EEVblog Electronics Community Forum
Products => Test Equipment => Topic started by: Wmacky on May 17, 2014, 03:09:06 am
-
Just bought an inexpensive lab supply ( I don't use them that often, but do need one ever so often). SoI throw it on the bench, hook up the fluke and give it a whirl! Everythimg seems great, The amps are dead on compared to the fluke, and voltage within .02 V. It all worked great!. Then I found it. I always have to find it.............. With the DC output turned off, my fluke reads -.112V. WTH? Hooking up the fluke as a dead short to read amps, shows no amp draw, but whats the deal with this .112 negative volt. GRRRRRRR. I'm testing sensitive, easliy blown expensive laser modules, and this worries me! Othewr Youtube videos show other getting 0 volts with the output off. Will this POS have to be returned?!?. Is there some way to remedey this to save time?
Whenever I get a new piece of gear. I never >:( get to the stage of wanting to :Take it apart"! I always end up wanting to smash it apart! :box:
-
The high input impedance of the multimeter on its low range makes it sensitive to the small amount of leakage through whatever the PSU uses to switch off the output. If it bothers you, try connecting a resistor inside the power supply directly across the back of jacks, the highest value that will pull the voltage down satisfactorily low. I suspect 10k-ish will do.
-
I'll look for some. I did try a diode across the terminals, as well as a RC hobby battery charger for a load, with no effect at all. Not even 1 digit. I guess the part that really gets me is the negative part. Isn't like like hooking up reverse voltage?
-
If there's "no amp draw", then it can't take much load to drop the voltage. Keep in mind - a diode will draw almost no current (a quick back-of-the-envelope gives picoamps to nanoamps) at a 0.1V bias, and the battery charger will be similar, as it's likely made from semiconductors too ;D
-
Sounds like a case of leakage or induction when the switch is in the off position. Like c4757p said, a resistor (say 10K) across the meter leads should drop the reading to zero.
-
So you guys wouldn't do a return? Also, Would that resister effect the other readings with output engaged?
-
What we are saying is that there is no fault with the PSU. The reading on your DMM is not valid due to it's high input impedance.
The resistor across the DMM leads is temporary and only there to prove (or not) that leakage / induction and its effect on the meter is what you're seeing.
-
Right. Adding the 10k will measure the voltage drop across the resistor, rather than rely on the multimeter's imperfect measurement ability at low impedance and low voltage. Since the circuit is only a single resistor, you will measure the voltage the power supply is actually outputting (which will be much closer to zero than your existing measurement) as the voltage drop across the now increased impedance voltage source.
That's my understanding, at least. I am probably wrong. Someone will correct me, if so.
-
I didn't have a 10K but it looks like 10K would have barely moved the digits?
Here's the results so far: All negative voltage.
DMM only - .112
5.6K - .096
2.7k -.086
1K -.058
.5k -.037
EDIT:
100 Ohms -.012
Seems to be taking a lot more resistance than original suspected? Thoughts?
-
What is the new power supply?
-
What is the new power supply?
I hate to say as it's not well loved, and I hate abuse! :palm:
-
What is the new power supply?
I hate to say as it's not well loved, and I hate abuse! :palm:
Kinda hard to give constructive comment on whether your PSU is broken or not without knowing what it is. A HP 6632B will have around 6mV on the output when disengaged, which is apparently normal for that device.
For a switching power supply, residual output when off would be a bit weird.
-
I hate to say as it's not well loved, and I hate abuse! :palm:
I don't think that abuse will follow, no matter what PS you're using.
-
My money is on it being a Korad, as I have one, and it's got the same behavior.
The -100mV on the output when (soft) off haven't caused me any issues.
-
100 Ohms -.012 V
12 mV across 100 ohms would correspond to about 120 µA. That's not enough current to startle a flea.
Remember a power supply is designed to supply power, so the important test is what it does when the output is on. What it does when off is of much less concern.
-
I doubt that -.112 V is going to damage anything. I would be more concerned about the supply's transient behaviour. Does the voltage overshoot when the output is turned on? Does it overshoot when the load changes? Can you rely on the current limit, or is there a dirty big reservoir cap waiting to vaporise your mirrors at the first chance it gets?
-
My money is on it being a Korad, as I have one, and it's got the same behavior.
The -100mV on the output when (soft) off haven't caused me any issues.
That's a pretty good guess! I wonder why ours has that same exact negative voltage, but others on you tube, and on the EEV review show 0 Volts? At least I know a replacement won't be worth doing. IF everyone agrees that I shouldn't worry with it, I'll move to more testing with my lowly little Rigol DSO. Good time ti learn how to do that single shot capture stuff! I don't think those tests have been done since the PS was redesigned!
-
My money is on it being a Korad, as I have one, and it's got the same behavior.
The -100mV on the output when (soft) off haven't caused me any issues.
That's a pretty good guess! I wonder why ours has that same exact negative voltage, but others on you tube, and on the EEV review show 0 Volts? At least I know a replacement won't be worth doing. IF everyone agrees that I shouldn't worry with it, I'll move to more testing with my lowly little Rigol DSO. Good time ti learn how to do that single shot capture stuff! I don't think those tests have been done since the PS was redesigned!
Yeah, it sounded familiar :) .. I wouldn't worry about it.
The PSU is a budget one, but I've found it will take a lot of punishment while remaining pretty accurate and performing well. Sure it's no SMU, but it's really good value for what it is, couldn't build anything similar for the same money.
-
[…] about 120 µA. That's not enough current to startle a flea.
I would love to see that proven empirically ;)