Great question. In the 80's, you could hear some things beyond measurement, but we could measure a lot. Some better than others using advanced equipment and techniques. Today, we can measure well beyond any limit of human perception, but interpreting the readings requires some skill. In the 70's we could measure more than what is popularly thought. If you only used a THD meter, you were maybe a little wiser, but pretty lost. Spectrum analysis gives a full picture, and way back I used a spectrum analyser on the output of my THD meter, also a 'scope. Today I use an audio analyser - a purpose built instrument. Many do, but they are expensive.
Massive false myth. People who measure don't listen. Completely false! You will find anyone that claims they can hear what can't be measured does not have access to good enough equipment to tell, or any at all. It is important to listen as well as measure. When the two agree you are on the right track. Each keeps the other honest.
Massive false myth. People who measure don't listen. Completely false! You will find anyone that claims they can hear what can't be measured does not have access to good enough equipment to tell, or any at all. It is important to listen as well as measure. When the two agree you are on the right track. Each keeps the other honest.
Look around and see how many fellow DIYers do this. The driving factors often are "they don't make that anymore" and "I have the knowledge not many have". If the vast amounts of parts and measuring stuff have a tenfold value of what is listened to maybe it is time to clean out 😀 Been there done that. I just did not get it until I noticed those large lots on the local second hand websites are from deceased hobbyists that leave the cleaning out to their peers. When a person of the female persuasion asked me why I needed so many devices while I only listened to just 1 my eyes opened 🙂
A recent visit to a fellow DIYer was shocking. There was no room in the whole house as various audio sets were stacked up against the walls. All rooms were full of equipment and daylight was scarce. For some reason the light often does not work either and broken stuff sits there for ages. The excuses are known and it is embarrassing to recognize them. And then the final stroke.... look carefully how often the hifi equipment is used.... in many cases less than a person does that simply uses devices.
Last edited:
This gear serves my purposes well. The Linear Audio Autoranger is a great tool that overcomes the weaknesses of a sound card measurement system, most importantly their maximum input voltage. Stuart Yaniger measured it in audioXpress using an Audio Precision APx525 and an APx555: https://audioxpress.com/article/fresh-from-the-bench-the-linear-audio-autoranger-mk-ii-interface
"...this product enables users with high-quality sound cards or recording interfaces to perform measurements to the level that is usually reserved for extremely expensive professional-grade audio analyzers. Indeed, its performance required the use of the newest and most elaborate Audio Precision system to characterize, and it greatly exceeded the performance of the APx525 I use day-to-day in my lab."
I pair the Autoranger with a MOTU M4, a budget-friendly audio interface, my loopback measurements here: https://www.audiosciencereview.com/forum/index.php?threads/motu-m4-loopback-measurements.10890/
Greater performance with this setup could be achieved pairing with something like a MOTU UltraLite-mk5 or an RME ADI-2 Pro FS, however I am working with tube-based circuits primarily, so the MOTU M4 is more than satisfactory.
Last edited:
Hi L0rdGwyn,
What you have is more than adequate for tube based equipment. Stuart is a friend of mine and I am more than familiar with his work. He is careful, capable and honest. His AP gear is easier to use than mine by a long shot. What I have is somewhat easier to use than what you are using, but you get excellent results if you are careful. Having said that, you are also aware that interpreting your results takes some experience, but it is all there to see. I'm sure you also know that your test setup for whatever you are measuring is important too.
My bench is an embarrassingly massive mess at the moment, but here it is. I have other equipment I bring out as needed and you can't see the GPS receiver for the 10 MHz frequency reference. Lower right is a Keysight AC power source (6812C) I can use to inject specific noise on the AC line, or very precise voltage and energy readings. The last shot is an RTX 6001 Audio Analyser. I use it with M.I., or Arta (paid) for quickie stuff. Some equipment is used in conjunction with other pieces which gives greater insight as to what is going on.
Hi Salas,
I only wish I could keep my work area as neat as yours is.
What you have is more than adequate for tube based equipment. Stuart is a friend of mine and I am more than familiar with his work. He is careful, capable and honest. His AP gear is easier to use than mine by a long shot. What I have is somewhat easier to use than what you are using, but you get excellent results if you are careful. Having said that, you are also aware that interpreting your results takes some experience, but it is all there to see. I'm sure you also know that your test setup for whatever you are measuring is important too.
