Scott,
So, you are saying that there are no observable differences between (say) a 1206 Thick Film Resistor and a 1206 Thinfilm Resistor?
Ciao T
So, you are saying that there are no observable differences between (say) a 1206 Thick Film Resistor and a 1206 Thinfilm Resistor?
Ciao T
In what context? Termination metalurgy, thermocouple effects, TC? Excess noise? Used in mV level circuitry I might be hard pressed to measure differences.
EDIT - Here's something for everyone, noise analysis of resistors from LIGO https://dcc.ligo.org/public/0002/T0900200/001/current_noise.pdf Yes thick film resistors have more excess noise but this is not manifest in many low level applications. We use Digi-key 0805,1206 thick film exclusively in all demo cards, etc. never had any unexplained performance shortcoming.
Last edited:
Hi,
How about something non too fancy and direct such as noise and HD/IMD non-linearity?
Any observable differences, using decent instrumentation, between thick film and thin film resistors, in your considered opinion?
Ciao T
How about something non too fancy and direct such as noise and HD/IMD non-linearity?
Any observable differences, using decent instrumentation, between thick film and thin film resistors, in your considered opinion?
Ciao T
Noise yes, but I would have to consider the effects in a particular circuit. Ed Simon needed carbon comp to see the effects of excess noise sidebands. Our tax dollars paid for that LIGO study no need to repeat the work.
The THD differences of -120dB vs -130dB don't excite me. Our audio guys use the Panasonic 1% thick film resistors for 24 bit test boards. So we have a suspect. Anyone up for a Panasonic thick film vs Vishay thin film shoot out? We could build two line amps (oops been there before) or how about a pair of reference DAC's?
Scott,
Hold on. So there are differences? Yes? Observable ones?
(of course there are - anyone in electronics should know that for at least 4 decades or so, I am always amazed by the denials of such differences routinely engaged in by people who should know better)
We may debate audibility and in a complete unit the differences between the resistors may be swamped by the systems own flaws.
However, there is a little corollary here...
In the end HD, IMD and Noise are indicators that there is something non-ideal afoot.
And stuff that shows rather low measured THD can have surprising audibility of this low, while a different source for HD can have very low audibility, despite being in single figures.
So, -120dB 4th and lower HD vs. -130dB 4th and lower HD does not excite me either.
BTW, what is it with that Vishay Resistor obsession? I don't really like them much, if you lay out that much long green, there is much better stuff. Reminds me of Black Gate cap's, never cared for them either...
Ciao T
Hold on. So there are differences? Yes? Observable ones?
(of course there are - anyone in electronics should know that for at least 4 decades or so, I am always amazed by the denials of such differences routinely engaged in by people who should know better)
We may debate audibility and in a complete unit the differences between the resistors may be swamped by the systems own flaws.
However, there is a little corollary here...
In the end HD, IMD and Noise are indicators that there is something non-ideal afoot.
And stuff that shows rather low measured THD can have surprising audibility of this low, while a different source for HD can have very low audibility, despite being in single figures.
So, -120dB 4th and lower HD vs. -130dB 4th and lower HD does not excite me either.
BTW, what is it with that Vishay Resistor obsession? I don't really like them much, if you lay out that much long green, there is much better stuff. Reminds me of Black Gate cap's, never cared for them either...
Ciao T
Who ever denied that. Resistors have a TC, thermal mass, and thermal capacity. An all physics simulator can derive the usually tiny thirds from this. This is LIGO, NIST, etc. stuff where people are trying to add the 10th digit onto some physical measurement. You can almost count the atoms in our thin film and we are down <-120dB. As for dead zones and discontinuities (speaking of some of ED's pictures), I can find no references in the physics community for these phenomena.
There are properties of components not explained by physics that matter only to audio? I find this a fairly extraordinary and fatuous claim.
So capacitors mis-behave in a sound field? Here's a quote from the AMA on MRI.
