1. Home
  2. Electronics Design
  3. EMI-resistant input transformer design
  4. Page 2

EMI-resistant input transformer design

Ouch! I test the teaching lockin we sell by looking at how much of the oscillator gets into the front end (at ~3 kHz.. highest frequency). The answer is typically ~10 nV. Which also about the noise level of the preamp...(10nV/rtHz) kinda the edge of what I can measure. (I assume you've got a low source impedance.)

As others said (and you most likely know) you want a foil and braid shield. With good grounding!... I remember explaining to my boss why the split steel lock washer could not go between the shield wire and connector body, but had to be outside that connection. (Well I had measurements to back it up, having done it wrong the first time myself.)

George H.

Reply to
George Herold
Loading thread data ...

Spehro does a lot of cryogenic stuff, SQUID magnetometers and so forth. The transformer ratio doesn't have to be that large in order to get a big noise benefit.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
ElectroOptical Innovations LLC 
Optics, Electro-optics, Photonics, Analog Electronics 

160 North State Road #203 
Briarcliff Manor NY 10510 

hobbs at electrooptical dot net 
http://electrooptical.net
Reply to
Phil Hobbs

I still do not quite understand your requirements.

The noise power generated by a resistor at the input is -174 dBm/Hz corresponding to a 300 K noise temperature.

Good UHF preamplifiers are capable of 1-2 dB below that noise floor, thus allowing for 1-2 dB input filter loss, we are still talking about

-174 dBm/Hz noise powers. At UHF a 1/4 input resonator is quite small and even at VHF and upper HF the dimensions for a helical resonator is quite reasonable.

Is the EMC issue really that bad ?

Reply to
upsidedown

So we are talking about 4 K devices...

In that case, maintaining a cryogenic temperature would be much harder than mainlining a proper EMC shielding.

Perhaps running the amplifier at 77 K might be an option in a cryogenic system.

Reply to
upsidedown

Depends. There could be VFDs around, which will put nasty ground loop junk even on conduit with no problems at all.

Last time S. and I talked about what he was doing (some years ago), he was flying low-TC SQUID sensors on aircraft, to detect buried kimberlites (diamond pipes). Just the vibration damping arrangements were super interesting.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
ElectroOptical Innovations LLC 
Optics, Electro-optics, Photonics, Analog Electronics 

160 North State Road #203 
Briarcliff Manor NY 10510 

hobbs at electrooptical dot net 
http://electrooptical.net
Reply to
Phil Hobbs

Hopefully they have a separate PE (Protective Earth) and TE (Technical Earth) networks in the plant. The VFD EMC filter and other mains filter grounds goes to PE and small signal cable shields goes to TE.

It is essential that the PE and TE networks are kept separate and there is exactly one connection point (a jumper) between PE and TE networks. The jumper can be removed, when the plant is powered down and it can be verified by measurement that PE and TE are really separate.

Using separate PE and TE networks helps keeping ground loops and common mode voltages away.

Reply to
upsidedown

A request to all designers doing low signal level system design, please keep the various grounds separate and bring them to the PCB edge connector and to box level connectors.

If separate grounds are not needed, it is easy to link together different grounds (PE and TE) outside the box. If plant level separate grounding is used, this can be easily implemented.

It is quite frustrating when a device has mains EMC filter grounds, signal grounds as well as chassis grounds all solidly connected together.

The RF currents from the EMC filters will quite effectively pollute the PE ground and if the grounding network has been badly designed, it will also pollute the TE network.

So please keep different grounds separate on the PCB or box level.

Reply to
upsidedown

Oh boy! This sounds like ground wars! What's the highest (signal) frequency you've worked with? At low frequency star ground (as you describe) is the way to go. At high frequency, ground early, often and where ever you can seems to work best. Depending on the system in between is 100 kHz to 1 MHz (Well that's my best guess.)

How is your 'scope grounded?

George H.

Reply to
George Herold

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.