Super precise PSU, what do buy?

[FriedMule]

I have already your "run-of-the-mill" power supplies that have 30V 5A.

What I need is a super precise PSU with minimum 5V 100mA but preferable 10V 1A.
I am hoping on being able to adjust voltage and current in minute steps, i.e. 0.01V - 0.001A

It shall be used for low power circuits and detecting their real cut-off power, behavior on low battery and stability.

[Micke]

Could a HP 6632B be sufficient?

Average programming resolution:
5 mV and 1.25 mA respectively.

Hmm, also appears in datasheet "Minimum current in constant current mode = 20mA" 

Very clean power and can be found cheaply with some luck...

https://www.keysight.com/se/en/assets/7018-03200/data-sheets/5990-9303.pdf

[gamalot]

HP/Agilent 66319B/66319D/66321B/66321D and my favorite one - 6612C

[tautech]

I have already your "run-of-the-mill" power supplies that have 30V 5A.

What I need is a super precise PSU with minimum 5V 100mA but preferable 10V 1A.
I am hoping on being able to adjust voltage and current in minute steps, i.e. 0.01V - 0.001A

1mV/mA resolution is standard in the very low noise remote sense SPD1000X range. 16V/8A and 30V/5A models

[nctnico]

HP/Agilent 66319B/66319D/66321B/66321D and my favorite one - 6612C

That would be my suggestion as well. Don't know about the 6612C but the others are intended to simulate a battery in battery powered equipment; they can also act as a (constant voltage) DC load.

[gamalot]

HP/Agilent 66319B/66319D/66321B/66321D and my favorite one - 6612C

That would be my suggestion as well. Don't know about the 6612C but the others are intended to simulate a battery in battery powered equipment; they can also act as a (constant voltage) DC load.

I have both 66319B and 6612C, yes, 66319B has more features, it can simulate battery and load, but I still prefer 6612C in daily use. The 6612C has front panel terminals, a shorter chassis, and less noise (I replaced the 6612C's fan with a low-noise one to make it quieter).

[nctnico]

I have replaced the fan in my 66311B and fitted it with front panel posts. Now it is a nice bench PSU (albeit having a long chassis).

[J-R]

I have already your "run-of-the-mill" power supplies that have 30V 5A.

What I need is a super precise PSU with minimum 5V 100mA but preferable 10V 1A.
I am hoping on being able to adjust voltage and current in minute steps, i.e. 0.01V - 0.001A

It shall be used for low power circuits and detecting their real cut-off power, behavior on low battery and stability.

What PSUs do you have now?

Even some of the popular Korad supplies can do 10mV and 1mA.  My specific sample of KA6003P is typically within about 2mV (0.01V display resolution) and right at 1mA (1mA display resolution).

Be aware that power supplies have different program/set-point specs from the read-back specs.  Then there are also specs for stability and noise.  Some of these are not always listed so you might have to check reviews.

[J-R]

1mV/mA resolution is standard in the very low noise remote sense SPD1000X range. 16V/8A and 30V/5A models

https://siglentna.com/power-supplies/spd1000x-series-programmable-dc-power-supply/

Specs are 0.03% + 10mV and 0.03% + 10mA for setting and read-back accuracy.  Kind of rough.  What's real-world?

[FriedMule]

Thank you all, I'll take a look at them now! :-)

[GerryR]

You might want to look at an older Power Designs 2005A precision voltage source.  They have two ranges 0-10V and 0-20V @ 500 mA.  You can download a manual online; if you can't, let me know, and I'll attach a file to a post.

[FriedMule]

I like your suggestions, what I am uncertain about is the ripple values. If I chose 1.19V 15mA and the ripple is i.e. 0.5V 10mA, does that not mean it can go up to 1.69V 25mA? Is a 0.3V 5mA resolution then not below the real world result? :-)

[GerryR]

If you had 1.19V @ 15mA, then your load would be 79 .  If you had +- .5V ripple, then the current would fluctuate between approx. 8mA and 21mA. (That much ripple on a 1.19V "supply" would be a poor design.)  The Power Designs precision supply that I mentioned above has < 100 microvolts ripple and noise from a 60Hz supply line, so even at a 1mV output that is only a +-10% variance.

[kwass]

I typically have the same needs as you do and use the BK 1739: https://www.bkprecision.com/products/power-supplies/1739-low-current-high-resolution-dc-power-supply-30-v-999-9-ma.html

[TheBay]

Have a look here too:
https://www.aimtti.com/product-category/dc-power-supplies

[Kean]

I agree with what gamalot and J-R said!

I usually would use my old HP 6612B for these types of test, although because it is in desperate need of a new fan I'll often just use the Tonghui TH6412 or my EEZ BB3 as they are already powered on most of the time and are also a bit easier to adjust up and down.

If I want really low noise and simple mV control dials, then I would use my Power Designs 2020B - but it only has simple current limit knob and a small analog meter.

Most PSU's don't have accurate low current read-out (HP 66xx/663xx and EEZ BB3 are exceptions).  If you want precise current read-out then you probably should use a separate ammeter, but keep in mind any burden voltage that adds.

[justanothername]

There is an interesting DIY solution, unfortunately about everything is in german (the UI is in english though).
There was (is?) a diy measurement system called C't Lab, that included a programmable power supply called DCG.

https://www.heise.de/ct/artikel/Kleinkraftwerk-291224.html
https://ctlabwiki.thoralt.de/wiki/index.php/DCG

With the 16 bit version, I guess you'll get the resolution you want. I own a complete system myself and use it regularily.
Surprisingly enough, you can still order kits for it:
https://www.segor.de/INFO/ct-lab/baugruppe-dcg.shtml

The precision of this PSU is based upon the quite tedious calibration routines you have to do after completing the build. The results are quite rewarding then.
There is still an active user group online here:
https://ctlabforum.thoralt.de/phpbb/index.php

[LazyJack]

If you do not need computer control, then the HP 6114A is an outstanding precision power supply. For precise current limit, you want it with the 10 turn current pot option.

[alm]

If you don't mind the size, then the HP/Agilent 6625A/6626A/6628A/6629A 2-4 channel precision power supplies are also pretty accurate and stable, and are programmable via GPIB. They can sometimes be found pretty cheaply on eBay. I think I paid $25 + shipping.

[PaulAm]

Another vote for the HP 6114a if you don't need computer control.  Although that would be possible with a small adapter, there is a programmable input for use with an external voltage.  HP made some adapter that you could cobble together or just use a microcontroller + DAC.

[paulbt]

Hi !
@FriedMule
Since you did not mention anything about the budget $, I will recommend you an Aim-TTi PL303 power supply, high quality, super reliable, linear, super low ripple even at high loads, 1mA or 0.1mA resolution, possibility to change the voltage range from 0-30V to whatever low and high level you wish (example 5-10V) and so your fine voltage knob will become finer, but the resolution of the voltmeter remains 10mV.
 

[HKJ]

For precision I use a Keithley 2280S, I am not going to list specifications, they are good and the PS is not cheap.