"Various types of acoustic noise are produced during the operation of an MRI system. Problems associated with acoustic noise for patients and healthcare professionals include annoyance, verbal communication difficulties, heightened anxiety, temporary hearing loss and, in extreme cases, the potential for permanent hearing impairment."
Considering this weeks events I will bow out of these discussions for a while.
Hi Scott,
So we agree, there are observable differences and they CAN, depending on circuit design be sufficient to be observable above the noisefloor?
That would be indeed such a claim. I do not remember anyone making any such.
Certainly, there are properties of many things in the world around us that current physics do not fully and completely explain (even it provides us with models that are usually sufficiently in day to day use to not show glaring problems) and electronic components, including those used in audio, are no exception.
However, one need not even to go to such length. There are some quite obvious problems with many "generic" components that become very obvious on simple disassembly. Terminations made by press fit between dissimilar conductive materials or by spraying on some metal dust, capacitors that are wound so they create a lot of inductance (variable with signal current and sound field) and methodes of connecting the wound part that use only two points, thsu maximising the inductive effects and so on.
Many of the "audiophile approved" components tend to be free from these gross problems and usually where not originally designed for audio either, the origins are usually military/industrial where there is no subjective "it sounds better".
They sure do. And it is easy to measure.
Even more, if you vary the current through them and or voltage across them enough some make as much noise as transistors in output stages or Tube output transformers (which is quite a lot actually).
What always amazes me is just how noisy all sorts of Amplifiers are, when you do a sinewave sweep on the testbench at something close to rated power...
We do not even want to talk about distortion in Cap's (hint, Black Gates make many a tube amplifier look quite good).
Incidentally, some types of Capacitors in my RIAA Circuit where so microphonic, it was distinctly unfunny. The choice of capacitors made from "silver leaf and slivers of stone" certainly put an end to that.
What I find staggering is that all this has been around and documented for decades, yet there are still those who argue "components make no difference" and you seem be taking their side, in direct contradiction of extensive and serious studies on the matter.
So, where does this leave us?
I know that components matter, that the differences are clearly measurable in many cases and that documentation of these phenomena has been on record for ages. It is also clear that there is at the very least the POTENTIAL for differences that may not be directly observed using traditional measurements which current physics is unaware off, cannot explain and quite frankly cannot even perceive.
I am willing to debate the possibility or not of audibility, how one may implement suitable listening tests (actually, I have made such suggestions in writing before and will happily quote myself).
I never got to meet Bob (shame about his passing and especially the way of it) but he actually was aware of quite a bit of stuff that seems at first blush just BS, parallels design techniques applied in audio and does seem to matter outside audio.
Ciao T
So we agree, there are observable differences and they CAN, depending on circuit design be sufficient to be observable above the noisefloor?
That would be indeed such a claim. I do not remember anyone making any such.
Certainly, there are properties of many things in the world around us that current physics do not fully and completely explain (even it provides us with models that are usually sufficiently in day to day use to not show glaring problems) and electronic components, including those used in audio, are no exception.
However, one need not even to go to such length. There are some quite obvious problems with many "generic" components that become very obvious on simple disassembly. Terminations made by press fit between dissimilar conductive materials or by spraying on some metal dust, capacitors that are wound so they create a lot of inductance (variable with signal current and sound field) and methodes of connecting the wound part that use only two points, thsu maximising the inductive effects and so on.
Many of the "audiophile approved" components tend to be free from these gross problems and usually where not originally designed for audio either, the origins are usually military/industrial where there is no subjective "it sounds better".
They sure do. And it is easy to measure.
Even more, if you vary the current through them and or voltage across them enough some make as much noise as transistors in output stages or Tube output transformers (which is quite a lot actually).
What always amazes me is just how noisy all sorts of Amplifiers are, when you do a sinewave sweep on the testbench at something close to rated power...
