# How to achieve very low output ripple with High voltage outputs

#### SiCEngineer

Joined May 22, 2019
299
For a PhD project I am designing a high voltage power supply which has a maximum operating voltage of 5kVDC. The output voltage ripple is very important for the converter. The maximum allowable output voltage ripple is 200mV peak-to-peak.

I would like to know peoples thoughts on how this could be achieved. Considering the high output voltage it is undesirable to have an LC filter, my only other idea would be to use an RC filter. But to get the attenuation I need I will either need a high value capacitor (which is limited due to the HV output) or a high value resistor (which will lead to high losses in the converter).

Any other suggestions that I could look into would be appreciated.

#### Alec_t

Joined Sep 17, 2013
11,981
How much current will be drawn from the 5kV supply?
How are you deriving the 5kV?

#### SiCEngineer

Joined May 22, 2019
299
How much current will be drawn from the 5kV supply?
How are you deriving the 5kV?
5kV is derived through an LCC converter feeding a step-up transformer and full bridge multiplier. 12mA current drawn from the 5kV supply. There is another output that draws more current however the ripple is not important on this output.

#### Alec_t

Joined Sep 17, 2013
11,981
What is the ~switching frequency of the converter and the output capacitor value?

#### SiCEngineer

Joined May 22, 2019
299

https://www.edn.com/design/analog/4326853/High-voltage-low-noise-dc-dc-converters

Edit: Something I did along those lines.

View attachment 183774
This is really interesting. So the resistive divider reduces the HV output to a value that can be handled by lower voltage analogue components. However, what will the voltage be across C4?

In the note you sent, the voltages are in the range of say 400V. Will this be applicable to outputs which are like 5kV?

#### DickCappels

Joined Aug 21, 2008
7,155
In the actual application the voltage across C4 was 27 KV. The output stage of the amplifier were some bipolar transistors from Sanyo with Vcb of about 1800 V (that was the hard part).

The amplitude of the signal depends on the ripple the circuit is cancelling, and that is determined by many factors.