Hi-Q
Junior Member level 2
Hi,
I want to design a test system with a multitude of modules that can deliver signals (DC and audio frequency AC) to, and measure signals ((DC, audio frequency AC and RF) from a unit under test that also consists of several modules.
As the modules of the unit under test share a common ground, especially due to the shields of the RF connections between each other, but each of them requires power supply voltages and control signals, I would like to have the several modules of the test system isolated from each other so that their grounds only meet at the unit under test (star grounding scheme with the UUT being the star center).
Nonetheless, the modules of the test system shall be located in a common housing (I am thinking of a 19 inch carrier with either the slide-in modules electrically isolated from the carrier or with the ground nodes of the several modules not connected to the casing and front-panel of the modules).
As each module needs +-15V and an additional 15V supply, having 8 modules I would need to generate 24 supply voltages within the 19 inch rack.
What would be the best way to do this?
My ideas:
1. 24 completely separate SMPS-modules like, for instance, RS-15-15 or RS-25-15 connected to a common AC rail, located in 8 slide-in modules plugged in from the rear end of the 19 inch carrier.
2. The 8 slide-in modules plugged in from the rear end of the 19 inch carrier, but in each of them one bigger 15V SMPS, for instance an RS-50-15, and two isolating DC/DC converters for generating the other two supply voltages for the according module.
3. One big SMPS for generating a convenient DC supply voltage and then using three DC/DC converters in each of the 8 modules.
4. One big SMPS and a 24-output DC/DC converter with 24 isolated outputs.
My concerns are mainly regarding the supply noise / EMI when having so many SMPS or DC/DC converters.
Are individual ones better because they will have slightly differing switching frequencies and therefore spread the EMI over a broader frequency band and generating lower peak levels?
Or would having many slightly different switching frequencies bear the risk of generating all kinds of intermediate frequencies mixing down into the audio frequency band where they would mess up the signals that are to be measured?
Thanks,
HI-Q
I want to design a test system with a multitude of modules that can deliver signals (DC and audio frequency AC) to, and measure signals ((DC, audio frequency AC and RF) from a unit under test that also consists of several modules.
As the modules of the unit under test share a common ground, especially due to the shields of the RF connections between each other, but each of them requires power supply voltages and control signals, I would like to have the several modules of the test system isolated from each other so that their grounds only meet at the unit under test (star grounding scheme with the UUT being the star center).
Nonetheless, the modules of the test system shall be located in a common housing (I am thinking of a 19 inch carrier with either the slide-in modules electrically isolated from the carrier or with the ground nodes of the several modules not connected to the casing and front-panel of the modules).
As each module needs +-15V and an additional 15V supply, having 8 modules I would need to generate 24 supply voltages within the 19 inch rack.
What would be the best way to do this?
My ideas:
1. 24 completely separate SMPS-modules like, for instance, RS-15-15 or RS-25-15 connected to a common AC rail, located in 8 slide-in modules plugged in from the rear end of the 19 inch carrier.
2. The 8 slide-in modules plugged in from the rear end of the 19 inch carrier, but in each of them one bigger 15V SMPS, for instance an RS-50-15, and two isolating DC/DC converters for generating the other two supply voltages for the according module.
3. One big SMPS for generating a convenient DC supply voltage and then using three DC/DC converters in each of the 8 modules.
4. One big SMPS and a 24-output DC/DC converter with 24 isolated outputs.
My concerns are mainly regarding the supply noise / EMI when having so many SMPS or DC/DC converters.
Are individual ones better because they will have slightly differing switching frequencies and therefore spread the EMI over a broader frequency band and generating lower peak levels?
Or would having many slightly different switching frequencies bear the risk of generating all kinds of intermediate frequencies mixing down into the audio frequency band where they would mess up the signals that are to be measured?
Thanks,
HI-Q
