Basic Power Supply

[Zorc]

Did aliexpress write worth value : €10,- on the box ?
I want no tax also, else to expensive.
What country are you from frex ?

On a box like that with corresponding weight you need to be really lucky if it gets through customs without getting noticed., regardless the written value

[Frex]

Did aliexpress write worth value : €10,- on the box ?
I want no tax also, else to expensive.
What country are you from frex ?

I always take in account that i have to possibly pay the customs tax
 when i buy something outside EU (i'm in France).
This time i had to pay it. The seller had reduced the real value,
 and because the product is heavy, customs asked me an invoice and to pay the VAT.
So i finally paid 49€ of VAT to customs.
Anyway, a great PSU for the price !

Frex
   

[psysc0rpi0n]

Hello.

Well, I'm getting a little bit frustrated now.
When I paid for the PSU, it said in the site that it was available only after 04/11/2019. It's 7th November and the PSU has not even been shipped. I sent 2 emails to the support and they only send a copy of the email I sent them to my email address.

Am I in trouble? I bought it at welectron.com but nobody replies to me. Anyone had bad experiences with this site?

[psysc0rpi0n]

They have replied to one of my few emails.

Quoting:

Regarding your order 34754: we are out of stock, incoming of goods is in about five weeks. However, we are glad to inform you that we still have two KA3005P Power Supplies left in our Warehouse. Through transport they got minor damages on the packaging unit. If you do not mind we will send those to you coming Monday.

Five weeks. And they want to send me damaged packages. This is not good for business. I asked them what they mean by minor damages in packaging. If they mean the hard paper box where they come is damaged or the PSU case itself. And asked them for pictures and also a compensation because I paid them as new.

What the hell??? Are they this bad? Jeeezzzz

[psysc0rpi0n]

Well, I agreed with them to take the PSUs they have available. The damages are in the cardboxes so no stress at all. I asked a compensation and they refund me 10%.
It was just a shame that communication was quite hard and delayed! Hope I get the PSU's somewhere next week.

[psysc0rpi0n]

Hello guys.

I just got my 2x Korads.

ISo, for the first test, should I just connect the PSU leds to my multmeter, set a value on the PSU and note it down in a Spreadsheet? Or what should I do? (I already did this).

For the Overshoot and other scope measurements, I think I need to spend some time learning how to properly filter out noise from scope and maybe other procedures to make it more accurate. I mean, if I short it's leads, I read weird things like too high Vmin and Vmax voltages of +/- 80mV with 1V / Div. If I zoom in up to 10mV /Div I get better values but looks like lots of noise.

Should I post the results of that first test I did? I'm not sure they are relevant.

I've just measured the following:
R - 470 ohm (464 ohm measured with Brymen)
V - 15V (14.97V measured with Brymen)

I = 30mA measured in Korad and 32.7mA measured with Brymen.

This might not be the best thing to do as I was using a 1/4 W resistor and drawing almost 1/2 W so temperature rises, resistance drops and current also rises, I guess. I don't have 1/2 W resistors I guess.

I did an OverShoot measuring with the scope but I'm not sure I did it correctly. I set the trigger level to 5V, Single Mode and turned the PSU ON. It read 5.439%.
I can take a picture tomorrow and upload it here if you want to see!

[psysc0rpi0n]

Of course, I will...

Site says it was not shipped yet!

Edited;

I guess it will take some time. It says there "Shipping from 04/11/2019"... So probably they have no stock or something like that!

Weren't you the one wanting to know more about the PSUs?
Well, I'm willing to run some tests but I need some guidance. I should mention I don't have a proper load to test the PSU's. Or better, I have 2 extreme loads. Regular resistors of 1/4W and an auto lead light bulb that draws almost the limit of the power supply. So, I on't want to force them this much.
Tell me what you would like me to do!

[Gandalf_Sr]

Glad your PSU eventually arrived.  A few comments:

1. Running a 0.25 W resistor at 0.423 W for a short time (5 Seconds) should be fine, just be careful as the resistor may get hot and switch off if you see smoke.
2. If you set a current limit on your PSU output, you should be able to short out that output with a piece of wire and see the display show that current flowing, the voltage should automatically reduce to enforce the current limit you set.
3. If you have 4 x 1/4 W 470 Ohm resistors you can wire them up to make one 1/2 W 470 Ohm resistor like this...
 a. Wire two 470 Ohm resistors in parallel, you will have a 235 Ohm resistor with double the W
 b. Make another 235 Ohm resistor
 c. Wire the two 235 Ohm resistors in series
4. Car stop/tail light bulbs have 2 different filaments, 5W and 21W. Use your meter to measure the resistance of the filaments and whichever is the lower, that is the 21W filament.  Be aware that measuring the resistance cold will not be the same as when it's hot (i.e. lit up).

Pictures would be good.  FYI, I own a Korad PSU

[Edit] Have you seen ? 

