Re: gEDA-user: basic anti-EMI design q

[Karel Kulhavy <clock@xxxxxxxxxxxxx>]

On Thu, Mar 23, 2006 at 10:05:53PM -0500, Dan McMahill wrote:
> DJ Delorie wrote:
> 
> >Keep in mind, this board is mounted inside a five-sided metal box,
> >which is mounted on the metal air duct coming out of our furnace.
> >This is the same furnace that includes a couple of induction motors,
> >an electrostatic air cleaner, fire, water, refrigerant, and 18 gauge
> >wires running all over the house.
> 
> last time I had to work in a hostile environment, I threw away the solid 
> state dumb waiter controller and replaced it with latches and a simple 
> state machine made with mechanical relays!  No more interaction between 
> the dishwasher and dumb waiter :)
> 
> >Needless to say, the existing board suffers from interference.  On the
> >current board, the latches occasionally change state on their own, so
> >I continuously set them to limit the fault time, but I can't reset
> >them often enough to keep the furnace from clicking on-off
> >occasionally.
> 
> Do you have a way to tell how the interference is getting in?  Is it the 
> supplies or the signal lines?

Through the space. The interference must run continuously in the space,
it cannot jump or teleport or go through 4th dimension. So interrupting
the space with a conductive plane always does the job.

> 
> >So, the big question is - ground/power planes.  I could probably
> >squeeze the design onto two layers, but power and ground traces would
> >be going all over the place.  Going to four layers gives me power and
> >ground planes, with easier signal routing.  An auxiliary question is,
> >should I split the planes?  I'm thinking, isolating the power/gnd
> >going to the I/O drive FETS and thermo power (i.e. the stuff going
> >around the house) back to a common point near the power supply, and
> >maybe splitting the 10baseT analog power similarly.
> 
> If you do split the planes, the big thing to keep in mind is all 
> currents flow in a loop.  So pay attention to the complete current loop 
> and make sure it flows where you want.  If you don't pay attention to 
> this, you can split the planes in a bad way and do worse.
> 
> 
> >Based on a goal of "minimize the effect of external EMI interference",
> >what makes sense?
> 
> Some generic EMI/EMC advice I read many years back which is simple and 
> useful is that it takes 3 things for you to have an EMI problem.  1)  a 
> source, 2) a path, 3) a receiver.  Look for ways to hammer on all 3 and 

Hmm if the neighbour is getting interference from Ronja then yes,
hammering his TV will do the job.

> you'll find a weak link (hopefully).  Sometimes you can only address 1 
> of them, but thats all it takes.  Sounds obvious, but I've found it 
> useful to approach these sort of problems with those 3 things clearly in 
> mind.
> 
> As an example, I helped someone with a board where the 13th harmonic (no 
> kidding!) of a clock oscillator was getting into the IF path of a 
> receiver via the power supply and then mixing with an LO (or maybe it 
> was a harmonic of the LO) and producing an in-band interference.  The 
> biggest benefit came from putting some series R in series with the clock 
> oscillator supply pin and a close local bypass cap.  Kept that HF 
> garbage very near to the clock.  The next benefit came from a small 
> series resistor at the output to limit the risetimes of the signal at 
> the load.  This helped on the source and the path.  The other path fix 
> was improved supply decoupling at the IF end.
> 
> In your case, I'd look at using opto isolators or transformers where 
> possible for driving those off board signals.  That confines the 
> currents induced in that loop to a specific spot which doesn't include 
> your power supplies.  If the supply into your board is particularly 
> nasty, which is likely the case, at a minimum you should put some 
> filtering on it right where it comes in.  Be sure and address both 
> differential and common mode signals.  X and Y caps in line filtering 
> terms.  In other words, put some C between the lines and also to ground. 

C between the lines? So garbage from one line can spread into all other
lines?

>  For an AC signal, you can put a common mode choke on the incoming 
> line.    By this I mean, take the pair of wires that bring 24 VAC in and 
> wrap them around a magnetic core.  Differential currents produce little 
> magnetic flux in the core and hence there is little inductance for the 
> differential circuit.  From a common mode point of view, you have a 
> large inductance.  In your case where you might be getting some really 
> large common mode glitches from the air cleaner and motors, this may 
> help.  Actually, on this topic, differential signalling for off chip 
> signals and some common mode filtering may give some benefit.
> 
> Also, many regulators don't have stellar rejection of glitches at their 
> input so you might look specifically for that when picking one.  If you 
> can tolerate it, you might be able to improve things by cascading 2 
> regulators.  Depends on spectral content of what you're trying to 

Put a PI RLC filter (R because you want to kill off the resonance),
simulate it in gnucap and tune it to get say 40dB wideband rejection,
and you might get stellar glitch rejection :) If you don't and the
garbage skips the filter by inducing from the incoming wire onto the
outgoing wire, solder the thing up watertight into a tin plated steel
case :)

CL<
> reject.  I've done that in a case where I really wanted a very quite 
> supply.  I used an off the shelf regulator to produce something larger 
> than I needed and used a custom design to further improve it.
> 
> Take a look at some of the EMI filters made by companies like murata. 
> Those can be quite effective on supply lines.
> 
> -Dan