Lab grade power supply outputs

[ginetto61]

Hi and very sorry for the belated reply.
I thought i had subscribed instead i did not

The ground is a connection to the case / chassis of the power supply. Regardless of whether the input is AC or DC it is often required that the case of all devices including whatever you power with this supply be 'grounded' to the same potential.
This will almost always be the same potential as the mains ground and in most cases but not all the B- will be connected to the same potential.
To power a device use B+ and B- only.
If it is a device with a metal, grounded case, use the ground terminal to ground it as long as there is a connection from the power supply ground to mains ground otherwise it's irrelevant.

Thanks a lot indeed for the very kind and helpful reply.
Yes the device has a metal case but i do not know it is grounded.
However i will connect the device case to the power supply GND terminal and see if the noise change. I am focused on noise these days.
Next step will be distortion but this is more difficult.
I have to understand how to do that.
Thanks a lot again, gino

 

[ginetto61]

If you're experimenting, why not just use a battery?

Hi ! thanks for the interesting advice.
Honestly i do not know which type would be the best option.
I have read an article very impressive about the extremely low noise of NiCd batteries.
The power requirements for the psu are 12VDC and about 1.5/2 A.
So my guess is that big, heavy and expensive batteries are needed.
I am thrilled by the idea. The noise with those NiCd was almost zero
I will try to find the article.
But 4-5 hours of function will require a 10 Ah battery
I do not know.
If i can find a good traditional psu maybe it will be the most handy solution.
Thanks again, gino

 

[ginetto61]

OK. I thought we were talking about the output DC arrangements from two different PSUs. Now we are talking about the input DC arrangements of the load. It is easy to confuse me!

Hi and thanks a lot again for the very kind reply. My fault sorry

So we have two supplies, which may or may not be grounded or floating, single or dual polarity. We have a load, which may or may not ground one side of its supply.
You do realise that the cables may pick up more noise/interference than the supplies put out?

No ! .. of course i would like to avoid this absolutely !!!
However i should see this noise on the graph i guess ?
This software seems quite sensitive to noise also at low level.
Maybe it just outputs data not real ?

Hence the final stage of filtering, when necessary, has to be at the device end not the PSU end.
If a device needs a very clean supply then it should arrange this itself, not rely on something else to do it

Thanks a lot and this is exactly my opinion. In a very good interface the performance should be quite independent from the actual psu used.
It should have a great suppression of the noise fed by the psu.
And to be honest swapping cheap smps this is what i am seeing.
Not significant differences.
I would have stopped if i had not seen some spikes ... low in level and very narrow at specific Hz, with all the psu i tried.
They could be like a signature of the interface, persisting also after changing the psu.
Unfortunately i had to visit some relatives for a week
When i will be back i will update with results.
I am attaching pictures of the 1st lab grade psu i have already at hand.
Thanks a lot again.
Kind regards, gino

 

[Pafi]

...low in level and very narrow at specific Hz, with all the psu i tried.
They could be like a signature of the interface, persisting also after changing the psu.

If this is important for you enough to write this much, then you could also tell the value of the frequency...

There is a high probability that the noise comes from the SMPS of your PC.

"Noise" is electricity. There are many different form of it. Current or voltage, differential or common mode, RF or base band... depending on the type they behave very differently. Saying "noise" alone is meaningless. Some noises can be reduced by connecting GND, some others are increased. There is no single good practice for all.

 

[ginetto61]

If this is important for you enough to write this much, then you could also tell the value of the frequency...

There is a high probability that the noise comes from the SMPS of your PC.

"Noise" is electricity. There are many different form of it. Current or voltage, differential or common mode, RF or base band... depending on the type they behave very differently. Saying "noise" alone is meaningless. Some noises can be reduced by connecting GND, some others are increased. There is no single good practice for all.

Hi thanks a lot for the kind and valuable advice
The noise from the pc it is why I prefer interfaces with external psu
I would prefer that manufacturers supply a USB data only cable and a very good psu
The idea now is to compare psu with ripple 100 times different one from the other
If the noise spectrum will be unchanged it's already a result
It will mean that the noise is typical of the interface
Regards gino