It shows that the PSU can get confused and output 56V !!!! FF to 29:00 onwards

[psysc0rpi0n]

Hi @Gandalf_Sr.

Ok, I asked about the method of drawing some current based on some Dave video about testing a PSU for noise and ripple. But he used something I don't have which is a constant current sink or whatever is the name of it. It draws the current from the PSU and then he uses a scope to show and explain the results. He used like 2A and that's why I wanted something more than just a few mA.
So I came across with a simple example of a current sink circuit:


So I used for this circuit:
LM358N OpAmp
10nF cap
1k Ohm, 100 ohm and 2.4 omh (5x 10ohm resistors in // - the 2.4 ohm is real value, not theoretical 10/5 = 2 ohm)
IRF640 MOSFET N-Channel
For R_L I used different loads such as lead auto light (about 4.5A to avoid dropping 12V), desk light bulb of 1.5W and 5x 100ohm resistors in //. These 5x 100 ohm resistors still get hot with about 200mA.

Dave's video is this:


Edited;

I also did what you told me about the 470 ohm resistors. But it's still very low current. About 40mA if I set the PSU to 20V.



Edited 2;
With the circuit I posted above with the OpAmp and MOSFET, I can draw about 600mA for about a minute or so before the MOSFET and the sensing resistors get too hot.
Here are 2 screens:

[Gandalf_Sr]

Dave has a BK programmable load (I have an old HP one), you set the current/Wattage/voltage you want and it trys to do that and shows you the actual current and voltage.  Inside it's got big power transistors fastened to big heatsinks with big fans.  If all you want to do check the ripple on the power supply when you pull say 2A from it, the simplest and cheapest way would be to use a 21W automotive bulb (turn signal) fed with 10.5 V.  The ones you bought are Linear so they should be pretty clean as far as ripple at half their rated load is concerned.

If you want something that's cheap and reusable you might try entering "electronic load" into eBay or Aliexpress, that will give you a bunch of cheap devices that range from some high-wattage resistors with switches ($3) to FET with a big (PC processor style) heatsink and fan with an LCD display that shows Voltage, Current, and Watts ($35).  I don't own one and can't tell you about its performance but beanflying has a couple he briefly mentions here.  Those loads probably use a microcontroller to regulate the drive to a FET. Your circuit is effectively a linear (rather than switching) solution and would work but it would cost you more to make it than the $35 one from China.

[psysc0rpi0n]

Dave has a BK programmable load (I have an old HP one), you set the current/Wattage/voltage you want and it trys to do that and shows you the actual current and voltage.  Inside it's got big power transistors fastened to big heatsinks with big fans.  If all you want to do check the ripple on the power supply when you pull say 2A from it, the simplest and cheapest way would be to use a 21W automotive bulb (turn signal) fed with 10.5 V.  The ones you bought are Linear so they should be pretty clean as far as ripple at half their rated load is concerned.

If you want something that's cheap and reusable you might try entering "electronic load" into eBay or Aliexpress, that will give you a bunch of cheap devices that range from some high-wattage resistors with switches ($3) to FET with a big (PC processor style) heatsink and fan with an LCD display that shows Voltage, Current, and Watts ($35).  I don't own one and can't tell you about its performance but beanflying has a couple he briefly mentions here.  Those loads probably use a microcontroller to regulate the drive to a FET. Your circuit is effectively a linear (rather than switching) solution and would work but it would cost you more to make it than the $35 one from China.

Well, for now I just wanted to check how good/bad are these PSUs.
You said to check for OverShoot/Noise/Ripple issues.

This was what I wanted to check but I have no fancy devices obviously. I'm just someone that likes electronics but my knowledge and experience is very limited.
I like to do simple things and see other things working and understand how they work. I don't do this for living as it is obvious.
I like to program micro controllers just for fun. I like to build very basic things with discrete components. But even this is now more challenging to me because I stopped practicing since I finished my graduation and my job, despite the fact that is kind of related with my graduation, has very little or nothing to do with electronics and programming (I'm working as a project designer for fiber optic networks).

I didn't understand what you said about

Your circuit is effectively a linear (rather than switching) solution and would work but it would cost you more to make it than the $35 one from China.

You mean the circuit I built based on that image I posted above? If so, how would it be more expensive than a $35 one from China?
If not, what solution you're talking about that costs $35?

[Gandalf_Sr]

All I meant was that, by the time you bought all the parts to build your circuit, including a heatsink, it would be cheaper to buy a $35 load from China.  Plus your time.

[psysc0rpi0n]

All I meant was that, by the time you bought all the parts to build your circuit, including a heatsink, it would be cheaper to buy a $35 load from China.  Plus your time.

If you're referring to the circuit I posted about to use as a current sink, I don't need to buy anything because I have all components here with me! And even if I had to buy them, they would be cheaper than $35. An OpAmp, a Mosfet, a couple of resistors, a capacitor and an heat sink would not be that expensive, I guess.