My bench is an embarrassingly massive mess at the moment, but here it is. I have other equipment I bring out as needed and you can't see the GPS receiver for the 10 MHz frequency reference. Lower right is a Keysight AC power source (6812C) I can use to inject specific noise on the AC line, or very precise voltage and energy readings. The last shot is an RTX 6001 Audio Analyser. I use it with M.I., or Arta (paid) for quickie stuff. Some equipment is used in conjunction with other pieces which gives greater insight as to what is going on.
Hi Salas,
I only wish I could keep my work area as neat as yours is.
Attachments
Lol, all of you guys, mods, administrator went off topic...
All of this will finish between a Sarmat satan 2 slam and listening music on our smartphone.
I agree with JP, we make it more by mono maniac curiosity than listen to music anymore.
Well as far we are aware of it...I like when you go outside ofnthe road guys...makes my day 🙂
Back to the worst ninjas worldwide organisation team or simple mono radio kolkoz survival grup ?
I knew Papa was a punk when he asked... I love him just for that 🙂
All of this will finish between a Sarmat satan 2 slam and listening music on our smartphone.
I agree with JP, we make it more by mono maniac curiosity than listen to music anymore.
Well as far we are aware of it...I like when you go outside ofnthe road guys...makes my day 🙂
Back to the worst ninjas worldwide organisation team or simple mono radio kolkoz survival grup ?
I knew Papa was a punk when he asked... I love him just for that 🙂
Well, you can't talk about capacitors without talking about test equipment!
On test equipment, I tend to look at it in terms of what can I measure and over what frequency range, and then try to plug the holes. Almost all of my equipment is old, but solid and reliable. I've yet to stumble onto a decent spectrum analyzer, so use the computer for that. Many years ago my bench was clean and looked like this-
There's a ton more stuff you can't see, mostly for component measurements.
As for the hoarding problem, don't worry, anyplace in the house where you can see light between the electronic equipment has been carefully plugged with DW's yarn and quilting fabric.
On test equipment, I tend to look at it in terms of what can I measure and over what frequency range, and then try to plug the holes. Almost all of my equipment is old, but solid and reliable. I've yet to stumble onto a decent spectrum analyzer, so use the computer for that. Many years ago my bench was clean and looked like this-
There's a ton more stuff you can't see, mostly for component measurements.
As for the hoarding problem, don't worry, anyplace in the house where you can see light between the electronic equipment has been carefully plugged with DW's yarn and quilting fabric.
There seems to be unanimous agreement that polyprop film caps sound better than a similar value electrolytic. With all this fancy equipment being shown off, has anybody been able to correlate measurements with what is being heard? I’m sure there are measurable differences but what I’m interested in is linking the differences to what is heard—that is the difficult part.
Hi Conrad,
Yes, nice setup. I can use the HP 4195A with the impedance test jig, but it is a pain. You are absolutely correct, you must understand and know the characteristics of the parts you use in order to do the work properly to the best of your ability. It is a learning process with a great deal of personal time invested. Many don't understand that and are happy to repeat what they have read without any investigation. Like anything I guess.
Hi ra7,
Decades of empirical data using actual real case work and multiple subjects that have correlated well. Always a good sanity test.
I wouldn't say that film types always sound better than electrolytic caps. It depends greatly on the situation, and a film cap will be massive compared to the electrolytic type. This generates it's own problems and brings in other effects. You can't blindly replace one with the other because "its better". The situation matters a great deal and you have to consider everything to make the right choice. If the part doesn't fit mechanically in the same space, don't use it. This brings to mind the insane capacitors bolted to chassis' with wires leading to a board. Absolutely stupid, and the other issues it creates outweigh any possible benefits. Clearly a sign where the person that did the work was fixated on one aspect and neglected to see the big picture. Also generally a sign that the person reads the internet and blindly follows bad advice without any real understanding of how things work.
I have had to remove so many stupid component choices to restore proper operation, and when holes are made in a chassis, it destroys the value of the equipment. This is especially sad when it is a classic piece worth thousands (well, it used to be).
Yes, nice setup. I can use the HP 4195A with the impedance test jig, but it is a pain. You are absolutely correct, you must understand and know the characteristics of the parts you use in order to do the work properly to the best of your ability. It is a learning process with a great deal of personal time invested. Many don't understand that and are happy to repeat what they have read without any investigation. Like anything I guess.
Hi ra7,
Decades of empirical data using actual real case work and multiple subjects that have correlated well. Always a good sanity test.
