Review of Korad 6003P DC power supply

[wls-canada]

New registrant, returning hobbiest, first post.

I discovered EEVblog a few days ago looking for reviews of cheapo-Chinese bench DC power supplies.  As a non-serious enthusiast looking to save a few $$ I couldn't pass up $145 CAD + free shipping Korad 6003P from Amazon.ca, although I was skeptical based on other reviews, in particular Dave's of the 3005P.

Anyways, I felt like sharing back with my own review of the 6003P.  Summary:

• There's a transient on turning on the power supply.  Don't connect to your circuit under test until it is on.
• After it's on, there is no transient when turning on or off output enable.
• There is considerable periodic noise on the output (140mV p-p)

I have to admit the last one concerns me and I'd welcome opinions as to whether this unit is a dud, or it's typical and expected of the breed.

[SharpEars]

New registrant, returning hobbiest, first post.

I discovered EEVblog a few days ago looking for reviews of cheapo-Chinese bench DC power supplies.  As a non-serious enthusiast looking to save a few $$ I couldn't pass up $145 CAD + free shipping Korad 6003P from Amazon.ca, although I was skeptical based on other reviews, in particular Dave's of the 3005P.

• There's a transient on turning on the power supply.  Don't connect to your circuit under test until it is on.
• After it's on, there is no transient when turning on or off output enable.
• There is considerable periodic noise on the output (140mV p-p)

I have to admit the last one concerns me and I'd welcome opinions as to whether this unit is a dud, or it's typical and expected of the breed.

I left a reply for you on YouTube, but I'll post it here as well:

If there is really an 8.6 V transient on turn on, I consider the supply useless and you should return it back to Amazon and get your money back. I wouldn't be surprised if there is a similar spike when you power down the supply from full on (i.e., not just turn off the output, but the entire supply as a whole via either the power switch or by yanking the power cord from the outlet, while it is sourcing say 15 V). Either behavior is totally unacceptable for a PS that costs more than $40.

You most certainly can beat the price/performance by buying a used HP/Agilent supply on eBay. Granted, it won't be as small, but it will have reasonable noise characteristics and will not have power-on/off transients.

As for the noise you are seeing, if your scope is not connected properly, you may be just picking up ambient power line noise. You have to be very careful, to do a proper noise measurement. For example, the alligator ground lead on your scope probe should be removed and only the spring pin, used. All non-shielded parts must be as short as possible and you should directly at the supply terminals (i.e., no connector cables or leads).

Try turning the supply off (or at least the output off) and see if the noise is still present. This is often a good indicator of ambient noise rather than PS noise.

[Electro Fan]

A 14V test with a 100 ohm resistor on a KA3005P shows less than a 4V Pk to Pk transient lasting 200ns during AC power on (with output off) but a 10V Pk to Pk transient that lasts 500 us during AC power off.  During output power on the DC ramps up to about 15V within 1.5ms and then settles back to 14V within another 0.5ms.  Noise is about 40mV Pk to Pk AC coupled on 14V and about 22mV Pk to Pk AC coupled on 3.3V.

[SharpEars]

A 14V test with a 100 ohm resistor on a KA3005P shows less than a 4V Pk to Pk transient lasting 200ns during AC power on (with output off) but a 10V Pk to Pk transient that lasts 500 us during AC power off.  During output power on the DC ramps up to about 15V within 1.5ms and then settles back to 14V within another 0.5ms.  Noise is about 40 mV Pk to Pk AC coupled on 14V and about 22mV Pk to Pk AC coupled on 3.3V.

So, why did you not send it back (to China)?? I am pretty sure you kept it, but why? It's seriously looking like a supply.

If it was solely because of its size, that I can can kind of, sort of understand.

[Electro Fan]

A 14V test with a 100 ohm resistor on a KA3005P shows less than a 4V Pk to Pk transient lasting 200ns during AC power on (with output off) but a 10V Pk to Pk transient that lasts 500 us during AC power off.  During output power on the DC ramps up to about 15V within 1.5ms and then settles back to 14V within another 0.5ms.  Noise is about 40 mV Pk to Pk AC coupled on 14V and about 22mV Pk to Pk AC coupled on 3.3V.

So, why did you keep it?? And I know you did...  on this supply

Not sure that a 500us turn off transient matters much to anything I'm working on.  I guess you would find that to be a problem?

