New Bench Power Supply - Chinese

[DW1961]

It's a good idea to check the security of all the connections in these, particularly the mains ones, they are not infrequently dangerously loose. Also the one I have (but never use, since it destroys its own display, faulty LED driver IC probably) had not only loosely terminated mains cabling, but some of the semis weren't screwed back tightly to their heatsinks. Remedying this cured a certain amount of audible noise at higher loads...

Now, this isn't from the same manufacturer as mine, or the self-igniting one that Big Clive looked at, but it's worth bearing in mind it'll be designed and built with much the same philosophy and methodology; pare back the costs by using the cheapest possible components (both electrical and mechanical) and assembled by horribly underpaid and overworked unskilled labour.

TLDR; I wouldn't leave it running unsupervised, or powering anything I would be upset to lose.

I'll get back to you after I do that. Chinese electronics and products overall are getting better, or I should say, can be better than in the past (UNI-T DMM, for example). Some of the speakers I've had for review sound quite good for $100 speakers. Some, not so good at all. Will they stand test of time, I wouldn't bet on it, but you never know. I have a Chinese class D amp that uses the Texas Instruments class D 3116D2 amp chip and a Qualcomm  Bluetooth chip, and I have had zero problems with it over the last two years, running 16 hours a day (low volume). 

[DW1961]

If you're posting a thread like this, then you should be posting traces of output performance - turn on / off and load change transient behaviour, ripple and noise etc.

A few oversized internal photos really don't tell you anything much (apart from grounding) about the desirability of the supply for actual use.

It can tell you how well they have use RFI and other filters and capacitors, grounding, etc. though.  Of course, the end result is a scope. I don't have a scope so I can't measure the output power cleanliness and efficiency of the unit. I was just wondering if even the unit is electronically or mechanically sound. I only saw one coil choke on it, for instance.

Unfortunately not. A really detailed analysis of the PCB layout (and schematic) might give a clue, but the effort would be considerable, for an experienced designer, and you wouldn't really have confidence in this without actual measurements. It depends on things like the internal construction quality of the magnetics etc, which are invisible.

Just wondering because when I read computer power supply reviews, where they do fully test the units using a scope, hot boxes, and at various loads, they always do a run down about what the board looks like, separation of mains , rectifier usage, filters, what brand caps they use and how many , soldering job, OCP protections, etc. ex:

COMPONENT ANALYSIS
General Data   -
Manufacturer (OEM)   CWT
PCB Type   Double Sided
Primary Side- Transient Filter: 4x Y caps, 2x  X caps, 2x CM chokes, 1x MOV
Inrush Protection: NTC Thermistor SCK-037 (3 Ohm) & Relay
Bridge Rectifier(s): 2x GBU1506 (600V, 15A @ 100°C)
APFC MOSFETs: 2x Vishay SiHF30N60E (650V, 18A @ 100°C, Rds(on): 0.125Ohm)
APFC Boost Diode: 1x On Semiconductor FFSP0865A (650V, 8A @ 155°C)
Bulk Cap(s): 2x Nippon Chemi-Con (400V, 470uF each or 940uF combined, 2,000h @ 105°C, KMW)
Main Switchers: 2x On Semiconductor FCPF190N60E (600V, 13.1A @ 100°C, Rds(on): 0.19Ohm)
APFC Controller: Champion CM6500UNX & Champion CM03X
Resonant Controller: Champion CU6901VAC