I wouldn't say that film types always sound better than electrolytic caps. It depends greatly on the situation, and a film cap will be massive compared to the electrolytic type. This generates it's own problems and brings in other effects. You can't blindly replace one with the other because "its better". The situation matters a great deal and you have to consider everything to make the right choice. If the part doesn't fit mechanically in the same space, don't use it. This brings to mind the insane capacitors bolted to chassis' with wires leading to a board. Absolutely stupid, and the other issues it creates outweigh any possible benefits. Clearly a sign where the person that did the work was fixated on one aspect and neglected to see the big picture. Also generally a sign that the person reads the internet and blindly follows bad advice without any real understanding of how things work.
I have had to remove so many stupid component choices to restore proper operation, and when holes are made in a chassis, it destroys the value of the equipment. This is especially sad when it is a classic piece worth thousands (well, it used to be).
Do try to understand the position of the guy that doesn't understand. Take me for instance at 48 years old trying to cobble together something that works. The fact is I could never reach an adequate level of understanding in this lifetime. Not enough time left unless I won the lottery and went to MIT. That's if a lottery winning can still pay for an Ivy league level of education. So I take what advice is reasonable. Why, because teasing usable advice out of really smart people can be tough. I have a friend with a father that was a genius re: radio antennas. Very smart man. As a layman ask him anything about the subject and he just couldn't articulate a single thing at my level. Add to that the fact some turn out to be smart and wrong. So you take a shot and live with the fact that at least your mistakes are your own. I had a very hard time finding a shop that said they'd refurb an old Adcom amplifier. 3 month wait. No idea if they're any good. So. soon, out comes the hatched.
48 years old and not enough time? Assuming you don't lose the big lottery before your time, you have plenty of time to learn far more than the basics. There's a huge amount of info on the 'net; all you have to do is sort the good from the bad. Personally, I like books. They go into more depth and I like paper better than a screen. I've spent a small fortune on technical books over the years, thinking they're appreciate. Silly me. Prices seem to be going down and, if you can stand the screen, pdfs are available for many classics. Radio antennas? Way too complicated for the likes of me.
Hi Conrad,
LOL!!!!
I'm in the same boat! When I think of how much my library cost me - holy crap! But I was also very lucky to learn at a young age to listen carefully to what people in the industry told me. That includes you, thank you. I would like to think I took the knowledge and did something good with it.
Hi iplayforme,
I do, but also consider others that read who don't post. Also those who profit from all this misinformation.
Just ask. Some people, and you will learn over time who they are, will give you the best information you can get.
I was authorized warranty for Adcom and several other brands, good quality ones. Always use common sense, and a good manufacturer will put out a good product. Always beware of people claiming to know more than real engineers who are not solid engineers themselves. There is always a good reason why things are done. I may not agree with them, but it doesn't mean what was done was wrong.
LOL!!!!
I'm in the same boat! When I think of how much my library cost me - holy crap! But I was also very lucky to learn at a young age to listen carefully to what people in the industry told me. That includes you, thank you. I would like to think I took the knowledge and did something good with it.
Hi iplayforme,
I do, but also consider others that read who don't post. Also those who profit from all this misinformation.
Just ask. Some people, and you will learn over time who they are, will give you the best information you can get.
I was authorized warranty for Adcom and several other brands, good quality ones. Always use common sense, and a good manufacturer will put out a good product. Always beware of people claiming to know more than real engineers who are not solid engineers themselves. There is always a good reason why things are done. I may not agree with them, but it doesn't mean what was done was wrong.
@iplayforme I started in this hobby in earnest at the end of 2019 and finished my first from-scratch design Q1 2020. I have no formal EE training, although some of my undergraduate coursework helps (math, physics, chemistry). Regardless, I think you can accomplish a lot as a hobbyist in a short period of time with the motivation and a singular goal in mind. Getting started, one needs to learn the hands-on skills to build, then the theory to design. For the former, a good DIY kit might be the way to break the ice. For the latter, books tailored toward the goal you hope to achieve in your first project along with asking questions to those with experience and are willing to share. I think you have plenty of time to learn and build something great!
I have now designed and built several tube amplifiers, a tube phono stage, a NOS DAC, along with a number of smaller projects along the way. Of the things I've made, this might be my favorite, an 801A single-ended triode design running in class A2, Sowter transformers. If someone had showed this to my 2019 self and told me I was going to build it, I would have laughed in their face.
Keeping things semi-relevant, it uses Jupiter copper foil coupling caps on its input (due to a mixed battery / cathode bias) as well as between the 6F5 input stage and the parallel-section 6BX7 cathode follower which drives the 801A grid positive. I've had good experiences with these caps, I use them when I'm feeling like something fancy as they are somewhat pricey.
