G.Kleinschmidt said:
Okie-dokie. Let us know how it measures when you actually build a real one.
Oh. And let us know how it sounds, too.
Posts that are not related to "BJT" or "Mosfet", attacking other members will go to Texas.jameshillj said:
Or lots of something else connected in parallel. The stuff that doesn't grow on trees is making progress slow, but when finished I'd be happy to run club demos (not that I actually know of any audio clubs in my vicinity/city ATM) with whatever speakers one cares to provide.
Parallel connection of multiple units is fine, so long as the minimum impedance is kept above about 0.5 ohms.
😎
> Does Class-A have any sonic advantage over a well-excuted Class AB amp that has equally good measured performance?
Yes, definitely in my experience. And the difference between a Class A amplifier of 100kHz and 350kHz of equivalent topology, output power and device types are immediately audible, according to my very limited experience. In that particular case, the only difference was the value of a silver mica cap across the feedback transistor.
> I have no explanation why this degree of bandwidth would be audible. I have no explanation of why a certain bandwidth with a bipolar design would sound duller than a lesser bandwidth in an all FET design. But as usual, the amp doesn't care what I think or whether I understand it. It just sounds how it sounds.
Very well put.
Patrick
Yes, definitely in my experience. And the difference between a Class A amplifier of 100kHz and 350kHz of equivalent topology, output power and device types are immediately audible, according to my very limited experience. In that particular case, the only difference was the value of a silver mica cap across the feedback transistor.
> I have no explanation why this degree of bandwidth would be audible. I have no explanation of why a certain bandwidth with a bipolar design would sound duller than a lesser bandwidth in an all FET design. But as usual, the amp doesn't care what I think or whether I understand it. It just sounds how it sounds.
Very well put.
Patrick
Charles Hansen said:
Charles Hansen said:
Hi Charles,
I understand what you are saying here, but I think you are confusing open-loop bandwidth and closed loop bandwidth in the context of amplifiers that employ negative feedback. In your designs, where there is no NFB, there is really no distinction; it is just the bandwidth of the amplifier. In an NFB design, the closed loop bandwidth is the bandwidth of the amplifier. I think you are just saying that the bandwidth of the amplifier wants to be well in excess of 100 kHz, and I completely agree.
An NFB amplifier can easily have a bandwidth of 500 kHz to 2 MHz (the closed loop bandwidth) while having an open loop bandwidth of less than 1 kHz.
Cheers,
Bob
Re: Re: Re: Re: VAS buffer
Hi Bob,
So this diamond buffer does work, not only in Sim City, but also in real life. The next step is to get rid of (the effect of) the capacitance of the Baker clamp diodes, which paves the way for a virtual distortionless front end.
BTW, I stumbled upon two different definitions of a Baker clamp on the web:
1. A diode between VAS input and VAS output.
2. A diode between VAS output and some fixed voltage.
Which one is correct? (I thought no. 1)
Cheers,
Edmond.
Bob Cordell said:
Hi Bob,
So this diamond buffer does work, not only in Sim City, but also in real life. The next step is to get rid of (the effect of) the capacitance of the Baker clamp diodes, which paves the way for a virtual distortionless front end.
BTW, I stumbled upon two different definitions of a Baker clamp on the web:
1. A diode between VAS input and VAS output.
2. A diode between VAS output and some fixed voltage.
Which one is correct? (I thought no. 1)
Cheers,
Edmond.
IMHO open loop bandwidth does matter. Assuming the same closed loop bandwidth (let's say 100kHz), in an amplifier with say 1kHz open loop bandwidth, the feedback loop has much more work to do between 1kHz and 100kHz, where as for an equivalent amplifier with open loop bandwidth of say 30kHz, the feedback loop only needs to start working hard from 30kHz onwards. Below 30kHz (in the audio band), the feedback loop is working with a linear open-loop transfer function. And there is probably not much energy in the music above 30kHz anyway.
Just an opinion.
Patrick
Just an opinion.
Patrick
Re: Re: Re: Re: VAS buffer
Hi Terry,
Whether I really need a damper resistor, I'm not sure. Anyhow, I'll reserve some board space for even two damping resistors, as these will also ease the PCB routing (acting like jumpers for the crossing collector traces).
Cheers,
Edmond.
Terry Demol said:It depends where and how it is used.
Inside a feedback loop there could be issues, as I can confirm I have
used this circuit myself and other variations of it.
Another more stable variation is join the collectors Q1/Q2 together
and from this point connect a damping resistor to 'out'. This will
resistively isolate collectors of Q1/2 from emitter of Q3/4 which
can be a good thing.
Yes, the connection scheme I described will loose some VCE on Q1/2
but usually Q1/2 in these configurations are run at very low currents
so it is not a big issue.
[snip]
cheers
Terry
Hi Terry,
Whether I really need a damper resistor, I'm not sure. Anyhow, I'll reserve some board space for even two damping resistors, as these will also ease the PCB routing (acting like jumpers for the crossing collector traces).
Cheers,
Edmond.