Topology   
Primary side: APFC, Half-Bridge & LLC converter
Secondary side: Synchronous Rectification & DC-DC converters
Secondary Side: +12V MOSFETs 6x On Semiconductor NTMFS5C430N (40V, 131A @ 100°C, Rds(on): 1.7mOhm)
5V & 3.3V: DC-DC Converters: 2x UBIQ QM3054M6 (30V, 61A @ 100°C, Rds(on): 4.8mOhm) & 2x UBIQ QN3107M6N (30V, 70A @ 100°C, Rds(on): 2.6mOhm)
PWM Controllers: UPI Semi uP3861P
Filtering Capacitors: Electrolytic: 7x Nippon Chemi-Con (1-5,000h @ 105°C, KZE), 7x Nippon Chemi-Con (4-10,000h @ 105°C, KY), 1x Rubycon (4-10,000h @ @ 105°C, YXJ)
Polymer: 37x FPCAP
Change Over Switch: 1x Sync Power SPN3006 MOSFET (30V, 57A @ 100°C, Rds(on): 5.5mOhm)
Supervisor IC: Weltrend WT7502R (OVP, UVP, SCP, PG)


The transient/EMI filter has all necessary parts and it does a good job, in restricting EMI emissions.

The main switching FETs, provided by Infineon, are installed into a half-bridge topology. An LLC resonant converter is also used to boost efficiency.

The APFC converter uses two Vishay FETs and a single boost diode, by On Semiconductor. These are high-quality parts.

The two powerful bridge rectifiers are more than enough to handle the max power that this platform can deliver.


Link: https://www.tomshardware.com/uk/reviews/corsair-rm850x-2021-power-supply-review

[DW1961]

Well, ok, so you would call this power supply noisy?
https://www.tomshardware.com/reviews/corsair-rm850x-2021-power-supply-review/4

Yes, they all are noisy.
I don't have exactly this model for testing, but I have Corsair HX620W. Just measured it, here is it's output noise:


For comparison, here is linear lab PSU KORAD KA3005D:

That power supply is 16 years old. Maybe check this one out?
https://www.tomshardware.com/reviews/corsair-hx850-psu,5257-10.html
(2017 model and I linked to the EMI test page)

The ripply was:
80% Load: 10.2mV

BTW, a line overlay would be more useful so we can see exactly how much difference there is between the perfect linear scope and the SPS. By their very nature, they will never be flat like a linear, but they may be "flat enough." In fact, everything I own runs off of a switching power supply, including my lamps (All 24V+ switching power supplies). Then I have chargers, amp power supplies, TV, monitors, routers, modems, all running 16 hours a day, and I dunno man. There must be something right with them.

As far as safety goes, China puts out millions of these $60-100 dollar power supplies, and they are in use all over the world. Believe me, I'm no fan of crap electronics. However, if they were catching fire or otherwise harming people, I think you'd know about it. Millions of bench supplies all over the world is a pretty damn good statistical safety test. The exception proves the rule.

[radiolistener]

That power supply is 16 years old. Maybe check this one out?

The issue is not it's age, the issue is that this is switching mode power supply. Any switching mode power supply use pulses for voltage regulation. And you can see these pulses on the output.

Very expensive elite branded SMPS has expensive filters, shielding and other measures to suppress these noise. They also have noise, but that noise is lower than for cheap Chinese SMPS.

The better switching mode power supply use MHz band working frequency and have frequency knob, which allows user to shift noise from the working bandwidth.

If you're needs noise free power supply, you're needs to look for linear power supply.

For ultra low noise applications it's better to use battery powered devices with no switching mode converters. And if it needs mains voltage you're needs to use pure sine wave generator with power amplifier, because connecting to mains may lead to RF leakage (from mains to device).

In fact, everything I own runs off of a switching power supply, including my lamps (All 24V+ switching power supplies). Then I have chargers, amp power supplies, TV, monitors, routers, modems, all running 16 hours a day, and I dunno man.

Yes, all these things with SMPS produce RFI. Some Chinese chargers have so high RFI noise that it is better to just put it into garbage bin. 

You can listen all these RFI noise on a short wave receiver with LW and MW band.

[DW1961]

That power supply is 16 years old. Maybe check this one out?

The issue is not it's age, the issue is that this is switching mode power supply. Any switching mode power supply use pulses for voltage regulation. And you can see these pulses on the output.

