Such as?
Resistors are made out of dirty sand. As a result, ICs sound dirty as mud.
Waly, it would be interesting for someone with up-to-date experience with IC design to tell us about the state of resistor non-linearity in modern IC's both for industrial and consumer applications. For example, what sort of linearity would we expect of the resistors in a device like the 4558, perhaps compared with the resistor linearity of a single resistor if it were measured separately, from an AD797?
The rest of that post, interesting though it may be, did not give an answer.john curl said:Max Headroom, you asked me a question. I thought that you deserved an answer. I gave you a short answer, based on my history and experience with IC's.
Now we have resistor non-linearity. Is this the answer? One answer (mentioned by someone other than JC), among 'lots of reasons'. Could someone in the know (Scott?) tell us about linearity (or otherwise) of silicon resistors?
JC has N reasons not to use opamps. He still has N-1 of them to tell us about.
Of course, unless you're going to somehow isolate those resistors, it's irrelevant. The issue is performance of the gain block, which is set by external resistors. I'm not sure we've even gotten to N-1, the only thing that's been revealed is that he had some problems in 1968, which isn't exactly a valid technical argument.
20 bit Vout DAC on a chip? = kinda, sorta linear resistors?
AD5791 14 bit R-2R resistor ladder DAC + 63 unit weighted lsb R
Analog Devices : Rarely Asked Questions (RAQs) : Resistors in Analog Circuitry
AD5791 14 bit R-2R resistor ladder DAC + 63 unit weighted lsb R
Analog Devices : Rarely Asked Questions (RAQs) : Resistors in Analog Circuitry
Last edited:
Oh dear - the implication of that resistor RAQ is that resistors in chips can be every bit as good as discrete resistors. N-1 +1 = N?
The rest of that post, interesting though it may be, did not give an answer.
Now we have resistor non-linearity. Is this the answer? One answer (mentioned by someone other than JC), among 'lots of reasons'. Could someone in the know (Scott?) tell us about linearity (or otherwise) of silicon resistors?
JC has N reasons not to use opamps. He still has N-1 of them to tell us about.
Thin film resistors would be the most relevant the non-linearity is hard to measure. Check the AD621 G = 100 with two on-chip resistors and a huge difference in power density. There is plenty of secret sauce involved, the thin film sputtering targets are very carefully made to an exact very pure formulation and the contact metallurgy is more complicated than you might think.
Last edited:
But of course we know that it is easy to 'hear', thus 'proving' that we can hear things that we can't measure.scott wurcer said:the non-linearity is hard to measure
About on chip evil resistors, makes-me remember a word of Karl Lagerfeld about fluos armbands at the time they wanted to make them mandatory in every cars: "It's ugly, it fits with anything, but it can saves lives !".
On the contrary, i thing we can measure things that we cannot hear.
Are-you so sure ?But of course we know that it is easy to 'hear', thus 'proving' that we can hear things that we can't measure.
On the contrary, i thing we can measure things that we cannot hear.
Last edited:
it really is amazing how hard it is to transmit sarcasm/irony/facetiousness with text on the interwebs sometimes. its unfortunate, as adding a smiley or other emoticon takes all the fun out of it.
there is even the possibility that ive missed some here 😉
there is even the possibility that ive missed some here 😉
We're not going to circle back to Bob Widlar telling JC that 709's and 741's were not god for highend audio (was that 1969?), are we?
Can't we just compare real world up to date IC's to currently available and in production, not save surplus discrete devices to make comparisons here? Going back to 1969 or even 2000 would seem disingenuous to an argument about the current state of the art. Discrete components seem to be dropping off the radar as fast as the companies can stop producing them, and we have no real ic's that are directed at audio circuits. So who can compare an up to date IC design to the best discrete designs with currently produced devices?
Bentleys Are Irrelevant, FFS !...
What is your current reasoning for using Jfet and Mosfet based op-amps, and which applications ?.
Thanks, Dan.
Ok thanks for going to the trouble to post such long reply....interesting history......Now, don't get me wrong, I use IC's in analog designs, just not my BEST DESIGNS. They are always jfet and mosfet based, and that will probably continue for the rest of my years, to be sure.
What is your current reasoning for using Jfet and Mosfet based op-amps, and which applications ?.
Thanks, Dan.
No.It is more (or less) an ellipsoid.We're not going to circle ... are we?
This thread needs to be observed by ALMA telescope to ensure, everything is not clear enough.
I started with a history of being enthusiastic about IC's even 45 years ago. However, they never quite did what I needed from them, SO I went back to using discrete component design with great success.
Now, what about more modern IC's? Some are pretty darn good, so long as you don't use them at their limits, however I can still get better subjective performance from discrete components, typically jfets, so I design almost exclusively with them.
If you, Max, do not know the intrinsic linearity of jfets compared to bipolars, then I just can't answer your question.
What it essentially comes down to, is the OPEN LOOP LINEARITY of an individual design. Negative feedback, while reducing steady state harmonic distortion, appears to convert the open loop nonlinearity into another form of distortion not picked up by THD or IM testers. In any case, the human ear tends to notice it.
Now, what about more modern IC's? Some are pretty darn good, so long as you don't use them at their limits, however I can still get better subjective performance from discrete components, typically jfets, so I design almost exclusively with them.
If you, Max, do not know the intrinsic linearity of jfets compared to bipolars, then I just can't answer your question.
What it essentially comes down to, is the OPEN LOOP LINEARITY of an individual design. Negative feedback, while reducing steady state harmonic distortion, appears to convert the open loop nonlinearity into another form of distortion not picked up by THD or IM testers. In any case, the human ear tends to notice it.
Too late
Have trouble reading an electronic DB hood ornament ?
SuperBob : http://sutromedia.com/published/ipad-sized-photos/170096.jpg
Happy
My ears are happy with SSM2135......
😀



I started with a history of being enthusiastic about IC's even 45 years ago. However, they never quite did what I needed from them, SO I went back to using discrete component design with great success.
Now, what about more modern IC's? Some are pretty darn good, so long as you don't use them at their limits, however I can still get better subjective performance from discrete components, typically jfets, so I design almost exclusively with them.
If you, Max, do not know the intrinsic linearity of jfets compared to bipolars, then I just can't answer your question.
What it essentially comes down to, is the OPEN LOOP LINEARITY of an individual design. Negative feedback, while reducing steady state harmonic distortion, appears to convert the open loop nonlinearity into another form of distortion not picked up by THD or IM testers. In any case, the human ear tends to notice it.
My ears are happy with SSM2135......
😀



Ok, thanks.What it essentially comes down to, is the OPEN LOOP LINEARITY of an individual design. Negative feedback, while reducing steady state harmonic distortion, appears to convert the open loop nonlinearity into another form of distortion not picked up by THD or IM testers. In any case, the human ear tends to notice it.
Dan.
- Status
- Not open for further replies.
- Home
- Member Areas
- The Lounge
- John Curl's Blowtorch preamplifier part II