We do not even want to talk about distortion in Cap's (hint, Black Gates make many a tube amplifier look quite good).
Incidentally, some types of Capacitors in my RIAA Circuit where so microphonic, it was distinctly unfunny. The choice of capacitors made from "silver leaf and slivers of stone" certainly put an end to that.
What I find staggering is that all this has been around and documented for decades, yet there are still those who argue "components make no difference" and you seem be taking their side, in direct contradiction of extensive and serious studies on the matter.
So, where does this leave us?
I know that components matter, that the differences are clearly measurable in many cases and that documentation of these phenomena has been on record for ages. It is also clear that there is at the very least the POTENTIAL for differences that may not be directly observed using traditional measurements which current physics is unaware off, cannot explain and quite frankly cannot even perceive.
I am willing to debate the possibility or not of audibility, how one may implement suitable listening tests (actually, I have made such suggestions in writing before and will happily quote myself).
I never got to meet Bob (shame about his passing and especially the way of it) but he actually was aware of quite a bit of stuff that seems at first blush just BS, parallels design techniques applied in audio and does seem to matter outside audio.
Ciao T
Thorsten,
I deeply respect your commitment to what you are doing. I never said I was on the components don't matter side, I just object to the "cut line" where it does not matter and the hubris with which the subjectivists are prone. I could easily find dozens of claims that a switch from one brand of decent resistors to another sent everyone running out of the room with their hands over their ears.
I have several unprintable missives from Bob that the audiophiles would not appriciate.
As for the pole at 20 Hz is forced not agree with you.
There's at 20 Hz there is no pole. Yes, the amplitude of the voltage at 20 Hz normalized. But this is not a pole. This is the origin.
About 50 000 Hz in general will not say - the question irrelevant. Hope among us bats?
Do not you think that automatically substituting in his brain frequency of 20 Hz and the beginning of the scale - 0 Hz you get into a trap?
You can easily and discreetly made a grave mistake! But do not worry you're not the first - this logical error made all bigheads, at the dawn of audio via the electrical path.
The paradox of difference to your unconscious reasoning and practical engineering experience of sound from the fact that govrorit to work with the sound must be from 0 Hz?!
Actually if you look at the frequency range from 0 to 20 Hz there is found an interesting frequency. Of course you can guess what is the frequency of the Schumann. The frequency of 7.5 Hz or so.
I know that the ability to accurately reproduce the audio device determines the frequency of their musicianship. It is an indisputable fact. Especially fast I understand those who are engaged in shellac 78 rpm.
At the moment of transition from shellac to vinyl, that is, a mechanical method of recording, the power storage technology of music has suffered substantial losses.
But all hurry to make money and no one noticed the degradation of sound. Or maybe it was done intentionally? Greens fellows no irrevocable. And our leaders ability to make some of them do business.
But where there is little money and little truth music.
From the natural range of audio lost 20 Hz but none of the engineers did not notice the loss!
Ha Ha Ha! And that's not what you an ultrasound to Batman - this is what one hears and feels. This discreet frequency of 7.5 Hz, bears information about the music! Eliminated all which is lower than 20 Hz. Psychophysics has not been canceled ... Who says that music is physical not psychological? Only liars and priests for their business psychophysics.
I know that the ability to accurately reproduce the audio device determines the frequency of their musicianship. It is an indisputable fact. Especially fast I understand those who are engaged in shellac 78 rpm.
At the moment of transition from shellac to vinyl, that is, a mechanical method of recording, the power storage technology of music has suffered substantial losses.
But all hurry to make money and no one noticed the degradation of sound. Or maybe it was done intentionally? Greens fellows no irrevocable. And our leaders ability to make some of them do business.
But where there is little money and little truth music.
From the natural range of audio lost 20 Hz but none of the engineers did not notice the loss!
