Well, order a million at 0.10 cts each, I'm sure they will reopen the line, with Scott putting in a good word for you. Just $10k!
You'll be the George Soros of Hi-End 8-legs! Howzzat for cornering the market
Jan
Well, order a million at 0.10 cts each, I'm sure they will reopen the line, with Scott putting in a good word for you. Just $10k!
You'll be the George Soros of Hi-End 8-legs! Howzzat for cornering the market
Jan
100K...
National Semi Linear dept. charge Ford Motor 50K for each discrete prototype IC for it built for testing on the Electronic Ignition system back in 73 when I worked there.
My last version won't fit in IC, even though my career started from IC laboratory in Tomsk Scientific-Research Institute of Semiconductor Devices (NIIPP)
Even though it is actually a fashionable CFA
Wavebourn - When David started this song I got goosebumps!...
Even though it is actually a fashionable CFA
Wavebourn - When David started this song I got goosebumps!...
you can buy AD846, just not for pennies apiece
Rochester Electronics Homepage >buy>search>ad846a
the quantities may be limited by Pass standards but should cover a fair bit of diy interest
I did build a prototype that used AD846 in the the early 90's - never got beyond the bench, I don't even remember if we bought a whole tube or just used samples - sorry, Scott
Rochester Electronics Homepage >buy>search>ad846a
the quantities may be limited by Pass standards but should cover a fair bit of diy interest
I did build a prototype that used AD846 in the the early 90's - never got beyond the bench, I don't even remember if we bought a whole tube or just used samples - sorry, Scott
Last edited:
The original meaning of the term current feedback (decades before a bunch of op-amp guru's hijacked the term) is exactly the same as what you call current output amplifier. Hence the misunderstanding, and hence my statement that it's good to define what one means by the term before using it.
Thanks for your comments, they belong to those which can further the debate.
Now say there is a distortion mechanism internal to the whole amplifier, well there is identically zero volts across the + to - input so the error appears totally as the voltage caused by the CURRENT through Ra1 into the - input.
OP3 is hoped to be a VFA....
OP3 inverting input which has a high impedance controls the voltage at T1 emitter (which is at low impedance) and across the Rb3 forcing it to be almost null, i.e. making them an almost perfect virtual earth.
But, despite the apparence, it is not the current in Ra3 which controls the circuit but the minuscule voltage difference (justifying the underlined almost above) between the op-amp in+ and in- inputs. Once again the feedback process is seen to rely on a voltage difference, the circuit is under the control of th VF op-Amp.
*
In the above circuit,
- discarding Tr3
- connecting the op-amp outpu to in- and to the output to the common point of Ra3 and R3b
- sensing the current circulating in the power supply connections
gives an equivalent which is sometimes seen.
The frequency compensation necessary for stability may compromise the performances at high frequencies when compared to those of a circuit based on a diamond input.
*
I have built these with IC's BTW and they can work.
I thought of such a circuit, something like below. The input stage is based on a Schlotzaur circuit. Of course, I won't never describe the complete amplifier as being a CFA because it is controlled by the voltage difference between the in+ and in- inputs.
The Schlotzaur circuit has very low distortion, but the current mirrors may compromise the overall performances.
Attachments
Sorry, Marcel, I've never understood it that way.
I spent 7 years earlier on in my career designing industrial instrumentation. One of the signals used extensively in industry is the 4-20mA loop. They used to be everywhere - maybe a bit less today because of fieldbus etc - to transmit process variables between instruments measuring temperature, pressure and so forth. I never once heard the term 'current feedback' to describe circuit operation and I certainly never used it. The output current was sensed across a resistor (typically 50 or 100 ohms) and that voltage fed into the f/back - input of a VFA opamp with the control on the + input.
Separately, the first article I ever read on CFA's was by SF in edn - probably around 1989 or 1990 and I don't mind admitting it took me a while to get my head around it.
But, I doubt this definition of terms will ever be fully resolved, and neither will the VFA/CFA thing either!
I spent 7 years earlier on in my career designing industrial instrumentation. One of the signals used extensively in industry is the 4-20mA loop. They used to be everywhere - maybe a bit less today because of fieldbus etc - to transmit process variables between instruments measuring temperature, pressure and so forth. I never once heard the term 'current feedback' to describe circuit operation and I certainly never used it. The output current was sensed across a resistor (typically 50 or 100 ohms) and that voltage fed into the f/back - input of a VFA opamp with the control on the + input.
Separately, the first article I ever read on CFA's was by SF in edn - probably around 1989 or 1990 and I don't mind admitting it took me a while to get my head around it.
But, I doubt this definition of terms will ever be fully resolved, and neither will the VFA/CFA thing either!
