Looking for a benchtop power supply, question about GPR-3060D and PSP-2010

[SmallWalrus]

I'm actually looking to get a benchtop power supply to save me from having to buy and wait for a new power adapter to arrive when a strange laptop shows up on my desk, and I am also hoping to be able to glean more information about what fault it might be suffering from by monitor its current draw during the boot up process.

The kind of power supply that I'd need will have to be able to supply around 20V, and to be on the safe side 5A or so. Since very few manufacturers have units with 20V output, this bumps me up a bit higher to the 30V/5A range of power supplies.

Was initially looking as Mastech, but an earlier video posted by another EEVBlog member showing that supply overshooting on power up made me rather worried, if a motherboard is fed 25V for 1s, it is gone. Hence I've decided to spend a bit more in hopes of getting a better unit, and my choice got narrowed down to the Instek GPR-3060D and PSP-2010.

What puzzled me is that the GPR-3060D being a non-programmable linear power supply, costs around 40% more than the PSP-2010 ($375 vs $275). Am I missing something? Aren't switching power supplies supposed to cost more? How is it then that a programmable switching power supply costs less than a linear model?

[alm]

These days with capacitor technology (polymer organics) SMPS can be low ripple the one you link specs 20mVrms

20 mVrms ripple is low these days? The linear GPR-3060D is specced < 1 mVrms, and this is quite average for a linear power supply. Even a decent ATX supply can meet 20 mVrms, and they're not known for their clean output.

Another issue is that some SMPSs really hate having RF leaking into the power supply, which may be an issue if you're working on an RF circuit. If you want lots of power in a small/affordable package, and don't care that much about output, SMPS is the way to go. For analog electronics were noise may be critical, I still prefer linear power supplies. Also because they're simpler (easier to fix). For laptops, power quality probably doesn't really matter, all those power bricks made for laptops in the last fifteen years or so are switching, and not particularly high quality.

One issue is that many manufacturers put some sort of EEPROM in their power bricks that identifies them to the laptop. These laptops may not work from a simple voltage source.

[SmallWalrus]

One issue is that many manufacturers put some sort of EEPROM in their power bricks that identifies them to the laptop. These laptops may not work from a simple voltage source.

Good point. Apart from the PSP-2010 bench top power supply, will there be any other idea whereby I can use an existing laptop power supply and still get to monitor real time current draw by the laptop though?

I imagine that various inefficiencies in the laptop's power supply may isolate away some of what's happening, but I am hoping that there might be another way short of splicing a multimeter into each brand's laptop power supply.

[alm]

What some people tend to forget when they get hung up on ripple is the test is done with a resistive load. So if you are just planning on hooking resistors up to it under the specific test conditions fine if you plan to actually use it the ripple will vary regardless of SMPS or linear.

Is your claim that an SMPS will have a better transient response and have a lower ripple with a dynamic load?

RF can cause abnormal behaviour on any electronic device if it’s strong enough and the device isn’t adequately shielded.

Some SMPSs are much more susceptible than the linear supplies I've seen. Lab supplies are often used to power an unfinished circuit without a case.

Good point. Apart from the PSP-2010 bench top power supply, will there be any other idea whereby I can use an existing laptop power supply and still get to monitor real time current draw by the laptop though?

You could monitor the primary side with something like a Kill-A-Watt type device. Of course the quiescent draw and inefficiency of the power supply will introduce errors. Just measuring the current will introduce additional errors compared to a KAW which measures the instantaneous product of voltage and current due to a non-unity power factor, especially at low power levels.

I imagine that various inefficiencies in the laptop's power supply may isolate away some of what's happening, but I am hoping that there might be another way short of splicing a multimeter into each brand's laptop power supply.

Measuring current will requiring either modifying each power supply or having a modified extension cord for each type of plug. Those universal power supplies they sell might be useful. You either need to open the circuit for a measurement with a DMM or an external shunt, or separate the conductors so you can clamp a current probe around just a single conductor.

[saturation]

I've found that a linear supply can provide transparent performance for most all facets required by DC designs, although my experience is limited.  I recall Agilent literature saying this too, when making general statements between linear vs SMPS.

Even cheapo Chinese linear ones can perform very well because the designs are robust and well tested [ all one need worry about is assembly and the parts quality!].  If you know how to test PSU, you are set, using Agilent guidelines are helpful and well documented.

Search for the title, "Power Supply Testing (AN 372-1)" via google to link to the proper pdfs.



 

[saturation]

Thanks AcHmed99, yes and neither am I being a linear zealot.  You can get SMPS as good as linear, but in regards to the topics lab bench supplies, most are typically like the OPs original query: noisy.

Will that matter?  I guess that depends on what you are doing. 

Can you share the reference you took these specs from?  This looks like a commercial SMPS more than a lab supply.

