Replacement Bench Supply

[ACharnley]

Hi All,

I've used a NRP-2050 for many years (https://www.manson.com.hk/product/nrp-2050/), it's a cheap power supply but has done the job, usually.

Today I've had a prototype smd circuit which was likely to have VBUS connected but not GND (I can't recall the order that I was connecting the wires) and when connecting only the GND to the power supply negative there was a small crackle indicating high current flow. After this one of the buck circuits on the prototype has blown, and it's not the first time but until today I didn't realise what was the cause.

After two things here;

1. suggestions why would this occur (one wire suggestions the supply is not isolated)?
2. suggestions for good quality power supplies (can spend a bit as I want something decent).

 

[JimB]

1. suggestions why would this occur (one wire suggestions the supply is not isolated)?

Looking at the pictures and the specification, the output of the PSU is floating, ie neither the positive or the negative output terminal is connected to ground (mains earth).

This is OK, it makes for a more versatile PSU, the user can connect either the negative or the positive output terminal to earth, or let the whole thing float above earth with some at a voltage provided by some other PSU.

But, there can be a little trap for the unwary, due to stray capacitance, especially the capacitance between primary and secondary of the mains transformer in the PSU, the output can be floating at some voltage above ground.
Usually this floating voltage is AC at mains frequency.
Use your multimeter and measure it.

I have come across this problem a couple of times in my home made PSUs.
The way that I have overcome the problem is to connect a 1uF non-polarised capacitor between mains earth and the negative output terminal.
This kills any stray AC.
Also connect a high value resistor, say 1 meg Ohm, to prevent the capacitor becoming charged with DC.

2. suggestions for good quality power supplies

Cannot really help there, all my PSUs are home made.
Typical example here:

Your thoughts on what makes a good and versatile bench power supply?

Hi to everyone, Further to another thread I started it was suggested that I start a discussion about what people like, want and expect from a bench power supply. Please feel free to add any ideas, wishes or even gripes about bench power supplies, and I mean anything you consider pertinent or...

www.electro-tech-online.com

JimB

 

[ACharnley]

Well I didn't expect that, ~+50v measured between negative and Earth (discharging like a capacitor) and about the same from positive to earth.

I take it two capacitors and bleed resistors to Earth?

I've researched but Manson seems to be by far the largest supplier with tons of rebranding (RS, Farnell etc).

 

[JimB]

I take it two capacitors and bleed resistors to Earth?

Just one capacitor and bleed resistor should be OK.
The pos and neg terminals of the PSU are tied together by... the PSU.

JimB

 

[ACharnley]

Of course, shouldn't have asked that.

Job done, now floating close to ground.

 

[Nigel Goodwin]

Well I didn't expect that, ~+50v measured between negative and Earth (discharging like a capacitor) and about the same from positive to earth.

That's what 'floating' means

However, there's no reason to suspect that's what caused your problem, as it's worked fine for years and is perfectly normal.

 

[gophert]

It may have works for years until he connected a device to his PSU that had one grounded component - (e.g. a scope) and he started to get his unexpected crackle of current flow as he connected the low side to a now-grounded PSU.

 

[Nigel Goodwin]

It may have works for years until he connected a device to his PSU that had one grounded component - (e.g. a scope) and he started to get his unexpected crackle of current flow as he connected the low side to a now-grounded PSU.

Except the 'current' is extremely low, and unlikely to cause any problems.

 

[gophert]

Except the 'current' is extremely low, and unlikely to cause any problems.

That's exactly why I don't wear anti-static wristbands, and I am sure to hold the gate with one hand and the source pin with the lot her as I drag my feet across my wool carpets in the winter. It's very low current - and CMOS devices only respond to current - WHAT?

Pull your head out! He measured 50v of float today, it could be 1500 tomorrow. Either way, it was way above the "absolute max " input voltage or supply voltage of most Modern logic or MCUs I've used. "Very low current" ha. Oh, good thing your comment is so rediculous, you'll be able to say, "I was obviously joking". Good grief.

 

[ACharnley]

It's not the first time I've had this occur - and it does tend to take out buck/boost chips for some reason. The 50v goes down quite rapidly but I'd wager initially it was lot higher prior to the multimeter (6KHz sampling) doing it's first read.

Question, why don't they do this as part of the design, is it because the outputs are no longer fully isolated?

 

[gophert]

It's not the first time I've had this occur - and it does tend to take out buck/boost chips for some reason. The 50v goes down quite rapidly but I'd wager initially it was lot higher prior to the multimeter (6KHz sampling) doing it's first read.

Question, why don't they do this as part of the design, is it because the outputs are no longer fully isolated?

yes, they want you to knowingly connect as needed - it becomes more versatile. I assume something in your setup is earth grounded to cause the spark sound when you connect he negative lead. Either your scope, signal generator something else on your bench that is plugged into AC power. Possibly even a bench meter?

without your path to ground, an isolated power supply can be stacked in series with a secon supply to make a bipolar supply or simply add the output voltage of the two or more supplies.

 

[upand_at_them]

I got this recently, and like it. Not brand-name high quality, but the one I have is the second or third revision, so they appear to have fixed issues with the original design. Many people have this. Several threads on EEVBlog about the original version. The non-programmable version is $40 less.

Amazon - KORAD KA3005P - Programmable Precision Variable Adjustable 30V, 5A DC Linear Power Supply Digital Regulated Lab Grade