Current Limiting for a Cheap SMPS

[rampra]

Hello All,

I am a Mechanical Engg. and I have very little idea about circuits. But I have a hired an electronics guy who can help me with the electronics portion of my project. He is not an Engineer so he cant really "think". He can do what is told and differ between safe and dangerous.

Well I have a cheap Chinese SMPS (notebook style) which has 24V and 10A output. Everytime I use to for my application which is as good as short circuiting (brush electroplating), the SMPS fuses and then we have to switch OFF the SMPS and wait for capacitor to discharge and then switch it ON again (I guess this is a safety feature).

I guessed, what i needed is a constant current device and SMPS is a constant voltage device maybe.

I have 2 questions for all the knowledgeable members:

1. If i add a simple limiting resistor like 1K pot or resistor in series to the output of the SMPS will I be able to use it for my application?
2. Is is possible to add a current limiting circuit outside the SMPS to regulate the current?

I used the "Search function" but found no queries on this topic. Related queries were either too complicated or not concerned to me. If there is any thread on this topic please feel free to guide me.

Thanks!

[Siwastaja]

SMPS can be either voltage (CV) or current (CC) source, or both (CC/CV). You need a current source. Most random cheap supplies are voltage sources, and even though they could implement the constant-current mode as well (being CC/CV), they often implement a latching or "hickup" protection mode instead, meaning that while they can and do limit the current just as you need them to, they only do it for very short bursts, and, depending on device, either perform these short bursts with low duty cycle, or turn off requiring a total restart. This is hard or almost impossible to "mod".

"Lab supplies" are CC/CV, but they tend to be a bit expensive for such a simple job.

LED power supplies are CC, and may work for your purpose - if you find one for a single low-voltage led. They may have protections for too low output voltage, however, so that they think the LEDs are shorted and turn off.

Adding a resistor after the voltage supply kind of works, and regulates the better the more voltage difference over the resistor. On the other hand, the efficiency is poorer, the more there is difference.

With 24V, 10A, and 0.5V electroplating output, you'll burn 23.5V @ 10A = 235W in the resistor, and 5W doing the actual electroplating! At 2% efficiency, this is a non-starter.

If you can find a very low voltage supply, like a high current 3.3V supply, you'll get the efficiency up to 10-20% range. Still a lot of heat to dissipate in the resistor, but possibly workable. You need a huge, fairly expensive resistor. You can sometimes make substitutions, like using large steel sheets, possibly submerged in water...

One practical solution to get you started, if you want to play with Chinese Ebay modules, is to try a CC/CV mode buck module from there. They do exist, pick one with an explicit current limit, meaning a separate potentiometer for current. You can use your 24V brick or similar (such as a 19V laptop power brick) as an input for such a module. Using a safe, existing mains supply makes life so much easier.

I mean something like this: https://www.ebay.com/itm/DC-300W-20A-CC-CV-Constant-Current-Adjustable-Step-Down-Converter-Voltage-Buck/401342111666?hash=item5d71da9fb2:g:nscAAOSwSeVaL7gG  , but do note that these tend to be of low quality, and with a minimum output voltage requirements around 0.8 or 1.2V, they may still need a bit of resistance in the output to work in electroplating. In any case, such step-down (buck) module multiplies the current manyfolds and easily offers >50% efficiency.

[rampra]

Thanks for replying.

I have this type of converter with CC/CV mode. Can you please explain how to provide this resistance?

they may still need a bit of resistance in the output to work

Please consider that my application is actually equal to short circuiting. The process gives results with welding machine but is definetely an overkill and doesnt provide great results.

Regards!

[David Hess]

A power resistor in series with the output is the easiest solution but there is something even better.  Get some big 12 volt incandescent automotive lamps, use them in series pairs for 24 volts, and put as many pairs in parallel to get a high enough current rating without tripping the power supply.

Use these instead of a power resistor to limit the current.  The advantages are that they will have no problem with the power and the non-linear resistance of the tungsten filaments means that at lower output currents, the resistance is also lower and more current is available than if a fixed resistor was used.

[Siwastaja]

By all means use incandescent lamps, but at least find some lower voltage supplies first, 12 or 5V supplies are even more ubiquitous than the 24V supplies, and 10% efficiency (with a 5V supply) makes a lot of sense over 2%... (Assuming 0.5V output voltage, which is approximately what I saw when doing electroplating (through hole copper plating).)

