# How does a power converter regulate voltage effectively at zero load current?

#### SiCEngineer

Joined May 22, 2019
320
I am trying to get my head around how power converters regulate the output voltage at no-load conditions. My output pulses from substantially full load ~200mA to substantially no load 0A at frequencies anywhere from 100Hz to 100kHz.

My question is quite clear already but, how exactly does a converter keep output voltage constant if output current and therefore power are zero? If P = VI, P and I are both zero but this does not mean V is zero. I get this. But by V = IR, if current is zero then voltage should also be zero. But we intend to regulate it to a certain value.

I understand there will be some current draw by the multipliers, and I have heard some people say we must place a bleeder resistor on the output to ensure a minimum load current.

But it is to my understanding bleeder resistors drain the charge from the capacitor. I do not understand why this is useful since we want to maintain the capacitor voltage, not drain it. If this is indeed the correct approach, is there a way to calculate the value to ensure a good level of Power dissipation.

I assume the answer is quite simple but Google searches come up with all sorts of things about burst mode control, etc which is complicated. I'd like to know first the simple principle. My topology is a buck-current fed push-pull converter, if this helps, but also interested in fly-back converter.

#### ci139

Joined Jul 11, 2016
1,696
from where such question
...
...
...
theoretically - if there's no consumption - you switch your converter off or to low-power/standby

#### SiCEngineer

Joined May 22, 2019
320
Okay, I get that. But if your load is pulsed regularly, how effective would it be to turn the entire supply of at high frequency? What if you turn the load off, but the output voltage drops too much?

#### ci139

Joined Jul 11, 2016
1,696
from where such question
what is that you want to achive . . . for example - you can use multiple converters - swapping them (electronically) with respect your varying demands

#### SiCEngineer

Joined May 22, 2019
320
what is that you want to achive . . . for example - you can use multiple converters - swapping them (electronically) with respect your varying demands
Sorry, I'm not sure what you mean. What would be the point of swapping to a different converter during the no load condition? And again, would you be able (realistically) be able to switch electronically between two converteres at such speeds?

My question still wouldn't be answered - how would the auxilliary converter be able to more effectively regulate the voltage, the current and power delviered will still be zero.

What I want to achieve is to regulate the output voltage as tightly as possible while the output is pulsed from full to substantially no load at varying frequency.

Last edited:

#### Alec_t

Joined Sep 17, 2013
12,201
Think of the converter as having two loads; an internal one drawing a small current used to do the house-keeping task of regulating the output voltage, and an external load which the user applies. A well-designed converter should have a transient response which allows for rapid changes in the applied external load.

#### WBahn

Joined Mar 31, 2012
26,398
I am trying to get my head around how power converters regulate the output voltage at no-load conditions. My output pulses from substantially full load ~200mA to substantially no load 0A at frequencies anywhere from 100Hz to 100kHz.

My question is quite clear already but, how exactly does a converter keep output voltage constant if output current and therefore power are zero? If P = VI, P and I are both zero but this does not mean V is zero. I get this. But by V = IR, if current is zero then voltage should also be zero. But we intend to regulate it to a certain value.

I understand there will be some current draw by the multipliers, and I have heard some people say we must place a bleeder resistor on the output to ensure a minimum load current.

But it is to my understanding bleeder resistors drain the charge from the capacitor. I do not understand why this is useful since we want to maintain the capacitor voltage, not drain it. If this is indeed the correct approach, is there a way to calculate the value to ensure a good level of Power dissipation.

I assume the answer is quite simple but Google searches come up with all sorts of things about burst mode control, etc which is complicated. I'd like to know first the simple principle. My topology is a buck-current fed push-pull converter, if this helps, but also interested in fly-back converter.
If you have no load, then R=∞, and so with I=0, I·R is indeterminate and can equal anything.

Think of a battery with no load on it. That does a pretty good job of regulating the output voltage.

Some voltage regulator circuits require a minimum output current in order to regulate properly because of the nature of how that regulator circuit operates. It is not something that is fundamental.

#### WBahn

Joined Mar 31, 2012
26,398
Sorry, I'm not sure what you mean. What would be the point of swapping to a different converter during the no load condition? And again, would you be able (realistically) be able to switch electronically between two converteres at such speeds?

My question still wouldn't be answered - how would the auxilliary converter be able to more effectively regulate the voltage, the current and power delviered will still be zero.

What I want to achieve is to regulate the output voltage as tightly as possible while the output is pulsed from full to substantially no load at varying frequency.
Sorry, but "as tightly as possible" is a useless and meaningless specification. What degree of regulation do you NEED? Beyond some point, factors such as WHERE you need it regulated become important. What voltage are you talking about? 5 V? 5,000 V?

#### puppydog846

Joined Apr 5, 2020
11
If you have no load, then R=∞, and so with I=0, I·R is indeterminate and can equal anything.

Think of a battery with no load on it. That does a pretty good job of regulating the output voltage.

Some voltage regulator circuits require a minimum output current in order to regulate properly because of the nature of how that regulator circuit operates. It is not something that is fundamental.
it requested the out via the feedback circuit. this circuit sense the output and produces a error voltage which will produce a short pulse of energy to the output. Under low or no load conditions this pluse will be more than is required and the output cap will increase in voltage if the voltage is over the set point the elthe error signal will go to zero. i believe the 100 hz is the time for the output to decrease enough to turn on the pwm bob

#### WBahn

Joined Mar 31, 2012
26,398
it requested the out via the feedback circuit. this circuit sense the output and produces a error voltage which will produce a short pulse of energy to the output. Under low or no load conditions this pluse will be more than is required and the output cap will increase in voltage if the voltage is over the set point the elthe error signal will go to zero. i believe the 100 hz is the time for the output to decrease enough to turn on the pwm bob
And since you are designing this, you have control over several variables, including the size of each pulse and the size of the capacitor and how much error signal triggers a pulse, so that you can achieve the amount of regulation that is needed.