Ha Ha Ha! And that's not what you an ultrasound to Batman - this is what one hears and feels. This discreet frequency of 7.5 Hz, bears information about the music! Eliminated all which is lower than 20 Hz. Psychophysics has not been canceled ... Who says that music is physical not psychological? Only liars and priests for their business psychophysics.
Hi,
Okay, I will have mercy and put this question and the Questor out of their misery.
The answer is of course "Both".
The minuscule conversion error due to fact that PI is involved in the conversion math is generally roundly and soundly ignored and left to obsessive mathematicians.
There are many areas where PI is involved in engineering and a reasonable approximation suffices even to ensure supersonic jet fighters keep flying.
Ciao T
Okay, I will have mercy and put this question and the Questor out of their misery.
The answer is of course "Both".
The minuscule conversion error due to fact that PI is involved in the conversion math is generally roundly and soundly ignored and left to obsessive mathematicians.
There are many areas where PI is involved in engineering and a reasonable approximation suffices even to ensure supersonic jet fighters keep flying.
Ciao T
I thought that by asking a technical (on topic) question we could move this thread along more productive lines.
Let's take the 318us.
Does the standard define this turn over point as 500Hz (318.31 to 6 sigfig) or 500.5 Hz (317.992 to 6sigfig) or as ~500.487Hz (318us to 6 sigfig) [sigfig = significant figures/digits]
I don't want a "don't know" answer, nor a "don't care" answer, nor do I want a "put down" answer.
What does RIAA Equalisation Standard define?
Let's take the 318us.
Does the standard define this turn over point as 500Hz (318.31 to 6 sigfig) or 500.5 Hz (317.992 to 6sigfig) or as ~500.487Hz (318us to 6 sigfig) [sigfig = significant figures/digits]
I don't want a "don't know" answer, nor a "don't care" answer, nor do I want a "put down" answer.
What does RIAA Equalisation Standard define?
Hi,
A variety of standards have defined this with the time constants and listed next to them the TRUNCATED (read round) numbers of Frequency.
So you can actually see on the paper something like "3180uS (50Hz)".
Clearly those setting standards thought the difference insubstantial and immaterial. If one wants to be anal retentive to power of 10 it would seem that statement of the time constant takes precedence over the frequency.
Now, given that difference between 3180uS and 50Hz is well below 0.1% there are more important things to keep awake at night before splitting hairs to this level, for example, where you are going to get the precise 0.01% tolerance resistors and capacitors to make such a precise RIAA Network and if Neumann ever ensured their cutters conformed and so on.
Honestly, you are just taking a gnat and you have so far blown it up to Elephant size and seem to be unwilling to stop. Have fun.
Ciao T
A variety of standards have defined this with the time constants and listed next to them the TRUNCATED (read round) numbers of Frequency.
So you can actually see on the paper something like "3180uS (50Hz)".
Clearly those setting standards thought the difference insubstantial and immaterial. If one wants to be anal retentive to power of 10 it would seem that statement of the time constant takes precedence over the frequency.
Now, given that difference between 3180uS and 50Hz is well below 0.1% there are more important things to keep awake at night before splitting hairs to this level, for example, where you are going to get the precise 0.01% tolerance resistors and capacitors to make such a precise RIAA Network and if Neumann ever ensured their cutters conformed and so on.
Honestly, you are just taking a gnat and you have so far blown it up to Elephant size and seem to be unwilling to stop. Have fun.
Ciao T
Hi Andrew
There are severall curves.
Riaa uses the following values:
RIAA f----------T us
Pole 50,05-----3180
Zero 500,5-----318
Pole 2122------75
EMI uses other values:
EMI LP f----------T us
Pole 70,00-----2273,6
Zero 500,0-----318,3
Pole 2500------63,66
IMO the majority of actual vinyl uses the Riaa standard so I always use 500.5Hz (318uS). It just sounds better this way.
Gentlemen, I've taken a chainsaw to the last few pages. I've tried to retain posts that may have been a bit off topic but still had technical content. - Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.