Using resistors (or lamps - a big 12V bulb is fine for current-limiting 5V) in series to limit current, allows you to easily parallel multiple 5V supplies as well so you don't need to try to find a single massive supply. Use one lamp/resistor per supply, and connect them in parallel after that. If you do the paralleling right at the electroplating mount site, this allows you to use distributed cabling with smaller gauge cables, as well, making the cabling more flexible and, again, cheaper and easier to source.

Note that when wasting power in the resistors, a 5V 10A supply is as good as the 24V 10A supply, but it's much smaller and a lot cheaper, and easier to find due to lower power.

[rampra]

Dear Friends,

So I changed my power supply to (12V, 25 AMP) as suggested by esteemed friends and I was "quite" succcessful.

Well now the problem is, the thing works for quite a good time but then it again fuses. Which means somewhere I need to limit the current to some upper value.

I believe disabling the short circuit protection is not an option as that is a recipe to burn the power supply. So I ask forum members to direct me to some kind of simple/moderate current limiter method. whereby the current is limited to some fixed value say 24A.

Regards!

[Whales]

I have this type of converter with CC/CV mode. Can you please explain how to provide this resistance?

No.  If you have a power supply that can do CC then you don't need any resistance.



If you have a CV (constant voltage) power supply: adding resistance is a cheap way of limiting current.  But it's power inefficient, so you need big resistors.

CV power supplies typically disable themselves when they hit over-current conditions, as you have discovered.

If you have a CC (constant current) or CC/CV power supply: you won't need any resistors, you can use the power supply directly.  If a CC/CV power supply is disabling itself then something else is wrong (faulty unit, using it beyond specifications, etc) or it's not actually a CC capable unit.



Sidenote: if a power supply says that it is "12V 24A" then that does NOT mean it is CC/CV.  Unless it says otherwise it is probably only CV.  The "24A" is a max before it self-disables.  It does not actually "limit" the current proactively, it simply fails if you try and go over.

A real CC/CV power supply will limit itself and never self-disable.  These power supplies typically have two knobs/controls on the front: one for voltage, one for current.

So I ask forum members to direct me to some kind of simple/moderate current limiter method. whereby the current is limited to some fixed value say 24A.

24A at 12V is a lot of power (300W).  There are no "simple" ways of creating a proper CC circuit at such a high power level, the cheapest and easiest option is to use lightbulbs in series.

You will have to vary the amount of bulbs and their wiring, depending on the bulb types and what you want.  Automotive bulbs are one option (high resistance), mains voltage bulbs are another (low resistance).  In most cases the bulbs will not light fully, unless you have a short.

[rampra]

Thank you very much!

I had a CC/CV buck boost convertor but that didnt make through as its max current was 10A.

Secondly my current power supply is CV.

Now as you are saying that at 12V 24A, its difficult to built a Current Limiter.... I have run short on ideas.

Can you please let me know that this can be used? (Ofcourse with modifications to Transistor)
https://maker.pro/custom/tutorial/how-to-make-constant-current-source-power-supply-circuit

Regards!

[Whales]

You would potentially be dissipating hundreds of watts in that final transistor.  In a standard size transistor die that's well into nuclear reactor power densities.

You would have to instead use multiple transistors in parallel AND try to solve the thermal runaway issues that parallel transistors suffer.  Ballast resistances, all sorts of problems.  Possibly fun due to the explosions, but not easy, and I'd probably still end up using lightbulbs as the ballasts for each transistor.

[rampra]

Thank you for reply!

Well I have come up with this solution.

I will try to use my power supply by connecting it with some High current motor driver module (ibt -2 H bridge). In this H bridge I will be using only 1 side i.e. forward or reverse. I will use an arduino and use its PWM function with a pot.

I know PWM is not equal to Current Limiting, but i hope to prevent the short circuit protection if I use it at 80% duty cycle.

I humbly accept your suggestion of automotive bulbs etc. but somehow I cant use a bulb for no reason.

Kindly let me know if this atleast sound reasonable and doable theoretically. I am sure this motor driver module is called something else in an expert's terminology. But currently I am only interested in getting it worked than bother about terminologies.

[Whales]

That could potentially work. 

You're dealing with power levels high enough to make things explode, so take precautions.