[Gandalf_Sr]

OK on building from parts that you have but be aware that power in Watts = volts x amps so, if you ran 12V at 2A through the circuit you showed with a 1 Ohm resistor and an IRF640:

a. The resistor would dissipate 4W (power in a resistor is amps x amps x resistance (I²R))
b. The FET would dissipate 20W (12-2 (the voltage dropped over the resistor) = 10 x 2A)

This means that they would both get VERY hot so you'd need high wattage resistor and a heat sink on the FET.

Sorry if this is telling you something you already know.

[psysc0rpi0n]

OK on building from parts that you have but be aware that power in Watts = volts x amps so, if you ran 12V at 2A through the circuit you showed with a 1 Ohm resistor and an IRF640:

a. The resistor would dissipate 4W (power in a resistor is amps x amps x resistance (I²R))
b. The FET would dissipate 20W (12-2 (the voltage dropped over the resistor) = 10 x 2A)

This means that they would both get VERY hot so you'd need high wattage resistor and a heat sink on the FET.

Sorry if this is telling you something you already know.

Yes, I'm aware but thanks for warning me.

My sensing resistor is not 1 ohm as in that circuit. It's 2.4 ohm. I didn't have any 1 ohm resistor. So I used 5x 10 ohm resistor (theoretical value should be 2ohm but I get 2.4 ohm).
Anyway, to prevent damaging FEt or resistors, I'm limiting the current to 600mA.

I also can use a light bulb from a desktop chandelier that can consume up to about 1.6A. And I don't need to be careful with heating components.
I did an experiment yesterday with the scope and this light bulb. The chandelier is rated at 12V, 20W max. This light bulb, according to the PSU, at 12V, draws about 1.6A (19.2W).

I don't have enough knowledge to analyse what the results mean.
I tried a few settings with the scope trigger. It was harder and uglier to measure the rising edge OverShoot (and I hope I did it right) than to measure the falling edge OverShoot.

For instance, when measuring the "connecting the load" scenario (rising edge), I can only get measurements from scope at time frames of 100ms and above.

Rising edge @ 100ms - 0.9644% OverShoot


Rising edge @ 200ms - 0.3624% OverShoot


Rising edge @ 500ms - 0.3496% OverShoot


Falling edge @ 2ms - 1.081% OverShoot


Falling edge @ 10ms - 0.3539% OverShoot


Falling edge @ 50ms - 0.3492% OverShoot


Falling edge @ 100ms - 2.306% OverShoot


Falling edge @ 200ms - 1.040% OverShoot


Falling edge @ 500ms - 0.3487% OverShoot


Power Supply:


My scope user manual says about OverShoot:

13. Overshoot: the ratio of the difference of the maximum value and top value of
the waveform to the amplitude value.

I can now test with an "heavier" load if needed. About 4A. Is it needed?

[Gandalf_Sr]

...I can now test with an "heavier" load if needed. About 4A. Is it needed?

Only you know what you "need".

If you want to know what the ripple is like at 4A, go ahead and test.  All power supplies have noise but linear supplies are similar and will typically have much lower noise than switched ones.  I own expensive HP and Rigol supplies plus cheap $99 linear ones including a 3-channel Korad that has the same button layout as yours.  I'd be more concerned that you could repeat Dave's case of confusing the Korad to send out almost 60V when it was set for 22V; a bit of noise is normal but sending out 3 x the set voltage is a potential disaster (pun intended).

[psysc0rpi0n]

...I can now test with an "heavier" load if needed. About 4A. Is it needed?

Only you know what you "need".

If you want to know what the ripple is like at 4A, go ahead and test.  All power supplies have noise but linear supplies are similar and will typically have much lower noise than switched ones.  I own expensive HP and Rigol supplies plus cheap $99 linear ones including a 3-channel Korad that has the same button layout as yours.  I'd be more concerned that you could repeat Dave's case of confusing the Korad to send out almost 60V when it was set for 22V; a bit of noise is normal but sending out 3 x the set voltage is a potential disaster (pun intended).

Ok, I thought you wanted to know the OverShoot of these PSUs for comparison... That's why I was testing this. It's not that "I need" to know for any specific load.

[Gandalf_Sr]

It wasn't me that asked for that.

[psysc0rpi0n]

It wasn't me that asked for that.

Ohh, I'm sorry. It was @Zorc. I really sorry sir!

[Gandalf_Sr]

Not a problem.

[eevcandies]

There been a bunch of these supplies sold the last few months...some went for approx $50  , I got one for about $80 , plus shipping, like new.

https://www.ebay.com/itm/Kikusui-PMP16-1QU-4-Channel-06V-0-16V-016V-025V-Regulated-DC-Power-Supply/264451210613?epid=15025745400&hash=item3d92853d75:g:-n8AAOSwk2tcpoxW