But in a two stage amplifier the second stage is the output stage.
What about The Intermediate Stage, abbreviated TIS?
Cheers,
E.
Hi Edmond,
Let's bear in mind that not all VAS are transimpedance amplifiers, even at HF. In fact, one could argue that those that do not use Miller compensation are not TIAs. For example, an amplifier that uses lag compensation, with, say a series R-C shunt at the VAS collector. That VAS would seem to be properly described as a voltage amplifier. Pretty much the same thing can be said of the VAS in amplifiers that do not employ negative feedback and therefore do not have feedback compensation.
Cheers,
Bob
TIS vs. VAS, the line is drawn while comparing the input stage output impedance vs. the 2nd stage input impedance. For a current mirror loaded input stage, it would be pretty hard to get a second stage with a larger input impedance, even at LF, and even if an emitter follower is used. The follower input impedance is about Beta*Re, say 100*1k=100K, while the input stage output impedance is Beta/2*gm or say 100/2*0.1=500K. So I think it would be fair to call the 2nd gain stage a TIS if the input stage is current mirror loaded.
Now if no active loads are used in the input stage, then definitely the 2nd stage could be called a VAS. The output impedance of a resistive loaded input stage is in the kohm range, while the 2nd stage follower impedance is 100kohm or so.
Hi Waly,
This is a good observation. So the truly correct semantic answer is... it depends...
Cheers,
Bob
I doubt this.
The second stage of Thompson's topology is a transimpedance stage (TIS) not only by virtue of the shunt derived shunt applied negative minor feedback loop due to the Miller capacitor, but also because the input stage, the transadmittance stage (TAS), provides current drive to the TIS which delivers a voltage output at all frequencies including DC.
Even if we're to assume your argument that "the Miller capacitor looks more like an open circuit, the VAS acts more like a voltage amplifier" holds water, we would find that at all frequencies of interest, say greater than 5Hz, the Miller capacitor CANNOT be considered an open circuit and that the monicker TIS, and not VAS, is applicable.
depends on fiction
Hi Bob,
It's a fictitious circuit that should not be taken into consideration. No one, unless he is completely mad, will skimp on a current mirror, replace it by a resistor, which lowers the gain and disturbs the LTP balance and then finally add an emitter follower, which does almost nothing. So it does not depend....
Cheers,
E.
Hi Bob,
It's a fictitious circuit that should not be taken into consideration. No one, unless he is completely mad, will skimp on a current mirror, replace it by a resistor, which lowers the gain and disturbs the LTP balance and then finally add an emitter follower, which does almost nothing. So it does not depend....
Cheers,
E.
I would be very surprised if you got less than 10ppm THD 20KHz with your arrangements, Bonsai.
The input stage is a voltage to current converter while the second stage is a current to voltage converter.
For maximal current transfer from first to second stage you need a very high output impedance for the first stage relative to the input impedance of the second stage.
As Cherry has observed, using an emitter follower before the second stage without using a current mirror for the first is a complete waste of time.
The input stage is a voltage to current converter while the second stage is a current to voltage converter.
For maximal current transfer from first to second stage you need a very high output impedance for the first stage relative to the input impedance of the second stage.
As Cherry has observed, using an emitter follower before the second stage without using a current mirror for the first is a complete waste of time.
Yes.
P.S. your amps use very crude short circuit protection which really does little to protect your output devices while, simultaneously, needlessly limiting the output current. send me mail and I'll send you paper on proper SOA protection.
Last edited:
The beta enhanced VAS gives about 12 dB distortion improvement with resistively loaded LTP vs without. I cascoded the TIS (sim) as well, and got a further reduction in distortion, but cannot recall the exact figure now. I ended up skipping the TIS cascode in the interests of practicality - I build what I design so have to draw a line in the sand at some point. Good, because it forces one to think about tradeoffs.
OK, I also got this 10ppm figure with TMC, which gave 15 dB of distortion reduction vs conventional Miller comp.
So, a little bit here and a little bit there and you can end up with a pretty damn fine amplifier!
Read the article, Michael before passing comments like that.
The peak output current from the e-Amp is 40A.
I don't do multislope protection in the e-amp (I did that on the Ovation 250). I use very fast load switching using SSR's and an oversized output stage, and don't plan to hang a speaker on my amp with a 60 degree phase angle anytime soon. The protection kicks in before the 10A rail fuses have time to react, and the oversized output stage lets me comfortably drive nasty loads, while also contributing to acheiving the < 10ppm distortion figure.
There's more than one way to skin a cat, as they say in my neck of the woods.
Hi Andrew,
That's a different kettle of fish (after Borbely?) and the 'Intermediate Stage' isn't just a voltage amplifying stage. It's partly a folded cascode.
Cheers,
E.
I have read the article. Send me mail anyway.
P.S: I like your Vbe multiplier on the ovation 250.
Last edited:
You're trying to lead a herd.
Its time to move on and stop quibbling about semantics.
Cheers,
Bob