Very expensive elite branded SMPS has expensive filters, shielding and other measures to suppress these noise. They also have noise, but that noise is lower than for cheap Chinese SMPS.

The better switching mode power supply use MHz band working frequency and have frequency knob, which allows user to shift noise from the working bandwidth.

If you're needs noise free power supply, you're needs to look for linear power supply.

For ultra low noise applications it's better to use battery powered devices with no switching mode converters. And if it needs mains voltage you're needs to use pure sine wave generator with power amplifier, because connecting to mains may lead to RF leakage (from mains to device).

In fact, everything I own runs off of a switching power supply, including my lamps (All 24V+ switching power supplies). Then I have chargers, amp power supplies, TV, monitors, routers, modems, all running 16 hours a day, and I dunno man.

Yes, all these things with SMPS produce RFI. Some Chinese chargers have so high RFI noise that it is better to just put it into garbage bin. 

You can listen all these RFI noise on a short wave receiver with LW and MW band.

What I meant about the PSU being 16 years old is that they get better with each iteration.

Of course I agree with you that is you need an extra quite power supply, then go linear. Ok. Sure. No argument there . I don't need that. My computer power supply will work to power my computer and other devices must fine, and I don't listen for RFI on a short wave radio. We're talking bout consumer electronics here, testing hobby things like LED lights, etc.

My original post was just tossing up the internal of this thing to see if anyone wanted to look inside to see if it was just garbage, from an external point of view, capacitors, soldering, filter use, etc., like so many other people do on the internet.

For example:

Kiss Analog did a review on a Kaiweets (Chinese) bench power supply, and he started by scoping it, then looking at the electronics inside of it, then put it on a scope. Overshoot was almost perfect. He got some ripple on it (about 1.5V peak to peak) but then just added a 1mfd poly capacitor. See below:
Video time stamp 28:35



The post wasn't about linear vs switching.

 It was about how the electronics look inside of the machine, something like tearing down a DMM to inspect it visually.

This isn't necessarily directed to you, but generally to anyone:

 My point in posting was to just have people get a look at it and see what they think (Like Eddie does in the above video, minus scoping it). It seems some people are more interested in simply shutting me down, instead of offering any constructive feedback. Like one poster said, what's with all of the attitude? Geeze.*

*radiolistener, I don't take your comments as attitude at all. It's interesting where you are coming from. It's just not really on topic, but I appreciate your explanations because I'm learning from them (I already knew SPS were noisy vs linear, but the conversation is still productive based on your explanations. I appreciate it).

[Gyro]

...
I don't buy them. I receive them for free for doing a basic review of the products.
...

...
 My point in posting was to just have people get a look at it and see what they think (Like Eddie does in the above video, minus scoping it). It seems some people are more interested in simply shutting me down, instead of offering any constructive feedback. Like one poster said, what's with all of the attitude? Geeze.*
...

Anyone feel that they're doing the work while somebody else is getting the free stuff? 

[Hamelec]

##
 ... It seems some people are more interested in simply shutting me down, instead of offering any constructive feedback. Like one poster said, what's with all of the attitude? Geeze.*
##

fully agreed !

[narkeleptk]

Radio makes a great point. These cheap power supplies can be useful but I will never go back to them as a main unit on the bench after using a linear supply. I used to wrongly blame my picoscope and pc for being so noisy when testing stuff. Had no idea and never thought to think about it just being the cheap power supply I was powering up the devices with it causing all the noise. Not just that but both of the cheap units I have went bad and I had to repair with better quality components. 1 of which was barley used before it failed.

[rsjsouza]

The power supply looks a typical standard issue from China, thus it does not seem to have anything terribly special about it. If the power supply can sustain a load for long periods, it can still be very useful to power beefier loads such as automotive lamps, for example. Also, not every circuit requires ultra low noise and I can also count a large number of development kits (Arduino, launchpads, etc) that could benefit from such power supply powering auxiliary circuits, for example. Just be sure to measure any voltage differences in the ground between the power supply and a laptop where the board is connected (Dave has a good video about this) and you should be alright.

