Say for construction reasons it was awkward or inelegant to solder a resistor directly up against a valve base, do ferrite beads generally work instead? I see it recommended a lot but I don't see it done often. The valves in question are high GM RF types.
I have some tiny toroidals with (presumably lossy) cores rated for EMI suppression. A couple of turns through them gives a measured inductance of around 10uH at 200Khz. Ideally above a certain frequency, a resistive lossy component would dominate, dampening any parasitic resonances.
Does anyone have any first hand experience with this working, or perhaps even making things worse? Any data points would be appreciated.
My confession is I absolutely hate having unsupported resistors poking out of valve sockets, which are then soldered to a wire. I would love to avoid this in a particular scenario where there is nowhere to secure a resistor. 😉
I have some tiny toroidals with (presumably lossy) cores rated for EMI suppression. A couple of turns through them gives a measured inductance of around 10uH at 200Khz. Ideally above a certain frequency, a resistive lossy component would dominate, dampening any parasitic resonances.
Does anyone have any first hand experience with this working, or perhaps even making things worse? Any data points would be appreciated.
My confession is I absolutely hate having unsupported resistors poking out of valve sockets, which are then soldered to a wire. I would love to avoid this in a particular scenario where there is nowhere to secure a resistor. 😉
If there was an advantage over carbon resistor, I am sure they would have been used from the onset of valves. If you use a 1/8W resistor and sleeve it, there is no weight and no need for support.
Inductors cause HF issues and parasitics, so it seems logical that ferrite beads will also.
Inductors cause HF issues and parasitics, so it seems logical that ferrite beads will also.
I fully admit to being awkward here 😀 I will probably end up just using resistors. The problem with the tiny ones is their leads bend too much and may end up snapping off - I'd rather use a 1/2w or even 1w as they are physically much studier, and of course heatshrink sleeve them.
For what it's worth the audio amp I am working on (not my design) used high slope RF valves all over the place with NO stoppers and long wire runs. 😱
I have used beads as stoppers in many situations.
Using them requires fairly careful design and selection, though.
The usual beads - in the forms of a hollow cylinder that is designed to slip over the lead of a component - can work well on some kinds of FETs, but the frequency range that they work over starts higher than chip-beads or multi-turn leaded beads, and will not be much help with most valves.
The slip-on beads require careful mounting. If they are allowed to move & vibrate around, (some types, at least) can generate noise voltages on the conductor! Very unpleasant. Fix them in place with something.
General considerations:
Think about what frequency range you want to suppress. If you run your circuit with no stoppers, can you provoke oscillation? If so, measure the frequency, and select a bead that offers resistive loss in that band.
Low frequency (<1MHz) oscillation is not a job for beads, and might well be down to poor construction (anode or grid wiring is too long, bad ground layout, supply cap too far away, etc).
Chip beads like Murata BLM21AG series are quite effective at HF, and do a good job of general oscillation suppression - but you need a PCB, or some cut FR-4 to mount them on.
But with grids where no current is expected, carbon composition resistors give trouble-free performance and are probably less hassle.
Using them requires fairly careful design and selection, though.
The usual beads - in the forms of a hollow cylinder that is designed to slip over the lead of a component - can work well on some kinds of FETs, but the frequency range that they work over starts higher than chip-beads or multi-turn leaded beads, and will not be much help with most valves.
The slip-on beads require careful mounting. If they are allowed to move & vibrate around, (some types, at least) can generate noise voltages on the conductor! Very unpleasant. Fix them in place with something.
General considerations:
Think about what frequency range you want to suppress. If you run your circuit with no stoppers, can you provoke oscillation? If so, measure the frequency, and select a bead that offers resistive loss in that band.
Low frequency (<1MHz) oscillation is not a job for beads, and might well be down to poor construction (anode or grid wiring is too long, bad ground layout, supply cap too far away, etc).
Chip beads like Murata BLM21AG series are quite effective at HF, and do a good job of general oscillation suppression - but you need a PCB, or some cut FR-4 to mount them on.
But with grids where no current is expected, carbon composition resistors give trouble-free performance and are probably less hassle.
Many thanks for the reply.
The amp is currently in bits for restoration and as I reassemble it I'm just struck by some questionable layout choices. As far as I remember the amp had no oscillation issues but probably only due to pure luck. If you saw the wiring you would understand.
The amplifier is loop stable but who knows what it could have done in the Mhz region local to the driver stages.
I am probably causing myself more trouble by going with beads and will just have to bite the bullet and use some sensible valued stoppers. Resistors poking out of B9A sockets drive me mad though 😀
In most cases I would use a 1/2W metal resistor attached to a single tag, so that it's mechanically secure at both ends.
Sent from my phone. Please excuse any typpos.
Sent from my phone. Please excuse any typpos.
Why?bigwill said:
Small ferrite beads are often used for suppressing VHF/UHF oscillation in HF equipment; they add some resistance but possibly too little for most valves. They may help with some UHF valves, but I suspect the reason resistors are almost always used for audio circuits is that resistors are the best choice.
