Current limiting in a CC/CV power supply

[Braeden Hamson]

I've been looking all over for a schematic for a good power supply, but I'm not having any luck finding one that doesn't use a voltage regulator IC. Those are inefficient and I'd like to avoid them. People on this forum as well as elsewhere recommend a buck/boost converter for the voltage portion.

However I haven't been able to nail down the current limiting section. A quick aside, as I understand it a CC/CV power supply is has a CC section and a CV section, they're never both running. One gets selected when by a controller when the current exceeds a set limit or when the voltage exceeds a set limit. And they're selected with a kind of switch. Back to the main topic. Again I can't find a current limiter that doesn't use a regulator IC. Whats a good way to regulate it? Can I just use another buck converter? Or am I wrong in my understanding that a current limiter is a voltage regulating type circuit in which a controller measures the current and adjusts the voltage accordingly.

Wonder if I could get my hands on a Keithley PS diagram.

 

[kellys_eye]

All regulated power supplies are 'constant voltage' - that's what they are meant to do, regulate the voltage. The output voltage is maintained steady (constant) despite variations of the input.

The current output is as much as the power supply can deliver unless you design a limiter as part of it, whereby the output voltage drops to maintain the set i.e. constant current (assuming you are exceeding the limit you preset).

I've been looking all over for a schematic for a good power supply

What, by your definition, is a 'good power supply' then?

How you specify your requirements usually sets how you implement the solution and there are many ways to 'skin that cat'.

The higher the spec you set (of course) the more complex the solution and the more costly too.

I have many, many established design schematics for power supplies - from the simplest voltage and/or current limiting versions right up to 'laboratory quality' power supplies with minimal ripple, fast slew rate, variable V/I (to mV and mA) and programmability (if required).

Quite often it is more cost effective to obtain a second hand PSU than build one from scratch though.

 

[Braeden Hamson]

Ahh of course you'd just reduce the voltage of the buck/boost converter for constant current mode. Also I'm looking for efficiency as top priority, followed by a somewhat high power output. Though I don't have a specific number in mind. I don't have a variable power supply so I figured it'd be a good project to make one. Now to figure out the control section.

 

[kellys_eye]

Efficiency is often a trade-off of other specifications. SMPS technology is the obvious for high efficiency but they require careful filtering of the output and have (usually) a poor response to transient requests i.e. have difficulty tracking fast-changing loads and poor ripple unless VERY well designed.... i.e. expensive and complicated.

Linear power supplies are laboratory favourites for their low ripple, low noise, fast response etc and you have to weigh up your 'real' needs for use. Will you really use (and quite often) the full 10 amps of a 'high power' version - or will your main use be 0-2A for most normal uses?

 

[AnalogKid]

All regulated power supplies are 'constant voltage'

Except for the ones that are not.

ak

 

[kellys_eye]

All GOOD regulated power supplies are 'constant voltage'

...when operated within their current limits.

 

[(*steve*)]

Many commercial power supplies are linear, but switch taps on the mains transformer to maintain a sufficient, but not too high, voltage overhead to the linear regulator.

Why is efficiency your first priority?

Have you considered noise or transient response? What do you consider "somewhat high power?"

Switch mode power supplies may generate too much noise for audio applications, but may be fine for many other applications.

More complex power supplies may combine switch mode and linear elements to gain the best of both worlds.

You can probably find some Keithley schematics, but it wouldn't be a good beginner project, and it would likely use voltage regulator ics (even if just for it's internal per supply)

 

[Braeden Hamson]

I didn't specify any specific requirements because I'm not looking for specifics yet. Though I realize those are important. I also don't know what I consider high power because I don't know what's achievable. I want efficiency because I thought it would be interesting to have a portable battery powered power supply. This is one of those cases of make it and find a use for it . And for specs 5A 30V or more just feels about right. Like I said no real requirements. I'm avoiding linear voltage regulators mainly because they don't have the kind of power outputs I want. Also, this is kind of a learning experience more than anything. And linear voltage regulators just aren't that interesting. Internally they are though. The buck/boost seems like what I want.

So my implementation would then be, a buck/boost whose output voltage and current are measured by an arduino and the PWM signal would be adjusted accordingly. I want to run some 7 segment displays and other functions so I figure an arduino isn't overkill. If not a bit of a cop out haha

 

[AnalogKid]

All GOOD regulated power supplies are 'constant voltage'
...when operated within their current limits.

Again, no they are not. There are thousands of applications for constant current power sources, including instrumentation, industrial controls, and ... LEDs. Linear Tech and Maxim have dozens of ICs for constant current supplies for LEDs.

ak

 

[(*steve*)]

I didn't specify any specific requirements because I'm not looking for specifics yet. Though I realize those are important. I also don't know what I consider high power because I don't know what's achievable. I want efficiency because I thought it would be interesting to have a portable battery powered power supply. This is one of those cases of make it and find a use for it . And for specs 5A 30V or more just feels about right. Like I said no real requirements. I'm avoiding linear voltage regulators mainly because they don't have the power outputs I want.

Take a look at the most recent video on eevblog (#1030). Although... there's not much construction there.

 

[kellys_eye]

Again, no they are not. There are thousands of applications for constant current power sources, including instrumentation, industrial controls, and ... LEDs. Linear Tech and Maxim have dozens of ICs for constant current supplies for LEDs.

ak

What is your point?

You make these repeated 'contradictions' of my post yet offer nothing to the thread or any explanation of why you think a regulated power supply isn't 'good' when it has a constant output voltage.

I'm not referencing 'constant current' - I'm talking about a REGULATED POWER SUPPLY whereby it is defined as having 'good regulation' by its ability to maintain a fixed output voltage regardless of the change in the input and the variation in output load 'providing it remains within the limits of the power supply'.

So, please explain why you're making 'negative' responses to my posts????

 

[(*steve*)]

It is fairly clear that @AnalogKid is referring to other regulated power supplies that output something other than a regulated voltage.

It's true, and he's also being hopelessly pedantic as was once started on his sig (if I recall correctly).

I would normally assume "constant voltage" if someone said "power supply" without additional description or context.

 

[twister]

Do a search for a data sheet for a tl494. It has a schematic for a switching power supply. I built a solar panel pulse width modulated voltage regulator using this IC. It has two inputs, one for voltage and one for current.

 

[Braeden Hamson]

Do a search for a data sheet for a tl494. It has a schematic for a switching power supply. I built a solar panel pulse width modulated voltage regulator using this IC. It has two inputs, one for voltage and one for current.

Now that's an interesting IC. So does it automatically adjust the PWM signal to the buck/boost circuit with changing loads?

 

[twister]

Now that's an interesting IC. So does it automatically adjust the PWM signal to the buck/boost circuit with changing loads?

Y es, that's what it does. In my circuit, I didn't use the buck boost, I just hooked it straight to the battery and the battery acts as a huge capacitor. It does create noise in radios but doesn't affect my TV or computer. The MOSFETs barely even get warm. If you want constant current, maybe just leave the voltage input open. I have tried using inductors and caps and using different frequencys to eliminate the radio interference but haven't had much luck.
Maybe using the buck, boost would eliminate the noise?
.

 

[Terry01]

I would normally assume "constant voltage" if someone said "power supply" without additional description or context.

I would assume the same Steve and k_e. My bench "power supply" is "constant voltage" probably the same as most unless it is for powering something in particular. There is always "the exception"