High voltage transformer design
Suppose I have a planar PCB transformer. The primary and secondary coils are printed on different PCBs. The core is inserted through these PCBs to link the two coils. The primary is at a low voltage and the secondary could be sitting at some 10 kV for example. Also, assume that the chosen PCB dielectric material has a high dielectric breakdown voltage capable of withstanding much higher voltages. The creepage and clearance requirements are also assumed to be met. I have the following questions:
- When should a transformer core be potted? Should it be potted at all? (the size of the planar transformer is less than 3 cm)
- Can I place the high voltage conductors of the secondary coil on the inner layers considering the high dielectric strength of the PCB.
- Air has a breakdown voltage of 3000 V/mm. If I wish to prevent arcing because of the 10 kV conductor, I can keep some 4 mm spacing between various components and the core. But, IPC 2221 specifies a much higher clearance requirement for the same. Is this to ensure reliability in various environmental conditions?
- "The core is at a floating potential that could be anywhere between 0 and 10 kV" - Is this correct?
EDIT: The secondary voltage is low (such as 5 V or 10 V) but with respect to a 10 kV reference.
2 answers
You are accessing this answer with a direct link, so it's being shown above all other answers regardless of its score. You can return to the normal view.
The following users marked this post as Works for me:
| User | Comment | Date |
|---|---|---|
| jonathan_the_seagull | (no comment) | Jan 20, 2024 at 14:16 |
When should a transformer core be potted?
Whenever you need some of
- Mechanical ruggedness.
- High dielectric strength.
- Minimize vibrations (and therefore audible whine) from capacitive, inductive, piezo-electric and magneto-restrictive effects.
Can I place the high voltage conductors of the secondary coil on the inner layers considering the high dielectric strength of the PCB.
Since you are using two separate PCBs for the primary and secondary, this doesn't really accomplish anything. You'll need at least one via, so there will be a point insulated only by the solder mask. You could us a blind via, but that's more complicated, and cuts down on the layers you can use for the winding.
Air has a breakdown voltage of 3000 V/mm.
Not really. There is no single figure for air breakdown voltage because there is no single spec for "air". Variations in humidity make a large difference, and pressure also to a lesser extent. 1 kV/mm is a more reasonable figure to use in the absence of a particular spec or standard. Creepage distance also matters since dirt and moisture can form a conductive path.
If this is something you plan to sell, then you need to look into the regulatory requirements and what kind of certification the customers or resellers will demand. There are extra-stringent requirement for some applications, like medical. Each standard will specify the minimum clearance and creepage distances for a particular voltage.
It's not clear whether both windings are low voltage with a high common mode difference, or if the secondary is a high voltage winding. If the latter, you probably can't do this with PCB coils. You won't be able get the turns ratio to go from something considered a "low" voltage to 10 kV. You might be able to use a PCB coil for the primary, then many turns of fine wire for the secondary. Either way, the power transfer will be "small".
1 comment thread
When should a transformer core be potted? Should it be potted at all?
For this type of job, you would pot the transformer when you can't obtain the necessary creepage and clearance between the secondary (raised externally to 10 kV) and the low-voltage parts of the transformer (including the ferrite core).
$$$$
The "typical" dielectric strength for 1 mm thickness of potting resin is 20 kV but, you need to verify this with the materials you want to use and, you need to ensure that your potting process does not produce internal bubbles or air-locks when setting. The bubbles/air-locks can expose high-voltage circuits to low-voltage circuits via a less-than-adequate creepage path. A "creepage" path is a much lower breakdown voltage (compared to the solid resin) because, it is a "surface path".
$$$$
Can I place the high voltage conductors of the secondary coil on the inner layers considering the high dielectric strength of the PCB
It makes no difference; you should regard all your secondary layers and windings as one block of metal raised to 10 kV.
$$$$
IPC 2221 specifies a much higher clearance requirement for the same. Is this to ensure reliability in various environmental conditions?
It certainly is.
$$$$
"The core is at a floating potential that could be anywhere between 0 and 10 kV" - Is this correct?
This is your call. If you say it is then, that's what it is.
0 comment threads
1 comment thread