Need to make a power transformer choice... Have a good option in both Torroid and EI...
What recommends one or the other please? Pros and cons.
Its for a line level tube audio circuit, fed by an SS supply.....
TIA!
Greg
What recommends one or the other please? Pros and cons.
Its for a line level tube audio circuit, fed by an SS supply.....
TIA!
Greg
Last edited by a moderator:
The trade-offs are in the realm of primary-secondary coupling capacitance (EI is considerably lower than toroid) and in leakage flux (EI is considerably higher).
Coupling capacitance matters in regards to mains-borne noise and leakage flux matters in regards to noise pick-up by adjacent circuitry. If you have plenty of space in your case to put the trafo distant from the electronics, go for the EI. If space is at a premium, go for the toroid and include additional mains filtering.
Coupling capacitance matters in regards to mains-borne noise and leakage flux matters in regards to noise pick-up by adjacent circuitry. If you have plenty of space in your case to put the trafo distant from the electronics, go for the EI. If space is at a premium, go for the toroid and include additional mains filtering.
That gets it...
Exactly what I needed to know....
Bit of a dilemma for me then, though....
Damned if do, damned if don't a bit....
This application has a problem prone main circuit.... so the EI would be better..
BUT, it also will use a pair of TVCs... which of course are also inductors... IN the same case....
SOOO...
I'm fkd..
Lol..
I think the best bet is the Toroid.. With mains filtering..
Just wondering now, how much would be needed... and truly helpful..
Would those mains filters that sub for the IEC jack be sufficient???
Space isn't a big issue...
Space is the case... 😀
Exactly what I needed to know....
Bit of a dilemma for me then, though....
Damned if do, damned if don't a bit....
This application has a problem prone main circuit.... so the EI would be better..
BUT, it also will use a pair of TVCs... which of course are also inductors... IN the same case....
SOOO...
I'm fkd..
Lol..
I think the best bet is the Toroid.. With mains filtering..
Just wondering now, how much would be needed... and truly helpful..
Would those mains filters that sub for the IEC jack be sufficient???
Space isn't a big issue...
Space is the case... 😀
TVCs in the same case does mean you'll need to be a bit careful with placement and wiring. Are the TVCs themselves shielded?
I'd not use one of those IEC inlet bricks for a few reasons. Firstly you probably don't need as much current as the smallest current rating of those (often 1A or more). So rolling the filter yourself will gain you more noise rejection. Second, you really don't want Y capacitors and they come as standard in pre-fabricated filter blocks. What you do want is an effective common-mode choke with the highest inductance you can reach. If you're up for winding your own you'll want to go segmented in construction to reduce the parasitic shunt capacitance as much as possible.
I'd not use one of those IEC inlet bricks for a few reasons. Firstly you probably don't need as much current as the smallest current rating of those (often 1A or more). So rolling the filter yourself will gain you more noise rejection. Second, you really don't want Y capacitors and they come as standard in pre-fabricated filter blocks. What you do want is an effective common-mode choke with the highest inductance you can reach. If you're up for winding your own you'll want to go segmented in construction to reduce the parasitic shunt capacitance as much as possible.
So what are Y capacitors, and why are they not desirable??
As to roll your own:
Interesting idea.... but frankly, I don't exactly know what/how as of this moment...
But also, I'm trying to initially use pre-made boards, since this is a totally scratch build... and my first at that.
If I get bogged down in details, with so much to do just to have a working thing (which btw I need, since I've nothing else to use), life will intervene and it will wind up in a box on a shelf for years....
I plan to go back later, IF its an overall success, and do everything more perfectly.
Maybe even a new (nicer) enclosure... probably point to point... more ideal circuits... etc.
That would be the time to implement a roll yer own...
This is all assuming that the overall project is a success.. or at least doesn't suck... LOL.
That way, I can take my time to learn about (SO many) things like what/how a mains filter should be... while having something to listen to!
(and that learning is so fantastic!)
How about these?
18A EMI Filter Two Inductor Filter Power Purifier Kits for 100W Amplifier Preamp | eBay
Easier to mod (if needed) than to start from scratch...
I ordered a torroid btw.... lesser of the evils, in this case.
As to roll your own:
Interesting idea.... but frankly, I don't exactly know what/how as of this moment...
But also, I'm trying to initially use pre-made boards, since this is a totally scratch build... and my first at that.
If I get bogged down in details, with so much to do just to have a working thing (which btw I need, since I've nothing else to use), life will intervene and it will wind up in a box on a shelf for years....
I plan to go back later, IF its an overall success, and do everything more perfectly.
Maybe even a new (nicer) enclosure... probably point to point... more ideal circuits... etc.
That would be the time to implement a roll yer own...
This is all assuming that the overall project is a success.. or at least doesn't suck... LOL.
That way, I can take my time to learn about (SO many) things like what/how a mains filter should be... while having something to listen to!
(and that learning is so fantastic!)
How about these?
