all this, but it helps to say it explicitly.

Regards,

Tom

Thanks Thomas Cuff , appreciating your support

It sounds like you are experiencing a common problem with high-frequency noise in inverter systems. This noise can be caused by a number of things, including:

• Switching transients: When the power switches in an inverter turn on and off, they can generate high-frequency noise. This noise can be conducted onto the low-signal wires through capacitive coupling.
• Electromagnetic radiation: The high-frequency currents in the power lines can also radiate electromagnetic noise. This noise can be picked up by the low-signal wires through inductive coupling.

The grounded aluminum foils that you have installed around the power conductors and low-signal wires will help to reduce the amount of noise that is conducted and radiated. However, it is likely that you will need to do more to completely eliminate the noise.

Here are a few other things that you can do to reduce the amount of noise in your system:

• Use ferrite beads: Ferrite beads are small, cylindrical devices that can be installed around the power and low-signal wires. They work by absorbing high-frequency noise.
• Use shielded cables: Shielded cables have a metal layer around the conductors that helps to block noise from entering the cable.
• Use a filter: A filter is a device that can be installed between the inverter and the grid. It works by filtering out the high-frequency noise.

If you are still having trouble with noise, you may need to consult with an expert in EMC. They will be able to help you to design a system that is immune to noise. Here are some additional tips for dealing with EMC noise in inverter systems:

• Keep the power and low-signal wires as short as possible. This will help to reduce the amount of noise that is conducted and radiated.
• Route the power and low-signal wires away from each other. This will help to reduce the amount of capacitive and inductive coupling.
• Use a variety of techniques to reduce noise. No single technique will be 100% effective, so it is important to use a combination of techniques.

By following these tips, you can help to reduce the amount of noise in your inverter system and improve its reliability.