The most recent topic on a similar theme seems to be from 2009 (either that or my searching skill suck!)
As I'm getting ready to build a few projects which at some point is likely to include making a power supply I was wondering if I could get some feedback and/or opinions on the merits of IEC inlet filters with earth line chokes?
I've located a data sheet containing some information on the filters made by Schaffner and was wondering if anyone has used these in the past and could explain some of the benefits and potential drawbacks of integrating them into a DIY design?
On a similar theme, the Tacima CS929 seems to perform a similar function (and has a 5* review on what hifi) would I be better off using something like this or are there again certain advantages and drawbacks?
Any information is much appreciated!
As I'm getting ready to build a few projects which at some point is likely to include making a power supply I was wondering if I could get some feedback and/or opinions on the merits of IEC inlet filters with earth line chokes?
I've located a data sheet containing some information on the filters made by Schaffner and was wondering if anyone has used these in the past and could explain some of the benefits and potential drawbacks of integrating them into a DIY design?
On a similar theme, the Tacima CS929 seems to perform a similar function (and has a 5* review on what hifi) would I be better off using something like this or are there again certain advantages and drawbacks?
Any information is much appreciated!
The earth line choke will help impede high frequency noise circulating on that wire. I have two issues with the chokes integrated within IEC inlet modules. One concern is that they are too low in inductance to impede noise in th audible range. While high frequency common-mode noise is a real problem in many applications, I don't see audio as being one of those applications, unless a switching power supply or an class-D amplification circuit is being employed. The other concern I have is safety related. Should the earth choke open, then the chassis looks to no longer be safety grounded to earth, making for a shock or fire hazard, although, I see the unit has several certifications listed.
I will admit, safety is a pretty high concern when building your own PSU. Mains voltage and current is enough to kill and one stray copper wire or short against the chassis and you're fried!
I believe earth impedence needs to be below 0.1 ohms to be safe? I'd presume even with the choke it should still be below this... Can anybody confirm?
I'm likely to be using class D amplification ICs in my projects along with DAC ICs so high frequency noise is a potential issue.
I believe earth impedence needs to be below 0.1 ohms to be safe? I'd presume even with the choke it should still be below this... Can anybody confirm?
I'm likely to be using class D amplification ICs in my projects along with DAC ICs so high frequency noise is a potential issue.
Looking at the plots for the parts with the choke vs those without, the differences seemed to be at most 6dB improvement but only in some parts of the spectrum.
The choke values as quoted in the DS look to be 100 or 400uH, a very small value in the scheme of things. Can't see the point myself - if more iron is to be added I reckon better to put a higher performance CM choke on the L/N rather than a low value choke on the E.
The choke values as quoted in the DS look to be 100 or 400uH, a very small value in the scheme of things. Can't see the point myself - if more iron is to be added I reckon better to put a higher performance CM choke on the L/N rather than a low value choke on the E.
What sort of inductance values would you suggest to place on the L/N lines? I'm rather new at this (as one may have figured) and looking to learn all I can before putting theory into practice. I'm loathed to mess with the earth line myself since that's in effect the last failsafe against my own stupidity! If a manufactured product can filter that sufficiently I'd be a lot happier going with it if it made a difference.
I've already determined that after the rectifier bridge I need a set of DC smoothing capacitors - 1pf (ceramic disc), 1nf (foil), 1uf (Polypropylene)and 2000uf (Polypropylene) or greater in parallel between the +V and Gnd lines should smooth most of the ripple. I've been told to avoid electrolytic as these are likely to heat and vent with constant switching between charge and discharge. Combining these with inductance loops would block out a lot of the extra high frequency noise left on the line?
As far as CM chokes go, inductance is one parameter to be concerned about (I aim for at least 50mH in lowish current applications - i.e. for signal sources, not poweramps) but equally important is the resonant frequency, above which the CM choke no longer acts as a choke. Parts with higher resonance freqs are constructed in segmented form to reduce the self-capacitance.
Your understanding of DC smoothing capacitors is at best incomplete - a 1pF capacitor will do nothing useful in a real world DC supply circuit, nor will the 1nF foil. A 2mF polyprop is going to be positively HUGE - partly because they never come in low voltages and partly because its not possible to make the dielectric anywhere near as thin as with aluminium electrolytics. Avoidance of electrolytics will be very costly indeed in terms of space but also financially - you look to have been misinformed about their drawbacks.
