Choosing Bench Top Power Supply

[kcs]

Hi,
I am willing to invest in a good bench top power supply. I have chosen these two:
1. EA-PS 8032-10 (manual)
2. EA-PSI 8032-10 (manual)

What can you say about them? Are they good choice for this price range? I do not want to buy a huge power supply (like this one).

[alm]

I missed your requirements, what will you use it for? GBP 400+ seems quite expensive to me for a lab supply without any fancy features, although 10A pushes the price up a bit. Just a single output seems kind of limited, since it's quite common these days for circuits (eg. FPGAs) to run on multiple voltages.

Just took a quick look at the specs, and the ripple looks horrible: 50mV. Most linear lab supplies are something like <1mV or <3mV. The small size requirement means you're limited to switching supplies, which tend to produce more noise, are more complex and can have some stability issues (I know some of them really hate getting RF in their outputs).

[saturation]

Excellent but expensive lab grade SMPS power supplies with a lot of digital interface options and 300W+.

Major trades off is this is clearly a SMPS, and its common for ripple >> linear design PSU for equivalent power rating.

More on the differences, you can use those values as sample specs to try to match:

https://www.eevblog.com/forum/index.php?topic=2002.msg27350#msg27350

As alm states, it common for electronic designs to have multiple voltage requirements, and if a basic clean adjustable supply is what you need, you can get several low cost linear ones than a single 300W SMPS. 

[Ernie Milko]

I see you're in the UK, I would look for a Farnell E30-2, 2nd hand, it'll set you back about fifty quid. It's pretty much oops proof and easily fixable.
It's also dual output; unlike the ones you listed.
I've been using one for 30 years or more.

Spend the remainder of your five hundred quid on a (2nd-hand analogue) Tek scope.
You should have some money left over.......................buy a Fluke meter and lots of beer........................

[kcs]

I am a hobbyist. I do not expect that my investments in any of electronic equipment will ever pay off. I do it for fun but I prefer to work with good instruments. That’s why I bought for example Fluke 87V instead of some cheap alternative.

Of course I would prefer 2 outputs instead of 1, but I could not find SMPS that meets my requirements and has 2 outputs. I want power supply, which meets these requirements: 0-30 Vdc, >=6A, PC controlled, relatively small.

Linear power supplies with 2 outputs are simply too big for me (look at this monster!).

[TopherTheME]

Have you checked out the supplies at Circuit Specialists?

http://www.circuitspecialists.com/level.itml/icOid/2795

They seem to have quite a few good deals. I'm thinking about getting one of their smaller ones for home use.

http://www.circuitspecialists.com/prod.itml/icOid/18

[Bored@Work]

Your power requirement (2 * 30 V * 6 A = 2 * 180 W = 360 W) makes it difficult to find a small one. I don't agree that the ones you keep rejecting are "monsters".  360 W output at, lets say 80% efficiency (mediocre SMPS), means 90 W of heat need to be dissipated.

I think you will have to adjust your requirements, either the size or the power requirement.

Do you really need 6 A? E.g. look at a UniSource21 P6100, half the current you request, but just 235 * 296 * 85 mm³. Triple output, -30 V / 3 A adj., +30 V / 3 A adj., 5 V / 2 A fixed.

[kcs]

I think I will consider then power supplies with 1 output. At the moment I am having an old 15V, 1A power supply, so it could be used along with a new one when I need a second output.

The question is, how bad is 10mV rms (or 100 mV pp) ripple, in which cases it is a big problem? alm noted that linear power supplies have only 1mV or 3mV ripple. Does it mean that no matter how good SMPS is, it will never have such a small voltage ripple?

Can anyone suggest any similar SMPS to the ones I have chosen?

[saturation]

If you know exactly what you want you can do the parametric search, there are cheaper models of that German PSU, at £200 each,

http://uk.farnell.com/ea-elektro-automatik/ea-ps-2042-06b/psu-0-42v-0-6a/dp/1840120

and they are computer controlled, the only issue is its electronics specs are not detailed enough as a PSU.

Likewise, if you don't need computer control, there are many dual PSU for lower cost.

http://uk.farnell.com/ea-elektro-automatik/ea-3048b/psu-bench-2-output-euro-uk/dp/1437527

You said 'linear ...big for me' but your choice is 2 slightly smaller PSU each with a volume of ~ 5.8L x 2 = 11.6L  that together, are far bigger than the ' ..monster' you linked, of 7.2L.