18A EMI Filter Two Inductor Filter Power Purifier Kits for 100W Amplifier Preamp | eBay
Easier to mod (if needed) than to start from scratch...
I ordered a torroid btw.... lesser of the evils, in this case.
Oh... the TVC's... I forgot..
Contacted the manufacturer (Silk Audio)...
Silk Audio 23 Step Supermalloy TVC (Pair)
He says they shield the TVC's with mu-metal, and the PWR trafo with a thick iron transformer shield, in one of their units, with success......
FWIW
Might that be adequate???
Contacted the manufacturer (Silk Audio)...
Silk Audio 23 Step Supermalloy TVC (Pair)
He says they shield the TVC's with mu-metal, and the PWR trafo with a thick iron transformer shield, in one of their units, with success......
FWIW
Might that be adequate???
Y caps are capacitors from the mains (both live and neutral) to earth. They're intended to perform some kind of low-pass filtering of noise on the mains, but their effectiveness does depend to a very large degree on the impedance of the earth (inductance especially). Given the that a low inductance earth connection is hard to arrange I feel they're best avoided and attention paid to increasing the common-mode impedance to reduce noise.
Those eBay 18A filters - well they're going to do something but they're using MnZn cores which means not much common-mode inductance. Great for high frequency mains noise attenuation but you'll probably want to combine them with something giving a lot more CM inductance to attack lower frequency noise (which I understand is the dominant kind of noise on mains, say 20k-100kHz).
Those eBay 18A filters - well they're going to do something but they're using MnZn cores which means not much common-mode inductance. Great for high frequency mains noise attenuation but you'll probably want to combine them with something giving a lot more CM inductance to attack lower frequency noise (which I understand is the dominant kind of noise on mains, say 20k-100kHz).
Last edited:
Uh, I salvaged some toroid filters out of PCAT switcher supplies for my disco mixer power. The ones after the X caps and before the mains rectifier & caps. I didn't do much engineering, but the result is dead quiet. I/ve got them filtering DC now, from a wall transformer, which gets the transformer flux out of the case entirely. Crackle from a florescent work light turning on does get through.
Televisions also have these toroid filters in the front end of the PS these days. Those are all over the curb now since the county imposed a $10 disposal tax.
Instead of an IEC filter, I'm using 15 mm MOS supressors between line & neutral on everything audio now: soldered on a cinch terminal strip. I use 300 vac rated since my mains are 125. Europeans may want a little higher trip voltage rating. I get MOS supressor at farnell/newark; littlefuse & bournes are two brands. I don't like the 7 mm ones. My Radio Shack "protected outlet strip" had one of those and the LED shows a failure after 18 months. Industrial motor drives has 22 mm MOS supressors, but I can't find those stocked at distributors. The more paste, the higher energy strike a supressor will handle. They are rated in joules.
I asked about putting .47 uf X capacitors from PCAT supplies on a new amp, but got told they were stupid and the proper value is .047 uf. Lacking a $15 box coming from newark this month, I used a salvage .01 uf 375 vac ceramic cap on my latest chip amp transformer input. We'll see how it handles shutoff pop.
As far as using IEC input filters, I don't like safety grounding every component. I don't even like putting AC driven transformers in every component. If I had babies or dogs chewing on the shields of the RCA cables I might rethink and put a +-15 v supply in everything and use op amps to cut the ground loops caused by safety wires on every component, but this obsolete old audio stuff I modify is not set up for that. Single ended interconnects and two wire power plugs, causes no ground loops. My safety ground to the power strip is the stud on the case of the mixer, where the green wire from the turntable ties to.
Televisions also have these toroid filters in the front end of the PS these days. Those are all over the curb now since the county imposed a $10 disposal tax.
Instead of an IEC filter, I'm using 15 mm MOS supressors between line & neutral on everything audio now: soldered on a cinch terminal strip. I use 300 vac rated since my mains are 125. Europeans may want a little higher trip voltage rating. I get MOS supressor at farnell/newark; littlefuse & bournes are two brands. I don't like the 7 mm ones. My Radio Shack "protected outlet strip" had one of those and the LED shows a failure after 18 months. Industrial motor drives has 22 mm MOS supressors, but I can't find those stocked at distributors. The more paste, the higher energy strike a supressor will handle. They are rated in joules.
I asked about putting .47 uf X capacitors from PCAT supplies on a new amp, but got told they were stupid and the proper value is .047 uf. Lacking a $15 box coming from newark this month, I used a salvage .01 uf 375 vac ceramic cap on my latest chip amp transformer input. We'll see how it handles shutoff pop.
As far as using IEC input filters, I don't like safety grounding every component. I don't even like putting AC driven transformers in every component. If I had babies or dogs chewing on the shields of the RCA cables I might rethink and put a +-15 v supply in everything and use op amps to cut the ground loops caused by safety wires on every component, but this obsolete old audio stuff I modify is not set up for that. Single ended interconnects and two wire power plugs, causes no ground loops. My safety ground to the power strip is the stud on the case of the mixer, where the green wire from the turntable ties to.