I generally use 1/2W 300R MF resistors, I've soldered them to the socket at the length of the lead without issue. Never had an audible oscillation. Needed 47R plate stoppers on a PPP 6P3S amp though.
I've used ferrite beads but only for a phono stage where the added gain and length of interconnect make it a tuned antenna/radio circuit. Without them I could pick up FM radio with the phono stage.
I've used ferrite beads but only for a phono stage where the added gain and length of interconnect make it a tuned antenna/radio circuit. Without them I could pick up FM radio with the phono stage.
I fully agree with the sentiment that a normal resistor is the best way to go. The only reason I enquired about the beads is because in a particular situation it may have been neater to use one. If it was my design I would have gone with a resistive stopper and a rigid way to mount it from the start.
A specific situation where I would favor a ferrite bead over a carbon composition resistor is at the grid of a cascode's "upper" triode. That grid is supposed to be at AC ground. Cement a tiny bead that surrounds, but doesn't touch, the grid's socket lug, to the socket body.
I've posted, on several occasions, a 6922 cascode voltage gain block, with paralleled caps. (1 RF and 1 audio) forming the "upper" grid connection to ground. Add the ferrite bead, if the caps. fail to suppress parasitic oscillation.
BTW, when gm is large, a 100 ohm carbon film plate stopper is a good idea. Given the miniscule current present, carbon comp. is fine on the grid. OTOH, plate current is substantial and CC noise is not welcome. You put up with the small inductance a CF part introduces.
I've posted, on several occasions, a 6922 cascode voltage gain block, with paralleled caps. (1 RF and 1 audio) forming the "upper" grid connection to ground. Add the ferrite bead, if the caps. fail to suppress parasitic oscillation.
BTW, when gm is large, a 100 ohm carbon film plate stopper is a good idea. Given the miniscule current present, carbon comp. is fine on the grid. OTOH, plate current is substantial and CC noise is not welcome. You put up with the small inductance a CF part introduces.
Last edited:
I have found that both the screen and grid of the 6P41S benefit from both stop resistors and beads. I slide the bead on the resistor lead, insert it through the opening in the tube socket pin tab, bend the lead around and solder it. This secures the bead in place.
6P41S is a sweep tube and is prone to snivets.
6P41S is a sweep tube and is prone to snivets.
Funnily enough a 6DJ8 cascode is the input stage of this amp - the lower grid does in fact have a stopper (though not marked on the schematic but found in the real implementation), but the upper grid has no stopper with a pretty long lead.
FWIW these are the exact beads I have: 2643250402 | Fair-Rite Ferrite Ring EMI Suppression, For: EMI Suppression, 6.35 x 2.95 x 12.7mm | Fair-Rite
Made from '43 material'
43 Material Data Sheet - Fair Rite
Last edited:
It's somewhat humorous to see parts that are obviously Imperial sized offered with Metric measurements. 6.35 mM. = 0.25 inch and 12.7 mM. = 0.5 inch. UK magnetics maker Sowter offers some products with 19 mM. mounting centers. That is a very clever choice, as 19 mM. is, for all practical purposes, 0.75 inch. FWIW, I've flip/flopped 3/4 inch and 19 mM. ratchet sockets, without incident.
That bead strikes me as being too large for cementing to the socket. I'll check Mouser for something smaller, when I get home from the office.
That bead strikes me as being too large for cementing to the socket. I'll check Mouser for something smaller, when I get home from the office.
One advantage of ferrite beads over resistors is their smaller loss resistance and hence smaller thermal noise voltage density at audio frequencies. You won't notice this in a typical valve amplifier, but for an amplifier for moving coil cartridges it is certainly something to keep in mind.
Hi Eli (or anyone who can help), I was wondering why the bead should not touch the socket lug. I have some beads to fit for a tube tester, and I was planning on slipping them over the stripped end of the solid core wire, and soldering the wire end to the lug. The bead is held in place, but will be in electrical contact with the wire and lug. Is that an issue?
Cheers, Richard
Some beads are conductive; you can test them with a meter. A problem arises if there is a risk of short circuit.
Another precaution to be aware of:
If a loose bead is threaded over a conductor it may generate a small voltage when vibrated. Again, not all types show this behaviour, but it is worth checking for, before settling down with a pile of record s.
Another precaution to be aware of:
If a loose bead is threaded over a conductor it may generate a small voltage when vibrated. Again, not all types show this behaviour, but it is worth checking for, before settling down with a pile of record s.
Great, thanks. So just must ensure they cannot cause a short, and are fixed in place.
Is there a trick for daisy chaining tube sockets? I am thinking I solder a wire to each pin, add the bead, then take two connections from the other side of the bead. Or am I over thinking?
Is there a trick for daisy chaining tube sockets? I am thinking I solder a wire to each pin, add the bead, then take two connections from the other side of the bead. Or am I over thinking?