I am a hobbyist. I do not expect that my investments in any of electronic equipment will ever pay off. I do it for fun but I prefer to work with good instruments. That’s why I bought for example Fluke 87V instead of some cheap alternative.

Of course I would prefer 2 outputs instead of 1, but I could not find SMPS that meets my requirements and has 2 outputs. I want power supply, which meets these requirements: 0-30 Vdc, >=6A, PC controlled, relatively small.

Linear power supplies with 2 outputs are simply too big for me (look at this monster!).

[saturation]

Most basic low cost, linear PSU have no issues with taking 2 or more and linking them together in series to increase the voltage, or in parallel to increase the current. However, if they have a form of digital control, such as its programmable or has memory, it may not work in series or parallel unless it says so on the manual; if its a simple linear design chances are it can be stacked, even if the manual doesn't say so.  This is often tied to whether internal circuits are earth grounded or float with voltage output.

Computer control is rarely needed by hobbyists, or even home labs.  Its usually found in automated testing, and you'd prefer a secondary monitor to insure the computer commands are correctly executed. If it glitches and you get incorrect output, how will you know? Its always best to manually check output voltages when designing, and if you do check it manually, the computer control is less helpful.

Ripple is a problem with lower voltage and currents needed in circuit, such as if you design for AA battery power.  As ripple and or noise is constant, it becomes a larger percentage of fluctuation as the required voltage drops.

As for the inefficiency of linear, that is true, but unlike an appliance its not on all the time, only when you are testing or troubleshooting.

A SMPS is mostly noisier, even the expensive ones.  At higher operating voltages, the ripple becomes less an issue, or often the output of the SMPS is further cleaned by added input filters and a linear regulator on the PCB of the circuit.

I think I will consider then power supplies with 1 output. At the moment I am having an old 15V, 1A power supply, so it could be used along with a new one when I need a second output.

The question is, how bad is 10mV rms (or 100 mV pp) ripple, in which cases it is a big problem? alm noted that linear power supplies have only 1mV or 3mV ripple. Does it mean that no matter how good SMPS is, it will never have such a small voltage ripple?

Can anyone suggest any similar SMPS to the ones I have chosen?

[alm]

The question is, how bad is 10mV rms (or 100 mV pp) ripple, in which cases it is a big problem?

Depends on application, which you failed to tell us. Probably not a big deal for digital (assuming enough noise margin) or stuff like motors, could be a big deal for analog, like ADC/DAC. For a 3.3V ADC/DAC, 100mVp-p noise will limit the effective number of bits to 6. For an 8-bit part, the least significant two bits will be garbage. Let alone higher resolution parts. But you typically don't need 10A for this kind of work either, so you wouldn't typically choose a SMPS.

alm noted that linear power supplies have only 1mV or 3mV ripple. Does it mean that no matter how good SMPS is, it will never have such a small voltage ripple?

I'm sure it's possible, put you have to work a lot harder to achieve this. A linear supply (in theory) doesn't produce any noise, the only noise on the output is either from the mains or from the load (in practice everything produces some noise, even resistors and capacitors). A SMPS converts the mains to a pulse train with fast edge rates to achieve high efficiency (MOSFETs are inefficient during the transition). This means tons of odd harmonics (infinite for the perfect square wave). You can filter this, but you're making it harder by first adding lots of AC and then trying to filter it.

[kcs]

So, now I am left with these two:
- Agilent U8002A (not PC controlled)
- BK Precision BK1788

Please recommend which one to choose out of these two. Agilent seems to make better quality products, while I am not sure about BK Precision. Should I choose Agilent's low-end PSU?

[saturation]

Agilent.  But that's still a lot of money for these types of supplies for hobbyist use.  They are both reputable brands, the B&K is costly due to the extra programming capability.

I personally would look into something closer to this.

http://uk.farnell.com/ea-elektro-automatik/ea-ps-2032-050/psu-bench-32v-5a-euro-uk/dp/1437494

So, now I am left with these two:
- Agilent U8002A (not PC controlled)
- BK Precision BK1788

Please recommend which one to choose out of these two. Agilent seems to make better quality products, while I am not sure about BK Precision. Should I choose Agilent's low-end PSU?