Last edited:
Y capacitors must be connected: Line to the enclosure.
The manufacturers do this using an ultra low impedance to the enclosing metal can and then electrically connect that can to the enclosure. This removes the need for us to be concerned about introduing inductance into the filter route. The can around the filter MUST be connected directly to the enclosure for which it is providing the filter.
If the can is an IEC input socket with integrated filter, then the can projects through the enclosure wall and electrically connects to the enclosure wall all around the hole through which it projects.
The length of the gaps between electrical contacts (of can to enclosure) determines the frequencies that can leak past the gap.
Last edited:
Sounds great in theory but if you recall the current through those Y caps doesn't want to return to the enclosure, it wants to return to earth. And the earth comes to the enclosure via the mains lead - plenty of inductance there in a couple of m of wire. Then there's the inductance of the in-wall earth wiring to consider. So that 'ultra low impedance' connection between cap and can is in series with something definitely rather a long way from even 'low impedance' at the frequencies of mains noise (above 20kHz).
Are you talking about MOV? Metal Oxide Varistor?
Why are you using 300V rated for a 125V mains?
You are allowing 2.5X overvoltage before it does anything.
For 230V mains in europe usually 275V MOV is used.
You will probably find many in any shops if you search for MOV instead of MOS.
You can use multiple ones if you are concerned for the current.
And you can't use MOV instead if IEC socket, because they do totally different thing.
MOV only works if a spike comes in that is higher than the rated voltage. IEC socket filters everything. MOV does nothing with "noise".
Last edited:
H.Ott tells me differently.
The High Frequency interference passes through the low impedance of the Y capacitor/s to the screening can. That is electrically coupled to the equipment enclosure.
The enclosure is capacitively coupled the "Earth". The PE route to "Earth" has too much inductance/impedance to pass any significant amount of the HF interference.
The enclosure provides the main screen.
It does not need a connection via PE back to the distribution board for the screening effect of the enclosure to be effective.
If you want to look at it from another direction:
Why do manufacturers of screening canned filters offer versions with Y capacitors and versions without Y capacitors?
Is it because they want to make their Y versions more expensive than their competitors equally good versions that manage without?
The specification data should show the improved filtering when the Y caps are implemented properly by connecting the filter can to the enclosure to make use of the capacitive coupling to "Earth".
"Earth" in the above is the world we stand on.
Last edited:
Always keen to learn from Henry Ott. How much capacitance does he say the enclosure typically has to earth and what's his estimate of the inductance of the mains connection? Also what frequencies was he considering?
In my understanding its because some applications cannot tolerate leakage currents to earth. Medical applications for example tend to have reduced Y cap values (or perhaps omit them completely) due to stringent regulations on the maximum permitted leakage.
Yes, medical filters have no Y caps, to remove the risk of Line getting onto probes attached to one's patient.
But excluding that special case, the manufacturers make Y versions because they have the data to prove that the Y caps improve the RF attenuation. This can only work effectively if the impedance of the shunting element to "Earth" is indeed a low impedance. Long wires won't help.
H.Ott is a long read. I'm not inclined to read him again, but I can recall he was referring to High Frequencies that see the Y cap as a low impedance. And it's that low impedance that needs to be maintained by the capacitive coupling from the equipment enclosure to "Earth". He actually drew diagrams of example placements to show that location can affect this air coupling.
But excluding that special case, the manufacturers make Y versions because they have the data to prove that the Y caps improve the RF attenuation. This can only work effectively if the impedance of the shunting element to "Earth" is indeed a low impedance. Long wires won't help.
H.Ott is a long read. I'm not inclined to read him again, but I can recall he was referring to High Frequencies that see the Y cap as a low impedance. And it's that low impedance that needs to be maintained by the capacitive coupling from the equipment enclosure to "Earth". He actually drew diagrams of example placements to show that location can affect this air coupling.
Last edited:
Under what conditions are they measuring that data? What inductance is assumed to earth and what capacitance to earth for the enclosure?
It must therefore follow that you have equally no idea whether their published data is relevant to the real world conditions.
That's all very nice...
And very fascinating, I'm sure....
And mine is the biggest of all!!! 😎 🙄
Now, how does one make a proper mains filter????????????
Meanwhile, back to the OP....
Here's a GREAT article on which transformer types are what, and the pros and cons....
Power Transformers for Audio Equipment
Totally answers the OP question....
And very fascinating, I'm sure....
And mine is the biggest of all!!! 😎 🙄
Now, how does one make a proper mains filter????????????
Meanwhile, back to the OP....
Here's a GREAT article on which transformer types are what, and the pros and cons....
Power Transformers for Audio Equipment
Totally answers the OP question....
Hi. Have you an oscilloscope? I would only bother with a mains filter if you can measure rf noise on the supply rail, and then use something very much like this http://uk.rs-online.com/web/p/power-line-filters/7672712/
Last edited:
- Status
- Not open for further replies.