I would even reserve space for capacitors (from the input's collectors to an AC-GND) in addition to the resistors (making them an RC in effect, and with the resistors going to the emitters to save most of the available Vce) ). With this type of cascoding a diamond buffer I've had problems with the direct connection and non-isolated capacitive loads, both in simulation and on the breadboard. Stopping the cascoding action for RF helped stability quite a bit.
- Klaus
- Klaus
Re: Re: Re: Re: Re: VAS buffer
Both.
A Baker clamp is a diode configuration that prevents a bipolar transistor to saturate. VAS, or non-VAS it doesn't matter. Take a look here:
http://www.ieeta.pt/~alex/docs/ApplicationNotes/Rectifier Applications Handbook.pdf
page 175. Download this handbook fast, before it's pulled. It's the best resource I've ever seen on the topic.
Edmond Stuart said:
Both.
A Baker clamp is a diode configuration that prevents a bipolar transistor to saturate. VAS, or non-VAS it doesn't matter. Take a look here:
http://www.ieeta.pt/~alex/docs/ApplicationNotes/Rectifier Applications Handbook.pdf
page 175. Download this handbook fast, before it's pulled. It's the best resource I've ever seen on the topic.
Bob Cordell said:
No, there is more to it than that.
An all-FET design open loop amplifier with a bandwidth of 100 kHz sounded very good -- open, extended, and detailed in the top octaves. But an open loop amp with JFET inputs and everything else bipolar with a bandwidth of 150 kHz sounded rolled off, with no air or detail in the top octaves. Extending the bandwidth of that design to 250 kHz solved the problem.
So something else is going on, but I have no idea what it is.
The main lessons here are:
a) The only way to know how various parameters (eg, distortion, bandwidth, et cetera) will affect the actual sound of the amp is to listen to the amp. Simulations are of no help. Theoretical discussions of the limits of human hearing are of no help.
b) There are still many, many factors that affect the sound of an audio circuit that are not well understood. Many people (even those who frequent these forums and should be expected to know better) act as though we have achieved a degree of technical perfection such that further improvements in sonic performance are either non-existent or negligible. However, anyone who takes the time to listen knows that this is not the case.
Re: Re: Re: Re: Re: Re: VAS buffer
Yes, a truly great book. It is the best resource available for choke-input filtering. For some unknown reason On-Semi removed it from their website a couple of years ago. I'm glad that somebody else is still making it available. As syn08 said, get it while you can.
syn08 said:
Yes, a truly great book. It is the best resource available for choke-input filtering. For some unknown reason On-Semi removed it from their website a couple of years ago. I'm glad that somebody else is still making it available. As syn08 said, get it while you can.
Charles Hansen said:
Charles,
This is the very first lesson I learned coming from tubes to solid state. A solid state circuit has to have half-again to twice the bandwidth--minimum--to sound as relaxed or open as a tube amp. I can't wrap my head around this point because it "shouldn't" be that way, yet it is.
No, I haven't done the bipolar/FET comparison, but having had exactly the same experience between tubes and FETs, I see no reason to doubt you.
This observation has led me to ever-increasing bandwidths in an attempt to duplicate or--a plea to the audio gods: exceed--that ease that tubes do so well. At this point, I'm running .5Mhz in the GR-25. At the rate I'm going, I'll be up into Spectral territory (ca. 1MHz) before long.
But for cryin' out loud...why?
I'm not so stupid as to tell myself I don't hear what I hear, but I'd love to have a rational explanation for it.
Grey
Re: Re: Re: Re: Re: Re: VAS buffer
A clamp is a clamp is a clamp, obviously.
But what is a Baker clamp? Both?
According to your reference, it's case no.1, i.e. a diode between the VAS output and the VAS input, exactly as I have always thought.
The other method is based on something different: brute force, while the former method is based on NFB.
syn08 said:
A clamp is a clamp is a clamp, obviously.
But what is a Baker clamp? Both?
According to your reference, it's case no.1, i.e. a diode between the VAS output and the VAS input, exactly as I have always thought.
The other method is based on something different: brute force, while the former method is based on NFB.
Re: Re: Re: Re: Re: Re: Re: VAS buffer
Not all transistors are in a VAS. The Baker clamp concept applies to switching stages as well. That would be barely NFB, as we know it from the linear world.
PS1: You are welcomed.
PS2: Both your GF and myself were right.
Edmond Stuart said:
Not all transistors are in a VAS. The Baker clamp concept applies to switching stages as well. That would be barely NFB, as we know it from the linear world.
PS1: You are welcomed.
PS2: Both your GF and myself were right.
Re: Re: Re: Re: Re: VAS buffer
Hi Edmond,
There are some advantages of the 1 x damper resistor with joined
Q1/2 collector arrangement, but as you say it is not always required.
The advantage is that you can run a larger damper resistor and
more current through Q1/2 without loosing voltage headroom
across those transistors.
FWIW, I have never seen this arrangement used elsewhere and it
is a useful design tool.
cheers
Terry
Edmond Stuart said:
Hi Edmond,
There are some advantages of the 1 x damper resistor with joined
Q1/2 collector arrangement, but as you say it is not always required.
The advantage is that you can run a larger damper resistor and
more current through Q1/2 without loosing voltage headroom
across those transistors.
FWIW, I have never seen this arrangement used elsewhere and it
is a useful design tool.
cheers
Terry