Ferrite beads on output transformers

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

warpie

Well-known member
Joined
Feb 7, 2009
Messages
1,566
I see that chips like THAT 12xx or 1646 use ferrite beads on pins 2 and 3 (XLR) to avoid any interference.

Is there any benefit of using ferrite beads in a circuit that has an output transformer instead of a THAT ic (i.e. on the out TX's secondaries)?
 
I agree but the reason I'm asking is because I'm thinking in modifying some of my gear with the option to select between TX/IC output and practically speaking it'll be a lot easier to switch the signal when it's still unbalanced (primaries). So, is it OK to have the ferrite beads on pins 2and3 regardless on whether there's an IC or a TX before? Hope I make some sence... :)
 
I agree but the reason I'm asking is because I'm thinking in modifying some of my gear with the option to select between TX/IC output and practically speaking it'll be a lot easier to switch the signal when it's still unbalanced (primaries). So, is it OK to have the ferrite beads on pins 2and3 regardless on whether there's an IC or a TX before? Hope I make some sence... :)

Are you implying that you would have the output of the IC and the transformer tied together all the time, and you would just switch the input between the transformer primary and the IC input?
 
So, is it OK to have the ferrite beads on pins 2and3 regardless on whether there's an IC or a TX before?
Yes and you would want to do that as long as the ferrites are right on the XLR at the entry point. The purpose of the ferrites is to block RF from getting inside the enclosure. The length of the the traces (or wires) from the connector at the entry point of the enclosure to whatever they're attached to (and regardless of what they're attached to) determines the frequency of RF that those wires can radiate within the enclosure. The longer the traces, the lower the frequency of RF that can get in. I have little circular PCBs that mount directly to the back of XLR in/out with chip inductors and caps right next to the pins. But I use chip inductors instead of ferrite beads because they're much smaller, less clumsy, cheaper, higher inductance, plentiful and they also stop COVID-19 from getting into the enclosure as well.
 
Are you implying that you would have the output of the IC and the transformer tied together all the time, and you would just switch the input between the transformer primary and the IC input?
Well, at least that's the plan. Something wrong with this?
 
Last edited:
Yes and you would want to do that as long as the ferrites are right on the XLR at the entry point. The purpose of the ferrites is to block RF from getting inside the enclosure. The length of the the traces (or wires) from the connector at the entry point of the enclosure to whatever they're attached to (and regardless of what they're attached to) determines the frequency of RF that those wires can radiate within the enclosure. The longer the traces, the lower the frequency of RF that can get in. I have little circular PCBs that mount directly to the back of XLR in/out with chip inductors and caps right next to the pins. But I use chip inductors instead of ferrite beads because they're much smaller, less clumsy, cheaper, higher inductance, plentiful and they also stop COVID-19 from getting into the enclosure as well.
The problem is that I already use XLRs with solder lugs so once I fit a PCB one the pins I won't be able to remove them from the enclosure in case I have to.
 
Just remember that not all ferrite is the 'same', and how much impedance is being applied depends on the ferrite grade, the length of the bead and if there are any turns. Aim for a ferrite that shows the lowest frequency impedance peak, and where the impedance has a substantial real impedance component.
 
Just remember that not all ferrite is the 'same', and how much impedance is being applied depends on the ferrite grade, the length of the bead and if there are any turns. Aim for a ferrite that shows the lowest frequency impedance peak, and where the impedance has a substantial real impedance component.

Thanks for the tip. I already got these ones. Do you think they'll be OK? They look good to me.

https://www.digikey.com/en/products...06673?s=N4IgTCBcDa4BwBkAMBGAzHAtCpAlbSIAugL5A
It's supposed to be "broadband material"...
https://www.digikey.com/en/htmldatasheets/production/578955/0/0/1/28r0880-000.html
 
Last edited:
I agree but the reason I'm asking is because I'm thinking in modifying some of my gear with the option to select between TX/IC output and practically speaking it'll be a lot easier to switch the signal when it's still unbalanced (primaries). So, is it OK to have the ferrite beads on pins 2and3 regardless on whether there's an IC or a TX before? Hope I make some sence... :)
I am a bit late to this thread and also a little puzzled about the idea of switching unbalanced (primaries). If you do this and do not switch the secondaries then this means you need an XLR for the TX and another one for the IC. In which case you might as well drive both at once. Or have I missed something?