Again, how much regulation is NEEDED?

#### SiCEngineer

Joined May 22, 2019
320
Hi,
And since you are designing this, you have control over several variables, including the size of each pulse and the size of the capacitor and how much error signal triggers a pulse, so that you can achieve the amount of regulation that is needed.

Again, how much regulation is NEEDED?
Apologies for being obscure. The regulation during pulsing is 20Vpk-pk on a -6kV output. It is relaxed quite a lot compared to steady state conditions.

#### SiCEngineer

Joined May 22, 2019
320
Think of the converter as having two loads; an internal one drawing a small current used to do the house-keeping task of regulating the output voltage, and an external load which the user applies. A well-designed converter should have a transient response which allows for rapid changes in the applied external load.
The applied external load is pulsed on and off at frequencies up to 100kHz. How exactly would a control system allow for this kind of rapid change? I assume additional external circuitry would instead be required?

#### ci139

Joined Jul 11, 2016
1,696
i have designed an experimental 500kHz bandwidth active feedback chain for the mc63034

you need to clear out your circuit - as what and when to regulate

there are likely many circuit techniques to achieve your goal - once you are able to clearly define one

#### SiCEngineer

Joined May 22, 2019
320
i have designed an experimental 500kHz bandwidth active feedback chain for the mc63034

you need to clear out your circuit - as what and when to regulate

there are likely many circuit techniques to achieve your goal - once you are able to clearly define one
My output voltage is a high, at 6kV and 3kV with 10mA and 200mA respectively, but due to the pulsing the maximum average output is 330W. The minimum is 3.3W at 1% duty cycle, 100kHz pulsing frequency. I have a buck current fed push pull converter which with average current control mode can accurately track the reference voltage and a second stage LC filter provides very low output ripple. The issue I am currently having is regulating the load when it transitions to no load state, at a varying frequency and duty cycle. I was wondering there may be some way of using active switches to improve the no load regulation of the push-pull/buck stages?

#### puppydog846

Joined Apr 5, 2020
11
now we are getting some where I think you need to increase the frequency response of the loop so it responses faster and the average current mode may also reduce the response

#### SiCEngineer

Joined May 22, 2019
320
now we are getting some where I think you need to increase the frequency response of the loop so it responses faster and the average current mode may also reduce the response
How do I increase the frequency response of the loop?

#### MrAl

Joined Jun 17, 2014
8,475
I am trying to get my head around how power converters regulate the output voltage at no-load conditions. My output pulses from substantially full load ~200mA to substantially no load 0A at frequencies anywhere from 100Hz to 100kHz.

My question is quite clear already but, how exactly does a converter keep output voltage constant if output current and therefore power are zero? If P = VI, P and I are both zero but this does not mean V is zero. I get this. But by V = IR, if current is zero then voltage should also be zero. But we intend to regulate it to a certain value.

I understand there will be some current draw by the multipliers, and I have heard some people say we must place a bleeder resistor on the output to ensure a minimum load current.

But it is to my understanding bleeder resistors drain the charge from the capacitor. I do not understand why this is useful since we want to maintain the capacitor voltage, not drain it. If this is indeed the correct approach, is there a way to calculate the value to ensure a good level of Power dissipation.

I assume the answer is quite simple but Google searches come up with all sorts of things about burst mode control, etc which is complicated. I'd like to know first the simple principle. My topology is a buck-current fed push-pull converter, if this helps, but also interested in fly-back converter.
It depends on the controller IC or the controller scheme in general.
Some converters can not output a zero width pulse so there has to be some load on the output.
Also, there really is no ideal plant there is always some loss. The inductor has ESR, the cap has ESR, that may be enough to eat up the minimum power that the converter outputs. If it is not, then a minimum load must be applied.

The opposite is also true. Some converters have a minimum dead time so they can not put out the maximum power that would be available from the input if there could have been zero dead time.

#### SiCEngineer

Joined May 22, 2019
320
It depends on the controller IC or the controller scheme in general.
Some converters can not output a zero width pulse so there has to be some load on the output.
Also, there really is no ideal plant there is always some loss. The inductor has ESR, the cap has ESR, that may be enough to eat up the minimum power that the converter outputs. If it is not, then a minimum load must be applied.

The opposite is also true. Some converters have a minimum dead time so they can not put out the maximum power that would be available from the input if there could have been zero dead time.
Okay, brilliant. My new question then, is what is the most efficient way of applying a "Load" to the output, in the case of my application there is absolutely no load current consumption during no load transient. Therefore I assume we must have, I don't know, a bleeder resistor or something?

#### ci139

Joined Jul 11, 2016
1,696
I have a buck current fed push pull converter which with average current control mode
LC filter
The minimum is 3.3W at 1% duty cycle, 100kHz pulsing frequency
? your ?load? has 1% duty at 100kHz = 100ns ON (event) ? then what's your push-pull controller frequency ???? (i assume i miss got that)
__________
• what is the minimum required response time for that 10mA ?load current? . . . for 200mA ?load current?
• how accurate the output voltage/current have (one and another) . . . to be
• what are the allowed over-/undershoot at transient events
• ? how fast the transitions in between the 3/6kV - ? is the load connected while you swap your voltages
. . .
• ? the feedback chain you have now (schematic) including the half bridge , (the) filter , (the) load and the pin(s) and the chip you feedback it to (if you allowed to share)

Last edited:

#### ci139

Joined Jul 11, 2016
1,696
? did you managed it already on your own . . . the HV and current feedback are not my strongest sides but ... i've done various exotic feedbacks - so probably i'm able to suggest something if i see the schematic layout