Ed, if you check the 'Append' box on the AP digital analyzer panel you can get two FFTs in the same window, that would help comparison.
But I guess you know that?
jan didden
I'll be glad to post those except they overlay so much it is hard to read them!
I don't think he was stating that as cause and effect. Both are the result
of lower open loop gain.
Nice to read Marcel van de Gevel. For same conclusions, see Bob Cordell's book, pages 500 to 505 and fig. 24.1
All this chatter over tired old papers. The world has moved on, there are many examples of amplifiers on high speed processes with simple signal paths that obviate all these old bogey men. GBW and slew rate to spare, why are you still talking about Baxandall's feedback around very low gain paper?
Thorsten, I don't need to hide from anything on Jan's site it's just that taking on these worshiped artifacts draws the pichfork wielding villagers.
Thorsten, I don't need to hide from anything on Jan's site it's just that taking on these worshiped artifacts draws the pichfork wielding villagers.
Last edited:
That's just one example of an amplifier that makes all those papers old news. It does make for no fun for some. Iggy Pop got a lot of money from Chrysler for "No Fun".
And more importantly, how does it sound?
Or are we to accept Bonsai's premise that since it performs at the measurement limits of today's THD measuring equipment that it must perforce also sound at (or beyond!) the limits of our hearing abilities?
John, why didn't you use these in the JC-3 phono stage?
I know that is a competitor's product. But obviously ADI has similar products. The TI datasheet gives very little useful information (except for the unusual fact that the output stage is a complementary common-emitter affair.
Just as many are interested to find out the circuit details of the Blowtorch, it might be interesting to find out the circuit details of these new op-amps that make everything else "old news". Is there anything you can share with us? Or is it like Bybees' classified quantum filters?
I found schematics, technical details, and measurements on TI's website.
I absolutely did NOT say that. And if my statement could interpreted in that way, then I was not clear enough.
What I have suggested is a bit less dogmatism, a bit more perspective and a great deal more latitude. I don't doubt your discrete designs sound wonderful (and I mean that). Now what I'd like to see from your side is an acceptance or an acknowledgement that integrated opamp solutions can also deliver a pretty fine listening experience. We are talking very small shades of difference here - no need bring out the pitchforks every time a contributor suggests an op amp might actually deliver a great result- measured or sonic.
Last edited:
Hi Scott
I don't know. Looking at the datasheets I'm very unimpressed.
The chip seems rather sensitive to common mode effects, slew rate and gain bandwidth is pedestrian. And no matter how hard I try, based on the datasheet I cannot forsee any Audio Application for this IC, where it would actually provide decent performance.
I think I take your AD797 or if I can tolerate 3dB more LF noise I'll take the OPA1632 at the very least.
BTW, has AD got a credible moder equivalent to the AD811?
I really liked the AD811 as MC Cartridge input AMp.
Ciao T
I don't know. Looking at the datasheets I'm very unimpressed.
The chip seems rather sensitive to common mode effects, slew rate and gain bandwidth is pedestrian. And no matter how hard I try, based on the datasheet I cannot forsee any Audio Application for this IC, where it would actually provide decent performance.
I think I take your AD797 or if I can tolerate 3dB more LF noise I'll take the OPA1632 at the very least.
BTW, has AD got a credible moder equivalent to the AD811?
I really liked the AD811 as MC Cartridge input AMp.
Ciao T
Hi Jan,
This is only partially correct. If I use up the excess loop gain in local degeneration and if needed local loop feedback my error value will be as before (unless it derives solely from the final stage), however the overshoot will be reduced or gone.
In other words, if we lower loop gain at low frequencies in a manner that makes good use of the reduction to linearise the circuit, we have only made improvements, no dis-improvements will be observed.
Sorry, I really do not feel like writing an LTE about such obvious obvious stuff, so don't ask.
Ciao T
Tsk tsk, what length people will go to to defend an orthodxy neither worthy of nor in any possibility of defending
This is only partially correct. If I use up the excess loop gain in local degeneration and if needed local loop feedback my error value will be as before (unless it derives solely from the final stage), however the overshoot will be reduced or gone.
In other words, if we lower loop gain at low frequencies in a manner that makes good use of the reduction to linearise the circuit, we have only made improvements, no dis-improvements will be observed.
Sorry, I really do not feel like writing an LTE about such obvious obvious stuff, so don't ask.
Ciao T
Tsk tsk, what length people will go to to defend an orthodxy neither worthy of nor in any possibility of defending
Wow. So you found details about this "pre-output driver" black box? Care to share a link?
Btw, OPA1611 has 4,5mA max quiescent current, 1641 half of that. How much flows through the output stage? Common emitter with low bias current into low impedances is not what i would call as linear as possible befor closing the loop. And loop gain depends on output impedance.
I like transimpedance outputs (as the blowtorch is btw) but they must run class a. If i had to use this thing, forcing the output into class A with a resistor or ccs towards a rail would be mandatory.
I don't think he was stating that as cause and effect. Both are the result
of lower open loop gain.
Not sure I get your drift Nelson. He was specifically commenting on the practise of loading the Vas to obtain a wider OL bandwidth and showed that that actually makes things worse. Because it lowers loop gain (and thus closed loop linearity) below Fc while doing nothing for the high-freq linearity. So yes it's a result of lower loop gain.
@Thorsten: Yes, there must be myriad ways to skin the linearity cat. But as mentioned above, Marcel's comment was to the Vas loading as a panacea.
jan didden
- Status
- Not open for further replies.