What is lab PSU output impedance ? (Inductance component)

[BravoV]

Well, I sort of understand that common simple term when talking about output impedance for a power supply, that lower means better ("ideally/theoretically" should be zero right ?), while higher impedance (which mostly bad) means the load will not receive a good regulation, CMIIW. 

Recently I just acquired an old HP 6115A lab adjustable psu which is so called precision psu, and it has this specification (quoting the exact wordings from the manual) :

Output impedance (typical) :
Equivalent to a 0.05 mili Ohm resistor in series with a 3 uH inductor

So far I've never encountered this specific details on other lab psu's specification, this is my 1st. Understand the resistance part, but in series with an inductor ? 

Any one care to explain it please ? What is this "specific" information is used for ? (presumably for professional)

Warning though, this is EE wannabe grade question. 

Thanks in advance

[SeanB]

What they are saying is that this appears to the load like an ideal power source, capable of infinite current at a precise voltage. However between the perfect source is a single 50 microohm resistor and a perfect 3uH inductor. Thus the output voltage drops slightly with increasing current, and the voltage drops as the AC component of the vurrent drawn increases. This will have a small effect on the circuit up to a certain frequency ( probably noted somewhere or likely where the inductive reactance is approaching dropping a half volt or so) where it might make the control loop unstable.

In constant current mode however you ideally want output impedance to be as high as possible so the current is not affected by the voltage dropped across the load. There you might see the impedance specified in Megaohms.

[rf-loop]

If there is 1m distance to lod and there is 1mm diameter copper wires (total 2m lenght) to load  this wire have around 3.3uH. So...
Just for thinking.

[Zero999]

It shouldn't be a problem if the circuit being powered has adequate decoupling capacitors.

[BravoV]

Thanks for the replies folks. 

What they are saying is that this appears to the load like an ideal power source, capable of infinite current at a precise voltage. However between the perfect source is a single 50 microohm resistor and a perfect 3uH inductor. Thus the output voltage drops slightly with increasing current, and the voltage drops as the AC component of the vurrent drawn increases. This will have a small effect on the circuit up to a certain frequency ( probably noted somewhere or likely where the inductive reactance is approaching dropping a half volt or so) where it might make the control loop unstable.

So it is a parasitic inductance rather than a designed feature ? And HP simply stated it clearly at the PSU's specification ?

In constant current mode however you ideally want output impedance to be as high as possible so the current is not affected by the voltage dropped across the load. There you might see the impedance specified in Megaohms.

The manual doesn't mention about that impedance at current regulation's specification. 

If there is 1m distance to lod and there is 1mm diameter copper wires (total 2m lenght) to load  this wire have around 3.3uH. So...
Just for thinking.

Sorry, still don't get it. Its like you're answering a riddle with another riddle to me. 

It shouldn't be a problem if the circuit being powered has adequate decoupling capacitors.

I see this as an alternate answer that the inductance is an unwanted side effect (parasitic component) ?

Btw, in the manual, its stated that this PSU has a feature to disconnect the internal decoupling capacitor, through the remote terminals at the back.



Ok, maybe I add another question :

At which "specific" situation and/or condition that this provided detail on it's impedance does really matter ? Please, an example would be nice. 

.

[SeanB]

The specified condition you are looking for is a test of a device where you want to measure the dropout voltage of some circuit at varying load. The power supply in this case keeps the test conditions constant as you apply the step loads to the DUT, so the PSU has no interaction with the test. Speccing the info allows you to have confidence in your testing.

[alm]

Parasitic inductance is not a design feature, just like noise, drift and other things frequently stated in specifications. An ideal voltage source has zero resistance and zero inductance, the higher the output impedance, the further from ideal. But as rf-loop states, wiring also adds inductance, which may swamp the inductance of the power supply.

Inductance is irrelevant at DC, but it is important for transient response. You can simulate this in LT spice: an ideal DC voltage source, a current source as load drawing a pulsating current of say 1 A, and some series resistance/inductance between voltage source and load. Without parasitic components the voltage will be constant. You'll see spikes appear if you add parasitic capacitance and inductance, although 3 uH is quite low, so you need a fairly fast pulse to see this.

[free_electron]

Basically this tells you how the supply will react to load changes. If you alter he load current very fast the series inductance will come into play. So the current will take a bit of time to change and during that time the output voltage will be off.
It tells you the response of the supply to pulsed loads.

[blackdog]

Hi,

If you can handle a translator (google?) take a look @ this Dutch Electronics website (Circuits Online).
I explane al lot about dynamic impedance testing for a power supply i am designing.
It is a longggg topic :-)

http://www.circuitsonline.net/forum/view/110029/8


Kind regarts,
Blackdog

[BravoV]

..... <snip>.... Speccing the info allows you to have confidence in your testing.

This  , is what I'm sort of expecting to hear, thanks SeanB.

Thanks also to Hero999, rf-loop, alm, Vincent & Blackdog, all your comments add up to my understanding, again, sorry guys, even this sound too overly noob matter to you all, it is really a precious moment at least for me.

Alm, also I really appreciate your idea on simulation, it improves by big margin on my understanding. 



Yep, also tried different values on the inductance, heck, I even tried without the cap or a really small value cap, the voltage regulation result is so horrible, really an eyes opener. 

[BravoV]

Hi,

If you can handle a translator (google?) take a look @ this Dutch Electronics website (Circuits Online).
I explane al lot about dynamic impedance testing for a power supply i am designing.
It is a longggg topic :-)

http://www.circuitsonline.net/forum/view/110029/8


Kind regarts,
Blackdog

Thanks, honestly, especially since English is not my native, those translation is giving me a headache, really. 

Any plan to re-post it in English here in this forum ? That will be really great contribution to this forum and I believe many noobs like me will appreciate that.

[ConKbot]

..... <snip>.... Speccing the info allows you to have confidence in your testing.

This  , is what I'm sort of expecting to hear, thanks SeanB.

Thanks also to Hero999, rf-loop, alm, Vincent & Blackdog, all your comments add up to my understanding, again, sorry guys, even this sound too overly noob matter to you all, it is really a precious moment at least for me.

Alm, also I really appreciate your idea on simulation, it improves by big margin on my understanding. 



Yep, also tried different values on the inductance, heck, I even tried without the cap or a really small value cap, the voltage regulation result is so horrible, really an eyes opener. 

And today we learned why decoupling on your 100MHz amplifier is important