What constitutes an RF choke? I mean, the impedance of an inductor at HF may constitute a choke at UHF. Is there a guideline?

 

Think about collector load resistors used in CE transistor amplifier circuits. They might typically be in the 1K to 20K range. For the frequency of interest and inductor in that range would be a choke in my estimation.

 

Hello,

A lot of information on building RF circuits can be found in the attached PDF.
There is also a block on RF chokes.

Greetings,
Bertus

 

It's as much the specific application as anything.

Consider a coupling capacitor vs. a filter/smoothing capacitor - identical 0.1uF caps could be used for both applications, depending where they are in the circuit.

A choke tends to be an inductor used for blocking or isolating high frequency signals (or noise), often in power connections.

It could also be an inductive element in a power supply smoothing setup, like a PI-network C-L-C filter.

Think of choke point or choke hold; a choke is an inductor used to control / limit / regulate rather than as a resonant element.

Chokes for high-frequency blocking or noise filtering may also be deliberately low-Q, using lossy core material or wound over lowish value resistors, to prevent any unwanted resonances affecting the operation of the rest of the circuit.

 

Good points all. The load resistor analogy is still apt in the sense that a low value resistor has a hard time developing a voltage across it without using large currents. The same is true for an inductor. A low impedance at the frequency of interest will pass an RF signal barely attenuated. A higher impedance inductor will, as you say, choke off the flow of RF current and allow a voltage to be developed across it.

 

Hi, Papabravo. My question is really 'what constitutes an effective choke?' I was hoping for a quantification such as an impedance of ten thousand ohms at such and such a frequency, or an impedance that drops the signal by such and such a factor, or an impedance that drops the signal so many volts or amps. I realize the inductor value is a function of frequency.

 

Hello yourself, Bertus,
Thanks for the link. I downloaded it and will print it out. It addresses precisely the type of problems I'm running into right now. I've got another month to play with electronics then I have to go back to work. Lately I've been striking out alot. I hope you all don't mind my questions; they arise from hand's-on fiddling and failure.

 

It's as much the specific application as anything.

Hi rjenkins. I'm actually aware of that. As I said above, I'm not really asking what an rf choke does, but rather a quantity of impedance such that an inductor can be considered a choke at some specific frequency.

Consider a coupling capacitor vs. a filter/smoothing capacitor - identical 0.1uF caps could be used for both applications, depending where they are in the circuit.

A choke tends to be an inductor used for blocking or isolating high frequency signals (or noise), often in power connections.

It could also be an inductive element in a power supply smoothing setup, like a PI-network C-L-C filter.

Think of choke point or choke hold; a choke is an inductor used to control / limit / regulate rather than as a resonant element.

Thanks, but as I said above, I'm aware of all of this. I'm looking for some designer's rule of thumb for determining an effective impedance that will satisfy the requirements of a given choke.

Chokes for high-frequency blocking or noise filtering may also be deliberately low-Q, using lossy core material or wound over lowish value resistors, to prevent any unwanted resonances affecting the operation of the rest of the circuit.

Now that's interesting. I haven't run into the need for such a choke as yet, but I'm sure to in the future. Thanks.

 

I'd think very basic rules would be:

The DC resistance is low enough for the specific application and current involved,
and
the self-resonant frequency is well away from any frequencies that occur in the circuit and allowing for any capacitive loading involved. I'd make sure it's resonance was lower than any circuit frequencies to avoid any possible harmonic resonances.

 

Agreed, but what is a figure of merit?

Say you're decoupling a 455 kHz signal from the positive power supply and you make an attenuater at the power feed (collector) of a CE amplifier by putting an rf choke (inductor) between the collector resistor and the supply. If you made it attenuate with a roll off frequency at 4.55 kHz, then the signal would be reduced by 40 dB at 455 kHz, right?

But what if we have a string of amplifiers -- as in an IF strip -- having discreet stages -- and an overall gain is 60 dB? We are decoupling the last amp in this string. Then how much attenuation do we need from this inductor? It seems to me that 40 dB would not be enough unless we could be confident that the attenuation at the front end of the whole strip could decrease the signal enough to make the feedback signal of no consequence.

 

In that example, the top of the collector resistor would also have a decoupling cap to ground as well as the inductor. Just the cap will give considerable attenuation.

The signal would already be far lower on the cap than the actual output, so the choke filter is furter attenuation on that.

Typically, in a multi-stage amp, there would either be a star-point feed to the various stages, or the power input would be at the output end of the amp chain and inductive or resistive isolation components at some points along the power chain to the input end.

Effectively, that gives the smallest signal stage a multi-point Pi network filter.

 

In that example, the top of the collector resistor would also have a decoupling cap to ground as well as the inductor. Just the cap will give considerable attenuation.

I realise there are multiple sources of attenuation between the output signal and the input of a multistage amplifier. The signal at the collector would attenuate across the collector resistor, then the choke, and then any decoupling capacitors between the power line and ground. Then it would be attenuated as it took a path to the input amp.

And yet I've seen many an amplifier turn into an oscillator, so evidently, we can't be careful enough in our designs or the physical implementation of them. I suppose the bottom line here is this: Use whatever amount of inductance for a given choke that makes sense with respect to the circuit at hand.