Lastly to the OP, on the Mastech, YMMV with Chinese label supplies.  A brand name like Instek, will help you if you can't test the device after purchase to see what it can do.  I got a Mastech that is designed like few you see for sale on the net, but available only from:

mastechpowersupply.com

I've checked other sellers, none seem to look closely like his offerings in the linear analog catalog.  These are more suitable as electronic lab supplies than the look alikes, they will cost about 50% more than many others, which are popular for use in tattoo parlors.

The older models are better made, and will give ~ no overshoot, and same noise specs as the Instek.  You can find review and pics I posted on eevblog.  Its priced from $50 for 18V, 3A.

I’m not saying linear aren’t superior in some aspects but an SMPS variable output has some advantages as well. For instance a linear regulator typical 0-30VDC 3A output only can supply the full 90W at 30V out and if you wanted to use it for testing 1.8V or 3.3V high current your SOL. With an SMPS 90W it may still be able to supply 10A down at 1.8V TO 3.3V. They both have there pros and cons. Personally I don’t think SMPS are as horrible for variable lab supplies as some people think. Size weight and cost is a definite plus for an SMPS.

I haven’t looked at the Agilent paper but I’m sure its old a well designed MHz smps can equal or surpass 20uS see the screen shots below. That’s with a couple of minutes of looking I know there is better out there.

[SgtRock]

Dear SmallWalrus:

--Check out the Heathkit 2718 Tri Power Supply on eBay. It should meet your specifications
and you can get one cheap. I have two and I love them.

""O Oysters," said the Carpenter,
"You've had a pleasant run!
Shall we be trotting home again?'
But answer came there none--
And this was scarcely odd, because
They'd eaten every one."
From The Walrus and the Carpenter
by Lewis Carroll (Charles Lutwidge Dodgson)
1832-1898
Best Regards
Clear Ether

[SmallWalrus]

Dear SmallWalrus:

--Check out the Heathkit 2718 Tri Power Supply on eBay. It should meet your specifications
and you can get one cheap. I have two and I love them.

""O Oysters," said the Carpenter,
"You've had a pleasant run!
Shall we be trotting home again?'
But answer came there none--
And this was scarcely odd, because
They'd eaten every one."
From The Walrus and the Carpenter
by Lewis Carroll (Charles Lutwidge Dodgson)
1832-1898
Best Regards
Clear Ether

The Heathkit looks surprisingly good. One more question about it though, does it take a universal AC input? (ie: a switchable 110V/220V mains voltage)

[SmallWalrus]

You realize that the heathkit is a far cry from the 20V 5A you said you needed?

I just noticed the 0.5A... 

[saturation]

Thanks Achmed99, I couldn't agree more.
The Heathkit also has no constant current capacity, at the least.

Hi Saturation

The first one is a commercial buck I cant find that now but the second one is just the AN for a National (now TI) demo board. Its here.
<snip>
Bottom line it depends on what you will be mainly using it for and what you're willing to pay.

Yes, but you can stack the tracking outputs in series to double the voltage, or in parallel, for current.  The main design was made for about 1A.

You realize that the heathkit is a far cry from the 20V 5A you said you needed?

I just noticed the 0.5A... 

[SgtRock]

Dear SmallWalrus:

--Please forgive me for my erroneous posting. AcHmed99 is indeed correct. Good hunting.

"He was born ignorant, and has been losing ground ever since."
Fred Allen 1894 1956

Best Regards
Clear Ether

[bfritz]

I think you will find the more problematic issue than buying a power supply, will be building a collection of the various sizes and types of connectors that fit the input power connector on the laptop.

I have used some of the Kensington and Targus products.  The nice thing is that some of their units come with multiple tips that you simply connect to the cord, and the tips alter the output voltage, and set the polarity correctly.  I believe Kensington also has a number you call, where you can get additional tips if you find a laptop that isn't supported by the tips that came with the charger.  I believe they will give you several for free, before asking for a small fee for them.

If you wish to use something like the above, and want the ability to monitor power levels, it would be fairly easy to cut a cable and add a DMM in line.  You would need to look carefully at the burden voltage of the meter you use, as at currents in the several amps, this could get large enough with some meters to cause problems with the laptop not having enough voltage.  (Caused by the voltage drop in the meter put in series to sense current.)

The other possible thing for power sensing, is to sense the input power to the Kensington or Targus you buy.  A unit like the "Watts Up?" or the "Kill-A-Watt" would likely serve your purpose.  I don't remember which had the better specs off hand, so do look at that before buying if you go that direction.

Anyway, this is an area that I've done a fair amount of work in, and I would suggest the Kensington units for you, as the connector problem is a real issue if you don't have access to all the connectors for the various laptops.  Unfortunately there are a huge variety.

If you do go the route of a Kensington or Targus, you will probably need to buy a higher wattage supply from Apple to support the MagSafe connectors on their products, as that connector is protected by patents, and as a result Kensington and Targus can't make connectors for those products.