[SharpEars]

A 14V test with a 100 ohm resistor on a KA3005P shows less than a 4V Pk to Pk transient lasting 200ns during AC power on (with output off) but a 10V Pk to Pk transient that lasts 500 us during AC power off.  During output power on the DC ramps up to about 15V within 1.5ms and then settles back to 14V within another 0.5ms.  Noise is about 40 mV Pk to Pk AC coupled on 14V and about 22mV Pk to Pk AC coupled on 3.3V.

So, why did you keep it?? And I know you did...  on this supply

Not sure that a 500us turn off transient matters much to anything I'm working on.  I guess you would find that to be a problem?

Hard to say, how confident are you in both the magnitude (across multiple turn off cycles from multiple "on voltages") and duration of your finding? I.e., was it the absolute max across multiple attempts at various voltages and loads that you presented, or just a one-shot deal at one setting?

For example, say I have a coil connected to the output. What happens when I power the supply off and the field on the coil collapses? Will the supply dampen it or will there be "infinite" voltage present to the rest of the circuit, "behind" the coil, because a relay in the supply completely disconnected its outputs?

[Electro Fan]

A 14V test with a 100 ohm resistor on a KA3005P shows less than a 4V Pk to Pk transient lasting 200ns during AC power on (with output off) but a 10V Pk to Pk transient that lasts 500 us during AC power off.  During output power on the DC ramps up to about 15V within 1.5ms and then settles back to 14V within another 0.5ms.  Noise is about 40 mV Pk to Pk AC coupled on 14V and about 22mV Pk to Pk AC coupled on 3.3V.

So, why did you keep it?? And I know you did...  on this supply

Not sure that a 500us turn off transient matters much to anything I'm working on.  I guess you would find that to be a problem?

Hard to say, how confident are you in both the magnitude (across multiple turn off cycles from multiple "on voltages") and duration of your finding? I.e., was it the absolute max across multiple attempts at various voltages and loads that you presented, or just a one-shot deal at one setting?

I tried it about 10 times at 14V - seems consistent.  I just tried the turn off at 3.3V and the turn off transient is about 4V Pk to Pk for a duration of about 150ns.  Turn off is also about 4V Pk to Pk for a duration of about 150ns.

[SharpEars]

I tried it about 10 times at 14V - seems consistent.  I just tried the turn off at 3.3V and the turn off transient is about 4V Pk to Pk for a duration of about 150 ns.

I updated with a question of how the supply will behave when an inductive load is connected to it. Will it dampen the field collapse? Will it fry? These are the sort of tests that separate the man supplies from the boys...

[Electro Fan]

I tried it about 10 times at 14V - seems consistent.  I just tried the turn off at 3.3V and the turn off transient is about 4V Pk to Pk for a duration of about 150 ns.

I updated with a question of how the supply will behave when an inductive load is connected to it. Will it dampen the field collapse? Will it fry? These are the sort of tests that separate the man supplies from the boys...

Don't know the answer to that.  If I was working on something worth more than the power supply it might be something worth better understanding

One other thought, the transients seem to get past the output control during AC power up and power down.  Seems like there are a fair number of Power Supplies (especially under $150 or so) in use that don't have an output control.  At least with this unit the PS can be turned on with the AC button, the DUT can be connected, then the output control can be used to turn on and off the DC, then the DUT can be disconnected, and then the PS can be turned off with the AC button.  So, if something is particularly valuable or sensitive there is a way to deal with it.

[SharpEars]

I tried it about 10 times at 14V - seems consistent.  I just tried the turn off at 3.3V and the turn off transient is about 4V Pk to Pk for a duration of about 150 ns.

I updated with a question of how the supply will behave when an inductive load is connected to it. Will it dampen the field collapse? Will it fry? These are the sort of tests that separate the man supplies from the boys...

Don't know the answer to that.  If I was working on something worth more than the power supply it might be something worth better understanding

LOL, I hear ya  .

[wls-canada]

• There is considerable periodic noise on the output (140mV p-p)

I have to admit the last one concerns me and I'd welcome opinions as to whether this unit is a dud, or it's typical and expected of the breed.

I left a reply for you on YouTube, but I'll post it here as well:

If there is really an 8.6 V transient on turn on, I consider the supply useless and you should return it back to Amazon and get your money back. I wouldn't be surprised if there is a similar spike when you power down the supply from full on (i.e., not just turn off the output, but the entire supply as a whole via either the power switch or by yanking the power cord from the outlet, while it is sourcing say 15 V). Either behavior is totally unacceptable for a PS that costs more than $40.

