Question About Ferrite Beads

[Ricardus]

I am going to be building some balanced output boards using THAT 1646s. The boards are the circuit right out of the THAT spec sheet, and used to be sold by DIY Recording Equipment. Peterson ended up putting the Gerbers up on Oshpark after removing the boards from his product line. Here's his build guide with parts list:

https://www.diyrecordingequipment.com/pages/balanced-input-output-assembly-guide
I asked him what Ferrite Beads he was using (L1, L2) and he said he was using these (the pic is wrong, this one has leads):

https://www.mouser.com/ProductDetail/623-277300112
A vacuum tube preamp I built a few years ago spec'd these beads and I have a bunch left over. Again, the pic is wrong, the actual beads have two ferrites on them:

https://www.mouser.com/ProductDetail/81-BL02RN2R1M2B
So is there any meaningful difference between these two parts in this case, and can I use the ones I already have?

 

[Ricardus]

The ferrite beads I have with the two ferrites on them look like this.

 

[Bo Deadly]

The second ones are better for sure since it's the same impedance at about half the frequency and since the impedance is still going up at 1GHz which is probably good in a 5G wireless world.

However, on that PCB they're not doing spit because the primary purpose (only purpose?) is to block RF on entry. So these things really need to be right at the input. This is why I use tiny chip inductors (like in the pic that's wrong) and put them on a little circular PCB soldered directly to the XLR connector so that they're literally a few mm from the pins. And the 100p caps also need to be on the pins shunting to pin 1. And pin 1 must be connected to the chassis over the shortest possible length of wire (integrated spike style preferred). And the chassis has to have good low impedance all around and finally a connection to earth ground.

 

[warpie]

Is there any good replacement for the BL02RN2R1M2B Murata?

 

[Ricardus]

The second ones are better for sure since it's the same impedance at about half the frequency and since the impedance is still going up at 1GHz which is probably good in a 5G wireless world.

However, on that PCB they're not doing spit because the primary purpose (only purpose?) is to block RF on entry. So these things really need to be right at the input. This is why I use tiny chip inductors (like in the pic that's wrong) and put them on a little circular PCB soldered directly to the XLR connector so that they're literally a few mm from the pins. And the 100p caps also need to be on the pins shunting to pin 1. And pin 1 must be connected to the chassis over the shortest possible length of wire (integrated spike style preferred). And the chassis has to have good low impedance all around and finally a connection to earth ground.

Thanks!

 

[Ricardus]

Is there any good replacement for the BL02RN2R1M2B Murata?

I think if you look around at other vendors you might be able to find someone who has them in stock still.

 

[Ricardus]

Is there any good replacement for the BL02RN2R1M2B Murata?

This however, looks like a pretty good drop-in replacement.

https://www.mouser.com/ProductDetail/KEMET/B-03-R?qs=gt1LBUVyoHlnHk7fD5GpWw==

 

[CJ]

i think they want those beads on the output, this is to prevent long cables from transmitting rfi back into the output section of the chip. once inside the chip via an output transistor(s), it will have no problems wandering around the entire 1646 circuit.

hard to know what frequency any rfi will be at.

increased inductance can be had simply by running more turns of perhaps small magnet wire through the bead. we used 11 turns on some beads on a GFI power supply to prevent false tripping, but this was aimed at spikes rather than rfi. you have to be careful with capacitance, too many turns around the bead and your inductor becomes a capacitor at hi freqs, thus defeating the purpose of the bead in the first place.

one time an attempt to cut down on labor costs of winding beads for the gfi's (we sold thousands of these things) was made via wire wound ferrite beads with a resistor like coating, these proved to have gobs of capacitance so be careful with those.

beads can be seen at top left of gfi schemo. i believe this is the original engineer's copy as he used to use a hi-lighter pen to check the schematic against the actual pc board after it came in, a simple trick that works very well when you need to use brute force to solve a tough problem.

 

[Newmarket]

hard to know what frequency any rfi will be at.

Quite. And you'll never know it's impedance either

increased inductance can be had simply by running more turns of perhaps small magnet wire through the bead. we used 11 turns on some beads on a GFI power supply to prevent false tripping, but this was aimed at spikes rather than rfi. you have to be careful with capacitance, too many turns around the bead and your inductor becomes a capacitor at hi freqs, thus defeating the purpose of the bead in the first place.

Yes. From EMC work with cable ferrites the advantage drops away with more than one loop through the ferrite.
ie it has two turns with a nominal x4 increase in impedance.
Ferrites are rather complex in detail as they transition from being inductive into being primarily resistive with frequency.
Ferrites to Kill Ringing or Not? - In Compliance Magazine

 

[warpie]

I think if you look around at other vendors you might be able to find someone who has them in stock still.

I've spent a decent amount of time trying to source them but without any luck. At least in Europe.

 

[Matt Syson]

I seem to remember some in packets from RS components but that was years ago.