Correct in case of series resonant tank but wrong with the parallel resonant tank
Yes it has but has its bias been understood so that it is used to advantage? The only place where I see its advantages exploited well is the circa 1960's circuit showing the front end of an Eddystone EC10 Communications Receiver (see above). The LTspice simulations given early on in this thread show that bias is critical and there is a "tipping point" the good properties of Common Base reach a maximum and then turn bad, the Eddystone EC10 circuit knows this because an operator can adjust the bias because he is turning an AGC (or RF gain) pot.
As regards your line level preamp example I agree that it is common base for the first and second transistor but is it? My understanding of Common Base is that the Base is connected to signal ground or shorted to signal ground for AC by a good size capacitor.
This so called Cascoode for tubes or transistors really is a terrible bad jargon thing and just guarantees that the two devices thus connected are not properly biased.
I just don't like Cascode much prefer the use of Cascade as it is a real word and explains.
For valves/tubes the Cascade is a compromise, if you go with setting the upper Triode Grid at 1/3rd of HT it means that the bias point for both Triodes is different and not ideal. If you set the voltage at the upper Triode Grid at a value that makes the bias point the same for both valves it pushes the HT up and then power dissipation is getting higher so what you gain by keeping Grid capacitance down you loose by higher HT voltage. They used the Cascade in Valve TV tuners to get VHF UHF performance at low cost.
I ran a simulation for a Cascade and copy the LTspice circuit below:
I'm just not getting excited about Cascade and any idea that (for solid state) a Cascade configuration is anywhere near correctly biased Common base.
As for Miller effect well hardly much justification for having to call it Miller effect because its glaringly obvious that the Capacitance between Grid and Anode will be static Grid Capacitance times the increased Anode voltages that increases as function of gain.
For valves/tubes the Cascade is a compromise, if you go with setting the upper Triode Grid at 1/3rd of HT it means that the bias point for both Triodes is different and not ideal. If you set the voltage at the upper Triode Grid at a value that makes the bias point the same for both valves it pushes the HT up and then power dissipation is getting higher so what you gain by keeping Grid capacitance down you loose by higher HT voltage. They used the Cascade in Valve TV tuners to get VHF UHF performance at low cost.
I ran a simulation for a Cascade and copy the LTspice circuit below:
I'm just not getting excited about Cascade and any idea that (for solid state) a Cascade configuration is anywhere near correctly biased Common base.
As for Miller effect well hardly much justification for having to call it Miller effect because its glaringly obvious that the Capacitance between Grid and Anode will be static Grid Capacitance times the increased Anode voltages that increases as function of gain.
Last edited:
It’s not bad jargon, and I don’t see how a name affects the use of a circuit. Wikipedia gives the origin of the name https://en.wikipedia.org/wiki/Cascode where it was coined for good reason.
In applications like the Eddystone receiver, the base voltage is varied to change the emitter resistance, and hence the signal current into the emitter. That is a cheap and nasty thing to do, but 60 years ago it was the best solution possible in RF design. That isn’t a typical use of a common base design as linearity changes significantly with bias (and in the case of the circuit originally posted, Q).
I just don't like the word Cascode, I know the explanation of how the term came to be but I don't buy it, terrible, and I don't want to carry the plethora of labels that electronic circuitry has collected over the years for often trivial reasons when in real circuits every case is usually different.
Yes changing the base voltage changes the bias of the Grounded Base and interestingly Grounded Base bias is critical, gain goes up reaches a peak and then drops off very fast. I noted it because I have seen how this works (without fully understanding) and have simulated it to find resistors for peak gain. I will repeat the circuit and my comments about it below. Try to see my point of view as to me the Grounded Base has very unusual surprising qualities that I have puzzled over for hours.