You most certainly can beat the price/performance by buying a used HP/Agilent supply on eBay. Granted, it won't be as small, but it will have reasonable noise characteristics and will not have power-on/off transients.

As for the noise you are seeing, if your scope is not connected properly, you may be just picking up ambient power line noise. You have to be very careful, to do a proper noise measurement. For example, the alligator ground lead on your scope probe should be removed and only the spring pin, used. All non-shielded parts must be as short as possible and you should directly at the supply terminals (i.e., no connector cables or leads).

Try turning the supply off (or at least the output off) and see if the noise is still present. This is often a good indicator of ambient noise rather than PS noise.

Thanks for the reply SharpEars.  I am not too concerned about the on-spike - unless it's the sign of a defect and not a deficiency - that is easily mitigated.

In Canada, counting shipping and import fees, I have never been able to find even a used EBay HP/Agilent for less that $350, and believe me I've looked. 

The noise is not present when the leads are disconnected so it's not environmental - also the noise bursts are periodic, in the range of about 300Hz.  In fact, that would suggest to me (based on theory, not real-world experience) that it is a switching artifact - but I don't have enough experience in DC switched power supplies to know if it's typical of the breed and, more importantly, it is an actual defect in the unit.  Is this something others have seen as "normal" in budget DC power supplies? 

[dom0]

300 Hz? Or (rather accurately) 217 Hz? If so I suggest turning your cell phone off while doing noise measurements

[SharpEars]

300 Hz? Or (rather accurately) 217 Hz? If so I suggest turning your cell phone off while doing noise measurements

Noice! 

[SharpEars]

The noise is not present when the leads are disconnected so it's not environmental - also the noise bursts are periodic, in the range of about 300Hz.  In fact, that would suggest to me (based on theory, not real-world experience) that it is a switching artifact - but I don't have enough experience in DC switched power supplies to know if it's typical of the breed and, more importantly, it is an actual defect in the unit.  Is this something others have seen as "normal" in budget DC power supplies?

Switching artifacts are on the order of hundreds of kHz (to even 1-2 MHz for some of the better switchers).

[wls-canada]

I think I have made a very basic mistake.  The noise is at 120 Hz (and yes, I've measured it without a cellphone nearby )  And it's present when the leads are plugged into the power supply, even if the latter is unplugged from mains.  When the leads are removed from the power supply, the noise goes down into 20mv range (and close to zero when there are no leads at all).  What this would imply is that (a) SharpEars was correct in his caution about the leads acting as antennas but also (b) there are significant 120 Hz antenna effects introduced by the circuitry within the power supply even when it is off and unplugged (large coils inside perhaps?).  I will add some capacitance across that resistor and see.  As this was a presumably wirewound power resistor in my original, I could see it being inductive as well.

This is all a great refresher in basics for me.  It's been 15 years since I worked with electronics, and a much longer since I did the studies.  I used to keep my thesis around to see how I understood it less each year.  Mostly I've done VHDL and DSP simulation even when I was a real engineer, very little practical bench work in decades.  Your patience is appreciated .

[DaneLaw]

So, why did you not send it back (to China)?? I am pretty sure you kept it, but why? It's seriously looking like a supply.

So' why did you keep it?? And I know you did...

sending it back to China (ref the TS), nahh..

Item of this size, is not really something you can ship back from the western world to China, unless it cost and arm and a leg, what these doesnt, so it useally a keeper or an dispute for some of the funds back..
Also got a ouple of these KD3005P purchased here in 4q2016 (from Reichelt Germany) and the KA3003P (from China), and if I should ship just one back, here from Denmark to China, it will cost more then i paid for both of them together including shipping, hence shipping to China for an 5.1kg cost whooping 150US alone, and Australia or USA, or Canada is not fare from these insane postageprices we encounter here..

wls canada' thx for the review, trying to get some pro & cons insight into these models and particaly the different batches.(since the KA model sounds like a JET when on, certainly not the case for the KD model, wich I guess is of newer heritage, and also with inserted-EUplugs..
the older KA seem to suffered from high-fan-noise based on currentload, but there where also some other problems that hazzled these older productionruns, so would be great if there where some info to classify, if this KA3003P that just arrived today from China is of old stock, since it was insanely cheap and ewen cheaper then what the postage cost from China to Denmark on an 5.1kg item with this  DHL-expressshippingform, this was send under from China.
any batch/serielnr overview, to sense what productionrun the KA model is/of, without going thrue the inside-spec in details.