Hi,
Is there any reason why you decided to use current measurement transformer to measure coupling inductor current? Is there any problem to use a shunt resistor method (apart from extra galvanic isolated power supply domain) or hall sensor for this?
The current measurement transformer transfers signal from 2kHz to 1MHz, it is any reason to blocking DC?
Regards,
Grzegorz
Hi Grzegorz,
Thank you for the query. There is absolutely no problem with the shunt based measurement system and an isolated amplifier based signal chain. Using TI's AMC3301/02 series of devices should be sufficient to run and update control loop variables every switching cycle. I understand from our previous interactions you are using UCC21732 gate driver for short circuit protection, so using shunt/hall sensor for control loop is definitely better than CT. In fact a better option over CT would be to use Rogowski coil based measurement. This would give you the fastest possible transient response with good accuracy. The implementation and schematicof Rogowski coil for transformer tank fast switching AC current is mentioned in page 16 of this TI design link http://www.ti.com/tool/TIDM-02002 and http://www.ti.com/lit/ug/tidueg2c/tidueg2c.pdf?&ts=1589547371599. Secondly, the Rogowski and the CT used here pass only high frequency AC signals to the measurement chain.
Next on the DC blocking capacitor, we ideally need to have them in the primry and secondary side of transformer to prevent DC offset settling in and saturating the transformer over time. We need to have them on both sides of transformer in case we operate them in a bidirectional way. These capacitors will be needed for doing average current control and will typically not be used in peak current control.
Hi Gregorz,
Thank you for the query.
You are correct that MLCCs can be used in place of film capacitors as can be seen in this table beow:
MLCCs only suffer the following disadvantages over Film caps:
1. Low capacitance per unit volume, especially with C0G dielectric
2. DC bias instability and derating for higher voltage operation.
Hence it would be better to use a combination of both to get better overall capacitance per unit volume, handling short term transient voltages, temperatures etc.
Thank you