I have one (not this exact model) that has been working quite well for a number of years.

[dophuc]

Well, ok, so you would call this power supply noisy?
https://www.tomshardware.com/reviews/corsair-rm850x-2021-power-supply-review/4

Yes, they all are noisy.
I don't have exactly this model for testing, but I have Corsair HX620W. Just measured it, here is it's output noise:


For comparison, here is linear lab PSU KORAD KA3005D:

You have a cheap, very old and bad PSU so you think all switching PSUs are the same as yours? Totally wrong, try to buy a new good one, for example Corsair RMX series, maybe it will change your mind.

[Trader]

Yes, they all are noisy.

To measure the PSU noise, I need to setup the scope to:

• Coupling AC
• Bandwith of 20MHz
• Probe in 1X (why not 10X?)

And the PSU must have a load to Max Current (10A) and Max Voltage (30V)? (or 90% of max?)

Any other setup?  Thank you for your measures.

[Someone]

I don't buy them. I receive them for free for doing a basic review of the products.

My point in posting was to just have people get a look at it and see what they think (Like Eddie does in the above video, minus scoping it). It seems some people are more interested in simply shutting me down, instead of offering any constructive feedback. Like one poster said, what's with all of the attitude? Geeze.*

Anyone feel that they're doing the work while somebody else is getting the free stuff? 

Not hating on the big pictures like some posters (most of them justify the file szie being in tight focus and showing lots of detail) which add some nice archival material for the forum, but the general vibe is farming out generation of content and editing for their external use/promotion.

P.S. the mains earthing arrangement looks "interesting" and soldered? to the PCB mounting tabs. That would fail most international safety standards.

[tautech]

Yes, they all are noisy.

To measure the PSU noise, I need to setup the scope to:

• Coupling AC
• Bandwith of 20MHz
• Probe in 1X (why not 10X?)

And the PSU must have a load to Max Current (10A) and Max Voltage (30V)? (or 90% of max?)

Any other setup?  Thank you for your measures.

When ripple is well down in the mV's most scopes have insufficient sensitivity to display such ripple with a 10x probe.
Further down into uV's of ripple it's even harder to measure even with a 1x probe............

[Trader]

For comparison, here is linear lab PSU KORAD KA3005D:

I was thinking about buying the KORAD PSU, but I gave up after reading these reviews:

https://www.amazon.com/gp/customer-reviews/R1QZ2T4NK0SU60/

https://www.amazon.com/gp/customer-reviews/R3PQK6U1SYZTL1/

Since you have the same model (KA3005D), please, could you check if your KORAD also has this Issue/Bug? Is this a real problem or maybe just a factory defect of that PSUs in the reviews? Thanks

[Electro Fan]

These are somewhat old reviews and any model from any manufacturer can have a failure. I have two Korad power supplies including one from ~2013 that continues to run fine.  I periodically check for overshoot with an oscilloscope - so far so good, and ripple is very low too, fwiw.  If you have a normal unit which you likely will I think you will be impressed with the performance, ease of use, and reliability.  If you buy it from SRA Soldering (which I think is possible direct from SRA or via Amazon) they will provide a good price and good support.  The Korad power supplies might be the best bang for the buck on my bench which includes gear from Rigol, Tektronix, HP/Agilent/Keysight, Brymen, and Fluke - all of which I am happy with but the point is $ for $ the Korad KA3005 (D and P) hold their own.

[J-R]

Switching power supplies run our current electronics.  They are everywhere.  If you want a switching or linear supply on your bench, then buy what you want.  Each have their place.  I have one switching supply on my bench in order to get 15V 40A, but rarely ever use it, preferring my linear supplies otherwise.  Bashing one or the other endlessly is silly.

