ferrite beads v. grid resistor

[Buckeye Sound]

I'm messing with a Carvin guitar amp and rather than a grid resistor such as a 68K at the input screen to V1 as on Fenders and others, they use a ferrite bead. I might swap it out just to see. Any thoughts on the pro's/cons of using a ferrite bead rather than a typical resistor here?

 

[cerrem]

Basically you have a 1 turn inductor when adding a bead...
Based on the permeability of the ferite..ect.ect... This inductance is tiny and makes a -40dB/Decade filter with the input capacitance of the vacuum tube... SO we have Double Pole low pass... The problem with the bead is the L is small and the filtering is effective for RF signals..from rectification in the valve...

Chris

 

[rdf]

I've done a number of Spice simulations on that scenario and was never able to make it work to my satisfaction. Any low-resistance inductance ahead of typical grid capacitances resulted in RF frequency peaking. The peaking could be tamed by adding a resistor in series with the inductor but acceptable results only came with values of R so large as to swap the contribution of L. By that point removing the inductor had little or no effect.

Those simulations didn't account for the inductance and capacitance of interconnects and should be revisited.

 

[Tom Bavis]

These ferrites are lossy, and should be modeled as an inductor in parallel with a 50 Ohm resistor. Especially good for supressing VHF oscilaltions, but I don't think they'd help with say, AM broadcast interference...

 

[Buckeye Sound]

Thanks. I'm going to try some standard value resistors in place of the bead and see what happens.

 

[rdf]

Hi Tom, the sims were to investigate wrapping a coil around a resistor in the hope of gaining the benefit of grid-stopping at RF while presenting an effective short between input and grid at audio frequencies. The model was a resistor in parallel with an inductor. The only way I found to tame the resultant RF peaking was adding resistance in series with the coil. It typically took so much series restance to flatten the curve that removing the RL element no longer made a difference. It's possible I didn't grind at it long enough though.

 

[cerrem]

Just a reminder to make sure to model the source impedance properly that is the preceeding stage.. SInce this will make a difference..

Chris

 

[coresta]

Hi , on old tubed Tektro scopes , especially 581 and 585 (85MHz bandwidth!) all the 6DJ8 distributed vert amps are simultaneously neutrodyned, ferrited and grid-resistored : The only way to get stable DC - 100MHz gain . If i remember, the symetric structure of the inputs and outputs needs even some twin lead ferrite cores too.

 

[richwalters]

The model was a resistor in parallel with an inductor. The only way I found to tame the resultant RF peaking was adding resistance in series with the coil.

This was the old amateur radio method of reducing volume control clicks to valve inputs. As a G8xxx er in my earlier days I always assumed that any inductance even though it suppressed also acted as an aerial. One can solve a problem as well as creating another. On tube amps I prefer using old carbon resistors or true unspiralled film resistors directly on tube inputs. Another way is to wrap a component in copper foil and ground it where it is with the foil.


richj