I also worked for a time designing 4-20ma industrial instrumentation, don't recall fellow engineers there being particularly concerned about academic classifications of circuits
so I would say you are reasoning from a "provincial" perspective if you didn't learn the feedback classifications in uni, didn't work with people that used them
you should welcome the broader perspective even if that happens to means you need to drop a certain defensiveness about your experience being sufficient to decide on your exclusive interpretation of 'current feedback' etymology, application
so I would say you are reasoning from a "provincial" perspective if you didn't learn the feedback classifications in uni, didn't work with people that used them
you should welcome the broader perspective even if that happens to means you need to drop a certain defensiveness about your experience being sufficient to decide on your exclusive interpretation of 'current feedback' etymology, application
Last edited:
Since we doing little more than name dropping now, DF (who hasn't as of yet replied to SF's response) cites Barry Gilbert to buttress his position, probably unaware that BG was a signatory to SF's original rebuttal to MK's Audio Express article.
And I am not sure what Kevin's initial response (the one beginning "Ho hummm") brings to the table. He just ranted about how VF amplifiers can be made to have the slewing capability of CF amplifiers. Had he actually read the piece by SF that he was directly replying to he may have noticed that SF actually schematically detailed and examined a buffered-input VFA having CFA-like slewing capability, and even gave the manufacturers part number for an example of one such device. In any case, KA's response doesn't directly address the issue currently being thrashed out here, so what is your point?
To a disimpassioned observer this is just unintentional comedy. In this increasingly ridiculous feud there seems to be a number of independent issues all hopelessly mixed up and conflated, such as:
1) A false operational equivalency being drawn between the established VFA and CFA typologies that ignores the role the feedback network plays in frequency compensating the CF amplifier. So long as op-amps that require an external compensation capacitor remain out of favor (for reasons logical or illogical), there will always be a marketing niche and an edge to plug for CFA amplifiers due to their degree of gain-bandwidth independence.
2) The absolute technical/theoretical correctness of the established CFA/VFA nomenclature.
3) That that "current-on-demand" slewing property is not topologically restricted to CF amplifiers only.
I want to thank whoever put it up, the tip for Rochester Electronics who actually has many, many AD846 IC's. What a resource! I would order some but their minimum is 29 for this device. However, IF I were to design a custom 1 off pro circuit that needed to be smaller than my normal designs, I would order these immediately.
John, that was discovered by octopart.com which is a search engine that YOU need to get familiar with.
I pledge, promise, and affirm that I will buy $100 worth of AD846s from you, if and when you purchase the "minimum order quantity" (however that works!) of AD846s from Rochester. It reduces your financial investment by a hundred buxx.
I pledge, promise, and affirm that I will buy $100 worth of AD846s from you, if and when you purchase the "minimum order quantity" (however that works!) of AD846s from Rochester. It reduces your financial investment by a hundred buxx.
what about the translinear second generation plus type Current Conveyor [CCII+] followed by a translinear voltage buffer------ The AD844? -- Includes pin for external compensation and is pin compatable with VFB amps.
One of the characteristics of CMOP's (Current-Mode op-amps) like the AD844 that I like is the expansive nature of the input - rather than compressive..... they take large signal inputs and pulses with high linearity.
THx-RNMarsh
One of the characteristics of CMOP's (Current-Mode op-amps) like the AD844 that I like is the expansive nature of the input - rather than compressive..... they take large signal inputs and pulses with high linearity.
THx-RNMarsh
Last edited:
That is sometging that also leads to recurring confusion here. Indeed, in the original 4-class classification, current feedback denoted feedback representing output current, but the feedback signal itself is a voltage (normally).
In the 'modern' CFA term, it denotes feedback that represents output voltage, but is returned as a current - or this is the claim.
The discussion raging now is whether, as is claimed, the feedback is returned as a current. This is separate from the classification issue, which I believe everyone accepts as an unfortunate state of affairs, but will not go away. ADI, TI, LTC, who have you, will not change their term CFA in the modern sense.
So we are debating: is the feedback returned to the inv input a current or not.
Unfortunately, when cornered, people like MK switch to the 4-classification issue again and again to cop out of the CFA discussion when it is not going well for them. A clear sign that they are not interested in solving the issue but more into arguing for it's own sake.
Such tactics will keep it going indefinitely.
Jan
A modern CFB amplifier that makes me happy now is the THS3001HV. It'll handle ±18.5V rails, has around -100dB distortion at 1MHz, it's built on a dielectrically isolated process, and it has a nicely high standing current. It's a relatively simple op amp with relatively low gain, but it's really easy to run it as a follower inside of the feedback loop of a higher gain, LF oriented op amp, with the point of removing all signal based dissipation in the main op amp that controls the feedback loop. I haven't finished a prototype with it, but it seems very promising. Further, if you don't need much OL gain or tight DC, I bet it would sound pretty good by itself at low gains. It's only available in a slightly small SMD package, but it's one of the few modern ultra high speed amplifiers that can run at high voltage rails.