My Korad supply does not have that voltage spike issue, but it did have a very loud fan that I had to replace with a temperature controlled one for my sanity.

But there is always perhaps some heightened risk with the budget brands and random sellers.  You may get some old stock with issues for example.

I have a couple of the Keysight power supplies and have been happy enough with their performance.  This has been sitting on ebay for quite some time:
https://www.ebay.com/itm/194385942444
Cute little supply with USB and ethernet.  Maybe it is what someone is looking for.  Ignore the 100V option if you're in the US, there are simply two switches on the bottom you can change.  An internal fuse change is only required when going between 100/120 and 230.


Back to the original power supply being discussed, yes, it has me concerned somewhat.
I have moved past such equipment in general.  But a tinkerer on a budget might go for it.

[Electro Fan]

Korad spike and ripple measurements in this thread:
https://www.eevblog.com/forum/testgear/korad-kd3005d-bench-supply-power-on-spike/msg3509200/#msg3509200

[Monkeh]

I don't buy them. I receive them for free for doing a basic review of the products.

My point in posting was to just have people get a look at it and see what they think (Like Eddie does in the above video, minus scoping it). It seems some people are more interested in simply shutting me down, instead of offering any constructive feedback. Like one poster said, what's with all of the attitude? Geeze.*

Anyone feel that they're doing the work while somebody else is getting the free stuff? 

Not hating on the big pictures like some posters (most of them justify the file szie being in tight focus and showing lots of detail) which add some nice archival material for the forum, but the general vibe is farming out generation of content and editing for their external use/promotion.

He's been doing this for a while.. receiving piles of free stuff from Amazon sellers and using the forum to help him create glowing 'reviews' of tools he doesn't really know how to use.

[DW1961]

Radio makes a great point. These cheap power supplies can be useful but I will never go back to them as a main unit on the bench after using a linear supply. I used to wrongly blame my picoscope and pc for being so noisy when testing stuff. Had no idea and never thought to think about it just being the cheap power supply I was powering up the devices with it causing all the noise. Not just that but both of the cheap units I have went bad and I had to repair with better quality components. 1 of which was barley used before it failed.

Yeah no problem with that logic. I'm just using it to test stuff like LED lighting and things like that. Nothing that requires s super quite supply. I mean, I'm mainly testing DC stuff that is powered by a computer power supply, so, yeah, it's ok. haha.

[DW1961]

The power supply looks a typical standard issue from China, thus it does not seem to have anything terribly special about it. If the power supply can sustain a load for long periods, it can still be very useful to power beefier loads such as automotive lamps, for example. Also, not every circuit requires ultra low noise and I can also count a large number of development kits (Arduino, launchpads, etc) that could benefit from such power supply powering auxiliary circuits, for example. Just be sure to measure any voltage differences in the ground between the power supply and a laptop where the board is connected (Dave has a good video about this) and you should be alright.

I have one (not this exact model) that has been working quite well for a number of years.

Yeah that, and I will only have it on to test devices  and not be powering anything from it for any period of time at all. I like to build computers and so I want to test specific runs of fans or LEDs for amps before adding more power to a circuit. Since most everything is powered from an SATA connector, and they SATA connectors are only my good for 4.5A, lest they melt, a 5A supply will work fine for me.

My latest build is below. You can't see the work that went into this by looking at it, but all of the 12V RGB runs were cut and soldered to size for their specific location. Then neodymium magnets were attached to the strings to hold them in place. The 140mm fans were dark gray, and the ARGB fan covers were also dark gray. Since the build is a white theme, I meticulously painted each fan and fan ARGB cover using a makeshift paint booth. Then re-sanded any imperfections, repainted, re-sanded. All of the screws were painted white in the same manner. The top grill cover over the radiator was an aluminum vent cut to size, sanded, and then also painted white in the same manner. The radiator and radiator fans were already white, so I didn't need to do those. However, I wanted 140mm Noctua fans, and they only come in dark gray and doctor office brown/tan. (I didn't paint the fan blades in order to maintain their balance.) The bottom pedestal is a repurposed acrylic monitor stand that I added a custom soldered and sized RGB lighting strip to the edge after sanding and diffusing the edges. I then found shorter legs for it, turned it upside down, and used it like it is in the pictures. In short, a lot of hours went into it.