Cheers

Ian
 
Well, this is a very rough drawing of what I have in my mind but come to think of it, probably switching the secondaries is the only way to do it...
 

Attachments

  • Untitled_1.jpeg
    Untitled_1.jpeg
    83.6 KB · Views: 28
Well, this is a very rough drawing of what I have in my mind but come to think of it, probably switching the secondaries is the only way to do it...
I think you are right because the way you have drawn it the IC is driving the transformer secondary. All you need is a SPST switch to select one of two balanced signals.

If you prefer to use a SPDT switch you could always use it to switch a DPDT relay.

Cheers

Ian
 
I think you are right because the way you have drawn it the IC is driving the transformer secondary. All you need is a SPST switch to select one of two balanced signals.

If you prefer to use a SPDT switch you could always use it to switch a DPDT relay.

Cheers

Ian

You lost me there Ian. How can I control both the hot and the cold signal with a SPST?

The LF ferrite would give you more impedance for frequency range where I think you would need it.

So I should have got the LF version of this bead right?
 
I agree but the reason I'm asking is because I'm thinking in modifying some of my gear with the option to select between TX/IC output and practically speaking it'll be a lot easier to switch the signal when it's still unbalanced (primaries). So, is it OK to have the ferrite beads on pins 2and3 regardless on whether there's an IC or a TX before? Hope I make some sence... :)
When I designed the A-Designs Mix Factory, we did the same thing, it was well worth doing, as one was clean and one was "dirty". I did play with the IC outputs with 2) 10 ohm resistors, so the IC out had the same loaded level loss as the their transformer, so when you switched, it was the same level. For this unit, 10R was right. Just play with the values and have a few different loads, 1K, 5K, 20K etc, to find the value.

As far as the ferrite beads, like John said, don't waste your time, it's only for CE certification when you are manufacturing, to insure that if you have any clocks inside the unit, they don't escape. A ferrite bead or small choke won't effect any tone at all, unless it's huge, and you don't need that.

When switching the outputs, use a double pole switch or relay, and I always hook 100K-200K between them so they are always at the same DC potential in case their is an offset or the next device has any offset, it just prevents a POP or Click. The IC out would have the output caps against the op-amp, and a value of 220uf to 470uf is fine depending on what the next stage load it, then the series resistors. Always put 100K to ground on the outside of the cap to ground (before the series Rs).
 
As far as the ferrite beads, like John said, don't waste your time, it's only for CE certification when you are manufacturing, to insure that if you have any clocks inside the unit, they don't escape. A ferrite bead or small choke won't effect any tone at all, unless it's huge, and you don't need that.
You seem to forget that the transformer has a significant leakage capacitance between its windings and its core, which may or may not be grounded.
I believe it's the main reason why Jensen recommends the use of an inductive "Output Line Isolator" (actually a glorified ferrite bead)
https://www.jensen-transformers.com/wp-content/uploads/2014/08/as023.pdfIt would be interesting to know Bill's take on the subject, and why Jensen took their OLI off their catalogue.
In the absence of it, the circuit that drives the xfmr sees a significant capacitance to ground (a few nF); the issue is not the capacitive load, since a well designed xfmr driver should be capable of handling such a load (particularly in the presence of an inductive load that somewhat compensates it, but mainly the problem that HF signal is dumped into the ground to which the xfmr frame is connected. It is the source of unwanted HF voltages that may interfere with other parts of the circuitry.
Another option is to let the frame float, which also presents another problem of electrostatically radiated field that may interfere with sensitive nodes.
 
Well, this is a very rough drawing of what I have in my mind but come to think of it, probably switching the secondaries is the only way to do it...

Suppose you switch to the transformer, (and presumably you remember to put a pull-down resistor on the input of the IC so it is not floating), now the transformer is trying to drive the audio signal on pins 2 and 3, and the IC is trying to hold pins 2 and 3 at 0V. Which one wins?
 
Back
Top