Please have a look at this circuit:
The idea is to get a sharp resonance (high Q) in the 10.427mH 8n2 tuned circuit. Simulating this while stepping the value of R1 shows that as R1 increases towards about 3K9 the resonance gets sharper and sharper. As soon as R1 exceeds 3K9 resonance is soon lost and performance quickly degrades. I can't see why but that's what happens. I see the Eddystone EC10 circuit using this "quality" to good advantage - giving the operator a pot to turn (RF Gain) that can be adjusted to a peak. By the way the EC10 was a very early design and used Germanium RF transistors such as Mullard AF116 so I think the designer was a very cunning skilled engineer. That said the Eddystone EC10 was a budget design what they today call an entry level communications receiver.
Cascode - Yuk.
Yes changing the base voltage changes the bias of the Grounded Base and interestingly Grounded Base bias is critical, gain goes up reaches a peak and then drops off very fast. I noted it because I have seen how this works (without fully understanding) and have simulated it to find resistors for peak gain. I will repeat the circuit and my comments about it below. Try to see my point of view as to me the Grounded Base has very unusual surprising qualities that I have puzzled over for hours.
Please have a look at this circuit:
The idea is to get a sharp resonance (high Q) in the 10.427mH 8n2 tuned circuit. Simulating this while stepping the value of R1 shows that as R1 increases towards about 3K9 the resonance gets sharper and sharper. As soon as R1 exceeds 3K9 resonance is soon lost and performance quickly degrades. I can't see why but that's what happens. I see the Eddystone EC10 circuit using this "quality" to good advantage - giving the operator a pot to turn (RF Gain) that can be adjusted to a peak. By the way the EC10 was a very early design and used Germanium RF transistors such as Mullard AF116 so I think the designer was a very cunning skilled engineer. That said the Eddystone EC10 was a budget design what they today call an entry level communications receiver.
Cascode - Yuk.
I understand your dislike for the word, but in a technical environment using the right word in the right context is important if you want to be understood, but as we are discussing a common-base circuit it is probably a distraction.
For the circuit you present, if I interpret the naming correctly then you have biased the transistor for about 1mA collector current. That means that, with a 3k9 collector load Vcb is 0. Without looking at the data sheet for the transistor I would expect the gain to drop and the transistor to start to enter saturation. The extreme of this with increasing load resistance is that the transistor degenerates to a diode to ground.
For the circuit you present, if I interpret the naming correctly then you have biased the transistor for about 1mA collector current. That means that, with a 3k9 collector load Vcb is 0. Without looking at the data sheet for the transistor I would expect the gain to drop and the transistor to start to enter saturation. The extreme of this with increasing load resistance is that the transistor degenerates to a diode to ground.
Having looked at the datashhet for the SSM2212, figure 13 might explain what you are seeing. The collector capacitance changes massively for small Vcb and forms a low-pass filter with R1. To check this, replace the antenna with an ac current source, and sweep the frequency for a range of R1 values.
Yes I will try that, I need to look for the simulations, from memory I think its to do with the bias voltages being such that its not possible for the Vbe voltage to reach the 0.65 Volts approx to start turning the transistor on - so it never gets turned on when R1 is much higher than 3K9.
The SSM2212 is a very well behaved transistor pair I use it because it has a very good LTspice model whereas a lot of common transistors don't have models or have third party models.
The SSM2212 is a very well behaved transistor pair I use it because it has a very good LTspice model whereas a lot of common transistors don't have models or have third party models.
The Cascade.
How I explain the fundamental electronic circuits to laymen:
A transistor has three legs, two of which determine the result.
Think of it like a water tap: there is a source, a regulator and a drain.
So you can measure the amount of water in these three ways, and you choose one leg as a reference.
But combinations are also possible.
Two water taps that share a common source: the differential.
Two connected water taps, one imitates the other: the current mirror.
A water tap that controls another water tap: the Darlington.
A water tap that controls another water tap in reverse: the Sziklai.
These are the two complementary sets that can also be made with transistors.
And then we have the lonely one, the beautiful cascade, the waterfall.
The lower water tap regulates with precision, the upper one regulates the enormous pressure.
We miss the counter to this cascade, which everyone seems to have forgotten.
That is the solid-state triode, which can be realized with only three resistors...
The resonance is intended to be around 17.2 kHz so quite low but even a small varying capacitance is likely to shift the resonant frequency.
Ah Estonia - very good country, good people !