lumanauw said:
PS : Sorry, I'm not EE, I try to understand these things by stone-age iteration. I imagine for full swing, the VAS will have to be driven by dV=20mV. You EE have the simpler way, one word "gm"
Don't be sorry. Many of us aren't EEs either, just trying to patch up some of what we didn't learn about EE earlier in life. You will find that some of the people who have designed well-known and acclaimed circuits here on the forum actually have quite a different training and profession than EE, but they have obviously learnt a lot about EE in one way or the other. This forum is a great place to learn about all the strange circuits and phenomenae that the textbooks don't mention, and also to find interesting and challenging circuits to learn from by trying to analyze them and understand them. And, of course, to learn by discussing with each others and with the experienced professionals who are here and willing to teach us a bit of their knowledge.
BTW, gm is just dIc/dVbe at the Q point, or in simple words, how many mA of change in collector current you get per mV change in Vbe. It is useful only for small signal analysis, where you neglect the non-linearities of circuits.
Hi Christer,
I agree for only using expressions as Rush circuit or Rush pair.
My question about folded cascodes is still semantic but technical too. I asked it to myself during the Blowtorch thread.
On the joined schematics, if the positive power supply is considered as non-ideal :
- in the left circuit, the second transistor decouples the first transistor from the voltage variations that could otherwise be seen by its collector (it's the main function intended from standard cascodes), however the current output of the second transistor is sensible to the variations of the positive power supply. The resistor common to both transistors, connected to the positive power supply rail, should preferably be replaced by a constant current source.
- the second circuit does not decouple the first transistor from the voltage variations seen by its collector output (so I think this is the reason why it can't be named a cascode). However, from a point of view of positive power supply rejection, the current output of the second transistor should be about the same as if it was connected as a common emitter transistor.
I agree for only using expressions as Rush circuit or Rush pair.
My question about folded cascodes is still semantic but technical too. I asked it to myself during the Blowtorch thread.
On the joined schematics, if the positive power supply is considered as non-ideal :
- in the left circuit, the second transistor decouples the first transistor from the voltage variations that could otherwise be seen by its collector (it's the main function intended from standard cascodes), however the current output of the second transistor is sensible to the variations of the positive power supply. The resistor common to both transistors, connected to the positive power supply rail, should preferably be replaced by a constant current source.
- the second circuit does not decouple the first transistor from the voltage variations seen by its collector output (so I think this is the reason why it can't be named a cascode). However, from a point of view of positive power supply rejection, the current output of the second transistor should be about the same as if it was connected as a common emitter transistor.
Attachments
Forr,
I do see your point, but I think that basic circuit topologies should primarily be considered as ideal circuits. There will always be practical considerations when implementing them, but these are so many and so dependent on situation that we would have to end up with zillions of similar circuits to invent names for. Your idea about a CCS instead of a resistor sounds sensible though, if the rail isn't stable enough.
Also don't forget that ordinary cascodes are not always properly referenced to the emitter of the lower transistor and thus do not provide a perfectly stable Vce for it, and that problem is there even in the ideal theoretical circuit.
I do see your point, but I think that basic circuit topologies should primarily be considered as ideal circuits. There will always be practical considerations when implementing them, but these are so many and so dependent on situation that we would have to end up with zillions of similar circuits to invent names for. Your idea about a CCS instead of a resistor sounds sensible though, if the rail isn't stable enough.
Also don't forget that ordinary cascodes are not always properly referenced to the emitter of the lower transistor and thus do not provide a perfectly stable Vce for it, and that problem is there even in the ideal theoretical circuit.
I think we are all a bit confused here. We have both been discussing what a Rush cascode is and what to call it.
Since the Rush cascode obviously goes under a number of different names and it is debatable whether it has anything to do with cascodes, forr and I suggested that it might perhaps be better to just call it a Rush circuit, but of course, then we introduce yet another name. Besides, thinking of it now, I think "Rush pair" might be an even better and more descriptive name.
It is a bit difficult to give crisp and clear definitions of circuit topologies, since there are usually so many variations one can make on them, that it is sometimes hard to know what is and what isn't a certain type of circuit. Add to that that many circuits go under several names (diff pair/diff amp/LTP, for instance).
My take, which isn't worth more than anybody elses, is that the Rush cascode is the concept of tying two complementary BJTs together at their emitters and use the bases as a differential input. Variations existing by swapping one or both BJTs to JFETs or maybe MOSFETs.
Enter the problem, what if we just tie one of the bases to ground or a fixed voltage, as Kanwar did? Is it still a Rush cascode? I don't know. Sometimes one of the inputs of a diff pair is tied to ground (or rather very often when using op amps). Does it suddenly cease to be a diff pair? I think not.
Since the Rush cascode obviously goes under a number of different names and it is debatable whether it has anything to do with cascodes, forr and I suggested that it might perhaps be better to just call it a Rush circuit, but of course, then we introduce yet another name. Besides, thinking of it now, I think "Rush pair" might be an even better and more descriptive name.
It is a bit difficult to give crisp and clear definitions of circuit topologies, since there are usually so many variations one can make on them, that it is sometimes hard to know what is and what isn't a certain type of circuit. Add to that that many circuits go under several names (diff pair/diff amp/LTP, for instance).
My take, which isn't worth more than anybody elses, is that the Rush cascode is the concept of tying two complementary BJTs together at their emitters and use the bases as a differential input. Variations existing by swapping one or both BJTs to JFETs or maybe MOSFETs.
Enter the problem, what if we just tie one of the bases to ground or a fixed voltage, as Kanwar did? Is it still a Rush cascode? I don't know. Sometimes one of the inputs of a diff pair is tied to ground (or rather very often when using op amps). Does it suddenly cease to be a diff pair? I think not.
Hi Folks,
The CC-CB stage is not that hard to Bias emitter degeneration resistors do alot from preventing oscillations as well as thermal runaway, only thing under consideration is benefit arised from using this stage...Acc to my experience it gives more Fast Transient response, very high slewrate...I have compared it with normal CE-CB stage and Slewrate was 35V/uS for CE-CB and 125V/us For CC-CB stage also the open loop gain was some what high than normal configurations...
regards,
K a n w a r
The CC-CB stage is not that hard to Bias emitter degeneration resistors do alot from preventing oscillations as well as thermal runaway, only thing under consideration is benefit arised from using this stage...Acc to my experience it gives more Fast Transient response, very high slewrate...I have compared it with normal CE-CB stage and Slewrate was 35V/uS for CE-CB and 125V/us For CC-CB stage also the open loop gain was some what high than normal configurations...
regards,
K a n w a r
Workhouse
---.Acc to my experience it gives more Fast Transient response, very high slewrate...I have compared it with normal CE-CB stage and Slewrate was 35V/uS for CE-CB and 125V/us For CC-CB stage also the open loop gain was some what high than normal configurations...---
Compared to a parallel CE-CB circuit, the series CE-CB benefits from the
the lack of the parasitic capacitances of the constant current source.
However, what about the linearity ? Is speed more important for audio than linearity ?
---.Acc to my experience it gives more Fast Transient response, very high slewrate...I have compared it with normal CE-CB stage and Slewrate was 35V/uS for CE-CB and 125V/us For CC-CB stage also the open loop gain was some what high than normal configurations...---
Compared to a parallel CE-CB circuit, the series CE-CB benefits from the
the lack of the parasitic capacitances of the constant current source.
However, what about the linearity ? Is speed more important for audio than linearity ?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.