What power supply do you use for emitter testing?

My bench power supply is an RD DPS5020 and while it’s quite nice I’m a bit disappointed that the ripple current is high enough to cause a visible fluctuation in the brightness of an LED.

I’m not using the power supply for LED output tests yet but I’d like to be able to when I find the time. I’m not sure whether the ripple current would be an issue for output tests and it doesn’t matter for the basic things I’m using the power supply for at the moment, but it is quite annoying.

I definitely don’t need a 20A power supply, but the 15A DPS5015 and 12A DPS3012 have the same current 0.01A resolution as the DPS5050

I’m considering getting one of the 5A models which has a 0.001A current resolution. I’m sure it will be adequate for 99% of my needs, but it might not quite get maximum output from some LEDs.

Can anyone recommend an adjustable power supply suitable for LED output tests for ~$50?

It’s nice to have some resolution, but the thing is - even if you have 0.1V/0.01A resolution, you can measure voltage and current with few multimeters, you will never have matching values. If you have a match, it’s not going to hold through the whole range. Some volt/ammeters have trimmers, probably these modular buck converters from RD do too, so you can calibrate a bit, but you’re always going to have some error, and the error will be greater than resolution, even if you spend a fortune on a ps. The thing with adjustable dc power supplies is that you can easily get to some power, you probably have a psu in your pc capable of few 20-30A rails at 12V, and you can buy Zhaoxin kxn3020d for 70 bucks (which I would warmly recommend), prices of ps units start multiplying with all the filters, high quality caps, all those things that lower the ripple as much as possible, then you have multiple outputs which are electrically isolated so you can put them in parallel or series if necessary, than there are fancy options like timers, constant power mode range settings, load settings, intelligent charging, and if you throw some money on top of THAT pile, only then you’ll have a high precision everything through the whole range. This is why there are European power supplies that will cost you 10$/W, and that is why I strongly suggest a kxn3020d if you’re not from a country that will bleed you dry with customs fees like I am.

I had one of these Ruideng units under a scope, I think that it was a DPH version, similar in power, really couldn’t say which one exactly. Anyway, these are great units for the money. Everything except the housing :wink:

Anyway, you don’t have a problem with switch ripple, you have problems with current spikes on (probably only) low load, if I understand correctly from videos. I mean, you have current fluctuations, you shouldn’t have voltage fluctuations in constant current mode? There’s a high frequency ripple, maybe a mV or 2 per W, that’s the way smps works, but you shouldn’t see any power fluctuations, other than ones caused by heating up the led.

I mean - do you? What happens in cc mode, and what happens with a bit more load? Not like induction heating, I mean, what happens if you power several straw leds, in parallel AND in series, have you tried, do the fluctuations persist?

I just assume that the problem is only with current regulation with very low draw load, that’s not to say that this is “ok”, it’s not ok and it should not happen.

Hey, btw, I just vote for 30V 20A ac-dc smps option because of convenience for me. I am used to it, I spend half of my life in wilderness, and I am used to my psu home - I set up the system to charge all my cells at the same time, I can charge a 12/24V lead acid cells, AND yes, I can test emitters. Not to publish the results with breathtaking accuracy, but to see what I like, find a balance between battery life and led output, choose driver accordingly, nothing else.

Don’t get me wrong, I am a flashlight hobbyist, enthusiast, I built some c8-s, some p60-s, I am even proud of some of those, but if I needed a power supply for led testing only, that would be:

An old computer psu
300W 20A buck converter 6$
Volt and current meter 2$
Two multiturn 1W potentiometers, so that I can solder them instead of trimpots that come on a buck converter and to round it up to 10 bucks.

Nothing else needed, everyone can do it, nothing can go wrong, shipping included, no customs fees, and you can push xhp70 to it’s limits.

Or, you can push some of those latest, highest possible quality leds such as xhp90, xhp110, xhp70.3, even the famous 700.000 lumen “one led - Sun” 9-die p10 wick that Cree unknowingly produced over the limit. Probably all at once :smiley:

@skinny

What’s the frequency? kenneth

the ripple frequency—how about put an inductor inline with the output of your power supply followed by a suitable sized capacitor? That would filter out the ripple and stiffen up the supply for heavy load changes. The coil will smooth out the current ripple and the capacitor will smooth out voltage fluctuations. i’ll bet that it will be cheaper cost and better than buying a new one and you might already have suitable parts in your scrap bin.