I have now designed and built several tube amplifiers, a tube phono stage, a NOS DAC, along with a number of smaller projects along the way. Of the things I've made, this might be my favorite, an 801A single-ended triode design running in class A2, Sowter transformers. If someone had showed this to my 2019 self and told me I was going to build it, I would have laughed in their face.
Keeping things semi-relevant, it uses Jupiter copper foil coupling caps on its input (due to a mixed battery / cathode bias) as well as between the 6F5 input stage and the parallel-section 6BX7 cathode follower which drives the 801A grid positive. I've had good experiences with these caps, I use them when I'm feeling like something fancy as they are somewhat pricey.
It’s true that physically large capacitors are often better antennas for picking up mains interference than their electrolytic counterparts. However, if one were building an amplifier from scratch and the use of any electrolytics could be avoided, why wouldn’t one use film?
An assortment of HV 100 & 220uf film capacitors:
Mostly because film caps are very susceptible to the peak currents found in a power supply, you need foil caps rated for far higher currents than you would expect. Electrolytic capacitors are actually perfect for that application. You can (and should) use local bypass caps if you are worried about HF impedance.
For coupling applications, the values should be much, much lower. You really want to avoid driving an output transformer near it's cut-off frequencies, it ain't pretty. It would sound much better with restricted frequency extension than over driving the core any day.
This is all about design and selecting appropriate values for the circuit and other components you are using. You know, electrolytic capacitors are not evil. Design is about intelligent compromise and that is entirely up to you. Also keep in mind what the frequency range of the entire circuit is, and sub-circuits. Designing for 1 MHz operation in a tube amp will tend to make you make decisions at the expense of things that really matter. You can bypass signal stages with high frequency capability, but most of the current will be low frequency stuff.
There is no best capacitor or capacitor type. There is a best capacitor type for a circuit and region of operation and that depends on the actual circuit and layout. Speaking of layout, spreading a circuit out due to large components is not going to help you. You have to consider many things.
For coupling applications, the values should be much, much lower. You really want to avoid driving an output transformer near it's cut-off frequencies, it ain't pretty. It would sound much better with restricted frequency extension than over driving the core any day.
This is all about design and selecting appropriate values for the circuit and other components you are using. You know, electrolytic capacitors are not evil. Design is about intelligent compromise and that is entirely up to you. Also keep in mind what the frequency range of the entire circuit is, and sub-circuits. Designing for 1 MHz operation in a tube amp will tend to make you make decisions at the expense of things that really matter. You can bypass signal stages with high frequency capability, but most of the current will be low frequency stuff.
There is no best capacitor or capacitor type. There is a best capacitor type for a circuit and region of operation and that depends on the actual circuit and layout. Speaking of layout, spreading a circuit out due to large components is not going to help you. You have to consider many things.
What if the goal is simply to avoid the eventuality of replacing electrolytics? It seems the two main advantages of electrolytics are 1) the range of large capacitance values, 2) expense per unit C and 3) compact size vs film, with the tradeoff of much shorter lifespan than film capacitors.
I’m asking if 1, 2 and 3 don’t apply, what is the argument for dealing with the tradeoff?
I’m asking if 1, 2 and 3 don’t apply, what is the argument for dealing with the tradeoff?
Last edited:
Mouser's parametric search tables will tell you which capacitors are rated for the longest life, at the widest possible temperature excursions. Copy down the part numbers and search all your favorite suppliers for the lowest price / best availability. Or let octopart.com do this for you.
Capacitor reliability and in-service lifetime are covered in Cornell Dubilier's application guide. Flip to page 13 and enjoy the story.
_
Capacitor reliability and in-service lifetime are covered in Cornell Dubilier's application guide. Flip to page 13 and enjoy the story.
_
Attachments
Mark, are there any particular features to look for in coupling applications and AC bypass applications? For example, a high dv/dt capability.
Douglas Self's book Small Signal Audio Design, 2nd edition covers that topic pretty extensively, pp. 63-75. Including lots of distortion measurements. One of his figures is below; it tells you that when a polarized electrolytic capacitor is used in an AC coupling application, distortion is unmeasurably low, if the capacitance is at least 100 microfarads per RMS volt of signal amplitude. So when the signal amplitude is 10VRMS or less, a 1000uF electrolytic has unmeasurable total harmonic distortion.
There are other curves for other capacitors, and I found it a delightful read.
_
There are other curves for other capacitors, and I found it a delightful read.
_
Attachments
Thanks! I did not know that. I have heard, and perhaps read on these forums, that the flaws of electrolytic caps reduce with increasing voltage ratings. That certainly seems to agree with my experience. Will have to test the “increase the capacitance” theory.