My latest iteration:

[DW1961]

Well, ok, so you would call this power supply noisy?
https://www.tomshardware.com/reviews/corsair-rm850x-2021-power-supply-review/4

Yes, they all are noisy.
I don't have exactly this model for testing, but I have Corsair HX620W. Just measured it, here is it's output noise:


For comparison, here is linear lab PSU KORAD KA3005D:

You have a cheap, very old and bad PSU so you think all switching PSUs are the same as yours? Totally wrong, try to buy a new good one, for example Corsair RMX series, maybe it will change your mind.

That PSU he has was a VERY high end unit when it came out. Also, any ripple might be noisy to him for his applications. I see what he is saying, but there are many other uses for a switching bench supply, for noncritical testing purposes, right?

[DW1961]

Switching power supplies run our current electronics.  They are everywhere.  If you want a switching or linear supply on your bench, then buy what you want.  Each have their place.  I have one switching supply on my bench in order to get 15V 40A, but rarely ever use it, preferring my linear supplies otherwise.  Bashing one or the other endlessly is silly.

My Korad supply does not have that voltage spike issue, but it did have a very loud fan that I had to replace with a temperature controlled one for my sanity.

But there is always perhaps some heightened risk with the budget brands and random sellers.  You may get some old stock with issues for example.

I have a couple of the Keysight power supplies and have been happy enough with their performance.  This has been sitting on ebay for quite some time:
https://www.ebay.com/itm/194385942444
Cute little supply with USB and ethernet.  Maybe it is what someone is looking for.  Ignore the 100V option if you're in the US, there are simply two switches on the bottom you can change.  An internal fuse change is only required when going between 100/120 and 230.


Back to the original power supply being discussed, yes, it has me concerned somewhat.
I have moved past such equipment in general.  But a tinkerer on a budget might go for it.

J-R, very interested what you see that is concerning you. Please elaborate? I can post more pictures if you need. Just let me now what you need to see.

[DW1961]

He's been doing this for a while.. receiving piles of free stuff from Amazon sellers and using the forum to help him create glowing 'reviews' of tools he doesn't really know how to use.

Why the personal attack?

[DW1961]

I don't buy them. I receive them for free for doing a basic review of the products.

My point in posting was to just have people get a look at it and see what they think (Like Eddie does in the above video, minus scoping it). It seems some people are more interested in simply shutting me down, instead of offering any constructive feedback. Like one poster said, what's with all of the attitude? Geeze.*

Anyone feel that they're doing the work while somebody else is getting the free stuff? 

Not hating on the big pictures like some posters (most of them justify the file szie being in tight focus and showing lots of detail) which add some nice archival material for the forum, but the general vibe is farming out generation of content and editing for their external use/promotion.

P.S. the mains earthing arrangement looks "interesting" and soldered? to the PCB mounting tabs. That would fail most international safety standards.

Good information and I'll check that out again. I did see that too when I opened it. I think someone else said that they didn't like the other one I posted because it had the mains screwed down on the board instead of soldered-lol. Interesting though, so soldering the mains to the board is a no go?

[radiolistener]

Since you have the same model (KA3005D), please, could you check if your KORAD also has this Issue/Bug? Is this a real problem or maybe just a factory defect of that PSUs in the reviews? Thanks

I think this is not PSU issue, but RF interference from power-on switch due to high current sparks in the mains switch. This interference has high amplitude because your scope has high impedance.

Here is KORAD KA3005D 12 V / 5 Amps transition on the output when you press ON button: