Is my chinese PSU broken?

Cyparagon, i'm not such a poor lad, i understand how a PSU works and I know i can leave it shorted.

I guess Apex meant i could do better, using a test load for example, it's much more like a diode and i'll be able to set the exact current without risk of turning too much the potentiometer etc instead of just shorting it.


BY THE WAY:

Is anyone there interested in drop shipping a bench PSU from amazon to Brazil?!!!???

Cyparagon said:

Bench power supplies are designed to be able to drive a short all day long.
Quit feeding the poor lad bullshit.

Click to expand...

They are indeed. On a well designed and constructed lab power supply, you can interconnect the output terminals with a crowbar for all i care, it should not sustain any damage due to that. What it will do is run the set current through the crowbar at a voltage very near zero. It should do that for all enternity, though most linear designs will dissipate a lot of power under those conditions, which could lead them to fail when in a hot and/or poorly ventilated area. Better designs have thermal protection as well, but not all of them do.



In this case it seems there has been some failure with the psu causing it not to withstand this. As you smell something burning, you can either return the psu since it simply was defective (current limitation should have prevented it from damage from a short circuit, no matter how you preset the voltage and current dials).

If sending it back is impractical due to shipping cost/time, i'd suggest opening it up and looking for whatever component burned down. Perhaps its just a sense resistor that is easy to replace, but it could very well be a power transistor that failed. I'm not familiar with the details of this power supply, but if you can put some good pictures of its inside on here i'd be happy to take a look at them.

luccax said:

using a test load for example, it's much more like a diode and i'll be able to set the exact current

Click to expand...

The current limit setting is the current limit setting regardless of the load. Hell, you could set the current limit with a 12V lamp if you wanted. If you set the current limit to 500mA at 0V, You can plug anything else into it, and it will still restrict the current to 500mA. You can use a test load if you want, but you're just wasting your time.

In fact, many PSUs have a starting spike in current mode (even some of the good ones) that poses a bigger threat than "slipping on the pot." The mastech I tested doesn't, however.

There's no reason to be afraid of setting the current on-the-fly assuming you start at zero, but if you're paranoid about it, shorting the output provides the same current set point as using a test load. In fact, one could argue that a length of wire is more accurate, because it guarantees the power supply will be in constant current operation during the current setting.

Everything you said is probably true. You seem to have missed my point entirely. Let me rephrase it for you.... These power supplies are not top grade laboratory power supplies. They are rather cheap chinese made power supplies that work but do not always have the highest grade components. This means (as seen by luccax's previous power supply) that they cannot always take the same abuse. Yes its in their spec to handle short circuit situations, it doesn't mean its good for it.

If you know how to properly use your test load you can, with 100% certainty, set your PSU perfectly every time without unneeded stress.

Anyway, I know what I am talking about contrary to what some may assume and what I am saying is basic logic. Obviously do what you like and what works for you, if you don't feel my suggestions are best for you then I will not care either way. Anything anybody ever says on here (contrary to what some may think) is either opinion or theory(its called scientific theory for a reason) and should only be taken as such.

BTW, that PSU has $20.99 shipping so it comes out to $80. I have Amazon Prime and get 2 day free shipping on certain items so I will see if I can find that model with my shipping. I would have no problem drop shipping that to you. PM me your address and I can calculate out what shipping would cost to send it to you.

EDIT: I found one that is eligible for my free shipping but the PSU itself is $80. I guess that just makes it 2 day shipping instead of that 4-5 day shipping.

Cyparagon said:

The current limit setting is the current limit setting regardless of the load. Hell, you could set the current limit with a 12V lamp if you wanted. If you set the current limit to 500mA at 0V, You can plug anything else into it, and it will still restrict the current to 500mA. You can use a test load if you want, but you're just wasting your time.

In fact, many PSUs have a starting spike in current mode (even some of the good ones) that poses a bigger threat than "slipping on the pot." The mastech I tested doesn't, however.

There's no reason to be afraid of setting the current on-the-fly assuming you start at zero, but if you're paranoid about it, shorting the output provides the same current set point as using a test load. In fact, one could argue that a length of wire is more accurate, because it guarantees the power supply will be in constant current operation during the current setting.

Click to expand...

The problem is: The adjustment potentiometer is not very trustable, it is a little loose and I could easily turn it too much overfeeding my diode and burning it. Using a test load is much better.

And I guess shorting the wires is a must do in my PSU to set it to the CC operation, without shorting it I wasn't able to limit the current while testing on a LED.

I'm not sure why everyone needs to have those digital psu's. Sure they are small and work. However I have TWO lab grade PSU's that can be shorted 24/7 at their max settings ALL day.

I paid next to nothing for both of my psu's (used on ebay) and they work great. I paid 100$ + shipping for my 7.5V 30A PSU and I paid 40$ shipped for my lambda 20v 6A PSU. Sure they are analog, but they both have current control as well as voltage control. You just need to know how to use a DMM to adjust your settings.

jeffreythe00 said:

I'm not sure why everyone needs to have those digital psu's...

You just need to know how to use a DMM to adjust your settings.

Click to expand...

The PSU itself is analog. It's only the built-in volt/ammeter that are digital. As you say, the main appeal is having extra resolution with less space. I like the look of the analog readout, myself. Precision isn't needed most of the time. And when it is, you can always attach an external DMM or two. There's always the option of upgrading the meters, too.

@luccax
By test load I am talking about something like one of these:
3 amp selectable test load [3AMPLD] - $19.99 : Cajunlasers Store

It creates a current load as well as it can be set to mimic the Vf specs of your diode. This is the same test load I use on my drivers and it does the same thing for my PSU. In fact you may want to look into getting one of these for testing your drivers even if you don't feel like using it for your PSU.

Cyparagon said:

The PSU itself is analog. It's only the built-in volt/ammeter that are digital. As you say, the main appeal is having extra resolution with less space. I like the look of the analog readout, myself. Precision isn't needed most of the time. And when it is, you can always attach an external DMM or two. There's always the option of upgrading the meters, too.

Click to expand...

Many seeries of power supplies are sold with analog and digital readouts, but are otherwise indeed indentical. The digital ones tend to be slightly more expensive, though thats hardly justified these days: a good analog panel meter costs as much as a digital one by now. They still fetch a premium price for many consumers though.

Digital or analog readouts do not determine the quality of the power supply though. It would still work if you completely removed the readouts, although you'd have to rely on multimeters to see what you're doing.

I like digital readouts because they are very quick to read at their maximum accuracy whereas you might have to look closer at an analog to make sure the needle is right where you want it to be. Not to mention viewing angle has no effect on a digital readout(LED like on mine is best for that) so my PSU can sit on the floor beneath my work space as long as I can reach the dials I can use it perfectly.

ApexProxy said:

@luccax
By test load I am talking about something like one of these:
3 amp selectable test load [3AMPLD] - $19.99 : Cajunlasers Store

It creates a current load as well as it can be set to mimic the Vf specs of your diode. This is the same test load I use on my drivers and it does the same thing for my PSU. In fact you may want to look into getting one of these for testing your drivers even if you don't feel like using it for your PSU.

Click to expand...

yes, i have two test loads, one from rckstr and one I've built myself.

I'm looking forward to buy one of those 'heavy duty' ones, i'll put it on my next cajunlasers order.

The cajunlasers one is totally beast. I even added a few small press on copper heatsinks to the diodes to extend its duty cycle. FlaminPyro makes a pretty solid test load as well.

Flaminpyro's is too expensive.

By the way, that PSU i sent you trough PM is manufactured here in Brazil, they have a website, manuals etc... They look serious, and i think i can return it if it fails. I'm in serious doubt...