DC Power Cables - Effects on SQ

[marce]

8 hours ago, mansr said:

Typically, the raw DC input feeds a buck (or linear if you're so inclined) regulator outputting 5-12 V. From this, low-noise LDO regulators provide the voltages (usually 3.3-5 V) required by various components. The noise seen by the critical circuits is mainly determined by the local regulator, not the external power supply, provided the latter is of reasonable quality.

Further to this, the main DC power that is distributed is often 12V or greater (15V is popular and sometime as high as 44V, generally below 50 then you don't have to worry about SELV requirements.)

If you use a cable for 12V it will pick up a certain amount of rf noise depending on the cable construction, the same cable carrying 3V will pick up the same noise, but with a far worse SNR. So if you are providing external DC, a main high voltage with local regulation as Mansr described is the best way in EMC terms. Providing multiple low voltage feeds via external cables increases the risk of cross-talk, rf pick up and other EMC issues. Then there is the cost and problem of filtering all these supplies when they enter the box being supplied. Filtering at the point of entry is the rule here.

I will dig into my notes and stuff on this, plenty off stuff from Henry Ott, Ralph Morrison, Keith Armstrong etc as well as many examples I can look at, there is a whole variety of filtering DC inputs and different components even active modules.

Sorry if you think its a bit OT but looking at the whole area of DC distribution will benefit the discussion. It could be the lack of proper filtering, bad wire layout etc. that does cause issues with some systems and noise pick up.

[marce]

Some interesting tomes (textbooks almost☹️) to promote discussion...

https://www.analog.com/media/en/training-seminars/tutorials/MT-095.pdf

 

 

If anyones got the kit.

http://www.hottconsultants.com/techtips/tips-cm.html

 

Based on a larger scale, but the ideas do filter down, we have motors being driven by amps and control gear for the amp (DAC), also shows good wiring techniques.

https://dl.mitsubishielectric.com/dl/fa/document/manual/plc/bqn-p0e100083e/bqnp0e100083e.pdf

 

Or if your really serious... a great Textbook:

https://www.amazon.co.uk/Electromagnetic-Compatibility-Engineering-Henry-Ott/dp/0470189304

or

https://www.amazon.com/Grounding-Shielding-Interference-Ralph-Morrison/dp/0470097728

 

And here a whole wealth fo practical advice and information.

http://www.emcia.org/keitharmstrong.aspx

 

 

 

[marce]

10 hours ago, John Dyson said:

- I am speaking of things like noise from CPUS or other fast digital circuits that have current surges in the order of Amps and transition speeds of nsecs or less.) 

 

https://blogs.mentor.com/hyperblog/blog/2015/06/02/simultaneously-switching-noise-an-overview/

Simultaneous switching noise, the digital power delivery system is designed to cater for this and the initial high dI/dt requirements, the on chip capacitance and local decoupling capacitors cater for the initial charge, then the on board reservoir caps. scattered about and finally the power supply as the power supply cannot handle the current requirements on its own.

We have to remember this noise is not stored in any digital files...