PS Electrolytics do get (slightly) warm but they get warmest when charged/discharged at the highest rate - which is typically in a switched-mode power supply. These run at least 1000X higher frequencies than mains. So given that switching supplies (which pretty universally use electrolytics) don't suffer the problems your 'knowledgeable' contact has mentioned, its vanishingly unlikely that caps running at mains frequencies will.
Your understanding of DC smoothing capacitors is at best incomplete - a 1pF capacitor will do nothing useful in a real world DC supply circuit, nor will the 1nF foil. A 2mF polyprop is going to be positively HUGE - partly because they never come in low voltages and partly because its not possible to make the dielectric anywhere near as thin as with aluminium electrolytics. Avoidance of electrolytics will be very costly indeed in terms of space but also financially - you look to have been misinformed about their drawbacks.
PS Electrolytics do get (slightly) warm but they get warmest when charged/discharged at the highest rate - which is typically in a switched-mode power supply. These run at least 1000X higher frequencies than mains. So given that switching supplies (which pretty universally use electrolytics) don't suffer the problems your 'knowledgeable' contact has mentioned, its vanishingly unlikely that caps running at mains frequencies will.
Last edited:
If the fault current pulse is only ½kA then the voltage drop for the 0.1ohms impedance would be 50Vpk
i.e. the protected chassis would sit at 50Vpk during the peak of the fault incident.
This is around the limit for maximum chassis voltage.
What about the other impedances in the return circuit that the fault current must take.
That does not sound safe to me.
I'm not 100% certain what you're talking about there... Remember it's current that kills - not voltage. Static electricity can be tens of thousands of volts but such a low current it's just a sting (1cm electric arc at normal room temperature and pressure is ~20,000V if I remember A-level physics). Ideally an earth should have a zero ohm resistance as the earth protects against chassis shorts - nothing else - and does this by blowing the fuse because it draws a significant current. This reduces the duration of any fault incident and the potential for serious bodily harm.
TLDR; Lower earth impedence = faster fuse blowing = less death. High earth impedence = fuse doesn't blow = death. (Sorry, felt like being funny)
The UK tests the current capability of the equipment flow and return resistance.
This when added to the supply resistance in both the flow and return, determines the total resistance of the fault current route.
To pass the test the total resistance has to be below some set value.
If you increase the resistance and/or the impedance inside the equipment you take with you the risk that you increase the time for the mains fuse to rupture. An inductive impedance has an even bigger effect for a fast changing current.
I think there is also a test of the installation to effect the same fast blowing requirement
This when added to the supply resistance in both the flow and return, determines the total resistance of the fault current route.
To pass the test the total resistance has to be below some set value.
If you increase the resistance and/or the impedance inside the equipment you take with you the risk that you increase the time for the mains fuse to rupture. An inductive impedance has an even bigger effect for a fast changing current.
I think there is also a test of the installation to effect the same fast blowing requirement
Last edited:
Ah good, we are thinking along the same lines then. From what I understand the set value to pass the test you're referring to is 0.1 ohms. You confused me by talking about voltages lol.
Basically even with an earth choke, we want the eath impedence to be as close to 0 ohms as possible. I was hoping someone may be able to confirm what sort of values these have or what values DIY inductor loops on the earth line tend to have.
I'm also interested in how placing inductor loops on DC Live and Neutral/Ground lines after the rectifier bridge may affect voltages and stability of devices. If I'm going to design a PSU that puts out +15 and -15V and then the inductor loops cause another voltage drop, my preamp chips won't be getting what they need to function as the design intends. Sure I could use voltage regulators but these are then introducing noise back into the circuits!
My safety concern is regarding what happens should the earth line choke suffer an open circuit failure mode. Such an failure would un-ground the chassis, creating a shock hazard. Because the chokes are small enough to be embedded in an IEC module, I might wonder about how robustly they are constructed. It's may be likely that the earth choke is simply comprised of ferrite beads clamped around the earthing wire internal to the module, which then would remove my failure mode concern.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.