Toroid or EI ? Pros and cons plz...

Status
Not open for further replies.
I don't have a scope at the moment... loaned to a friend who can't seem to part with it..
lol..

A scope isn't the best choice for examining mains noise. What's needed is a spectrum analyser - trouble is they tend to be rather expensive in comparison with scopes. But they are more affordable than they used to be.

You can build yourself a common-mode choke which will do a good job of attenuating mains noise. You really want one which is bifilar wound on a segmented bobbin. Rather like these ones - http://www.whatsbestforum.com/attachment.php?attachmentid=8174&d=1360589767
 
Okay, read the article.
It left out some things.
For one thing everybody should know that successful amps put the mains EI transformer several inches away from any high gain wires or devices. Can't believe the author built an expensive box to stuff an unbanded EI transformer in 3" from the tubes, with the rectifiers over next to the tubes, too. He could have bought a stocker BUD 19" relay rack 5.5" chassis NHC-14156, as I am building my current amp in. The transformer is in one end with the bridge & the filter cap, then a steel mesh bulkhead, then the nearest high gain part is 3" from the steel bulkhead.
Examples of proper practice, the dynaco PAS2 PAS3 preamps that has a steel wrapped transformer in a separate steel box from the high gain wires. See Greg Dunn dynaco history website or dynacodoctor for pictures. Another example I own, Peavey PV-1.3k, has a copper banded EI transformer behind a steel bulkhead 4" from the high gain parts. The distance and the steel box kills the hum. A little sheet steel doesn't cost very much especially if it doesn't have to be finished to look good.
The disadvantages of toroid transformers, not listed in the article. 1 they have tremendous inrush current for their size, typically handled by putting a NTC resistor in series with the AC feed after the circuit breaker. 2. The ones for audio power amps, like the ones from antekinc.com in NJ, or this from Newark VPT100-5000 - TRIAD MAGNETICS - Toroidal Transformer, World Series, 500 VA, 2 x 50V, 10 A, Panel, World Series | Newark element14
don't come with mount pins or plastic cradles for the pc board. they are sold naked. So the mount is some sort of a wood cradle with an inverse donut you wallow out with a router or knife. I've seen some mounts between two steel or plastic plates. The first captures half the heat with an insulator. The second method captures the heat of both halves with insulators. Whereas the EI transformer couples the heat into the chassis with the mounting feet. The whole hunk of iron is out in the breeze if you have a fan, radiating heat out.
As far as line noise filter, in a dynaco ST120 EI powered amp, It originally had a .01 uf 1000 v disk cap out of the transformer into the diode bridge. It sounded fine as was. I've added a MOS surge supressor of about 200-300 vac rating ahead of the transformer primary, to keep lightning surges from breaking down the transformer insultation of about 600 v rating. Lots of lightning in this area. This amp sounds fine the 14 hours a day I use it, pops a bit when I turn on a flourescent light on the same circuit, also when lightning strikes nearby. It doesn't pick up lamp dimmer hash in the same room, switcher supply howl from the PC 30' away, or AM radio transmitters 4 miles away.
Have fun designing.
 
Last edited:
Main filtering it's an science ,not an joke , your not gonna fix it with an simple caps across the trafo or the main .

there is a guy here in France who knows very well about main filtering ( Francis Ibre ), if you understand French I can give you some usefull links .

anyway , common mode filters are useless , you need differential ones .

in my setup I use Shaffner FN 9675 , the FN 680 and FN 2080 are also good .

ask if you want more infos / pics 😉


.
 
Last edited:
Hi Fabrice, and thanks...

I can read French..
I won't try to SPEAK French...
And forget completely about writing in French! lol..

I would be greatul for more infos/pics... I'd like to better understand this subject!

Thanks again,
Greg
 
Main filtering it's an science ,not an joke , your not gonna fix it with an simple caps across the trafo or the main .

there is a guy here in France who knows very well about main filtering ( Francis Ibre ), if you understand French I can give you some usefull links .

anyway , common mode filters are useless , you need differential ones .

in my setup I use Shaffner FN 9675 , the FN 680 and FN 2080 are also good .

ask if you want more infos / pics 😉


.



Fabrice
Please Send me the links.
Thank you


Sent from my iPad using Tapatalk
 
A standard mains filter has both common mode and differential mode filtering inside the screening can.
The X rated caps typically 47nF to 100nF across the line are part of the differential filter.
The Y rated caps typically 1n5F to 4n7F are from line to Earth and are part of the commom mode filter.
Both of these capacitor types require an impedance in the line to allow the capacitance to become a filter. That line impedance can be R or L.
 
For those who want's , like me , to learn about main filtering , here are some links :

all in French , to read carefully , words of Mr Francis Ibre are those who are interesting .

I made my filters that way ( the second topic ) and it really make the difference .

remenber , real world testing and experience is the only way , not bla bla bla .....

have a good read 😉


ELEKTOR FORUMS • View topic - Filtres secteur Francis Ibre

ELEKTOR FORUMS • View topic - Filtre secteur 2. /an very good example .

ELEKTOR FORUMS • View topic - Filtre secteur 3

ELEKTOR FORUMS • View topic - [ACH] Filtre secteur Schaffner Francis Ibre

ELEKTOR FORUMS • View topic - filtre secteur + symétriseur



.
 
Last edited:
you'll find more infos reading the post , my understanding and my english are not enough to be more specific 🙂

the only thing i can tell you is that they do their job properly , especially in audio bandwidth , that's what we look after 😉



.
 
Status
Not open for further replies.