Hi,
last night I was looking at some new electrolytic capacitors for the power supply of some of my amps and I noticed how old e-caps showed small differencies in impedance between 100 Hz and 10 kHz whilst new ones, due I guess to the egemony of SMPS, have highly improved performances at high frequencies.
So I was thinking how to restore this small difference between low and high frequencies, and first thing that came into my mind are those small cheap inductors seen in battery chargers.
After a short calculation, I saw that in the range of 10-30 uH we can restore that ratio of impedances of the cap. Other point is current. I would consider around 2 Arms to be on the safe side.
While ordering some stuff from mouser, I would add some small cheap inductors (we are talking aroun 20 cents each) to perform some tests.
Has anyone experience in anything similar?
Thanks in advance
Roberto
last night I was looking at some new electrolytic capacitors for the power supply of some of my amps and I noticed how old e-caps showed small differencies in impedance between 100 Hz and 10 kHz whilst new ones, due I guess to the egemony of SMPS, have highly improved performances at high frequencies.
So I was thinking how to restore this small difference between low and high frequencies, and first thing that came into my mind are those small cheap inductors seen in battery chargers.
After a short calculation, I saw that in the range of 10-30 uH we can restore that ratio of impedances of the cap. Other point is current. I would consider around 2 Arms to be on the safe side.
While ordering some stuff from mouser, I would add some small cheap inductors (we are talking aroun 20 cents each) to perform some tests.
Has anyone experience in anything similar?
Thanks in advance
Roberto
If you want old saggy capacitors, fit a resistor in series with them to increase the ESR.
Inductors are frequency dependant and will give odd results.
Inductors are frequency dependant and will give odd results.
Hi JosnSnell,
So you suggestion is to add a fixed resistor (some hundreds of mOhm) to get the same ratio by increasing the whole ESR instead of increasing it with the frequency with a small inductor. I will try both ways.
Thanks
So you suggestion is to add a fixed resistor (some hundreds of mOhm) to get the same ratio by increasing the whole ESR instead of increasing it with the frequency with a small inductor. I will try both ways.
Thanks
This MUST be tube stuff. Why would you make performance worse? A tube device will also perform better with better power supply caps.
Only section where this kind of things occur. Next thing is to sand off the gold plating of brand new Elma switches contacts to have them behave like the older scratching switches.
And cotton isolated mains cabling of course.
And cotton isolated mains cabling of course.
Thank you for the capital letters jean-paul, I really appreciate you teutonic MUST.
I'm exactly talking about tube stuff, so about designs made with components with different specs compared to modern ones. Why are you so sure this attempt will make the performance worse?
I'm exactly talking about tube stuff, so about designs made with components with different specs compared to modern ones. Why are you so sure this attempt will make the performance worse?
Dear Jean Paul,
you are free to not agree, you are free to think it's useless, but as a moderator you cannot joke users. Thanks.
you are free to not agree, you are free to think it's useless, but as a moderator you cannot joke users. Thanks.
Yes I tend to think I am normal but some might disagree 🙂
Mostly because better parts are simply better parts and since most want to squeeze the maximum out of their tubes it would seem logic that using better caps will give that possibility. The attempt with coils will make performance of the caps worse than their specifications and it will then not improve a thing using these. Maybe one could then use NOS NIB caps or just leave the old ones to enjoy more bad performance 😀 Resistors would be more adequate but why?
In general the thought is the more parts reach the specs of ideal parts the better the PSU will perform with regards to ripple, HF/RF behavior, output impedance etc. Our granddads also knew this but they had to do with what was possible then and they probably would envy the quality of todays caps, semiconductors etc. Many a trick was used to enhance performance to overcome non idealities of parts available then. They also wanted maximum performance maybe it was more a goal then it is today. They invented to term HiFi when devices were starting to perform live like. The old guys I know were very happy when they could get rid of the tar caps that were quite bad.
Sometimes it is different when renovating old radios where the looks of the old parts are important (some even use the old caps casings and solder new ones in the old casings). In case of maximum performance in 2021 with polluted mains voltage having caps perform better in a broad frequency band will also filter that pollution better. When the old amplifiers were designed mains voltage and the aether were way more clean than they are today. For this very reason one can not get away anymore with wooden casings and no shielding....
So when wanting to have old performance one better uses a time machine to go back in time. The teutonic MUST was right as there is no other section where one would expect this.
Mostly because better parts are simply better parts and since most want to squeeze the maximum out of their tubes it would seem logic that using better caps will give that possibility. The attempt with coils will make performance of the caps worse than their specifications and it will then not improve a thing using these. Maybe one could then use NOS NIB caps or just leave the old ones to enjoy more bad performance 😀 Resistors would be more adequate but why?
In general the thought is the more parts reach the specs of ideal parts the better the PSU will perform with regards to ripple, HF/RF behavior, output impedance etc. Our granddads also knew this but they had to do with what was possible then and they probably would envy the quality of todays caps, semiconductors etc. Many a trick was used to enhance performance to overcome non idealities of parts available then. They also wanted maximum performance maybe it was more a goal then it is today. They invented to term HiFi when devices were starting to perform live like. The old guys I know were very happy when they could get rid of the tar caps that were quite bad.
Sometimes it is different when renovating old radios where the looks of the old parts are important (some even use the old caps casings and solder new ones in the old casings). In case of maximum performance in 2021 with polluted mains voltage having caps perform better in a broad frequency band will also filter that pollution better. When the old amplifiers were designed mains voltage and the aether were way more clean than they are today. For this very reason one can not get away anymore with wooden casings and no shielding....
So when wanting to have old performance one better uses a time machine to go back in time. The teutonic MUST was right as there is no other section where one would expect this.
Last edited:
Thank you guys. And I'm not jocking thanking you.
Not because you answered my question, but because you underlined me some aspects that, after two decades of moderation in other forums it may have been happened that I reacted the same way without noticing: how key words made us react with preconcepts, and how this influences the perception of other people.
"Vintage", as can be easily understandable by jean-paul's first intervention, readable between the lined of Mooly's post and clearly underlined by bucks bunny's post is clearly one of those, as it forwards us to all the snake oil seen undreds of times at thousands bucks and all the frustration we felt in seeing people convincing others to have invented the chronovisor.
Another aspect is the request to "reinvent the weel", that in this case is a vintage octagon used as a weel chiseled starting from a round weel.
The "newbie" aspect, together with the aggregation (Le Bon's "Psychology of crowds" applies to the web too) to older users are certainly important factor of the recipe too, and often make people believe to be justified to act out of netiquette, or simply out of general etiquette rules that in real life would have been automatically applied.
That said, let me explain the etiology of this thread.
Substituing old caps for friends, I often heard them say "it isn't softer as it was", even in instrument amplification the most common comment after a recap is "it isn't as creamy (you know, we are Italians, we need to refer to food) as before".
That made me think about a possible solutions to substitute old caps with new ones, trying to join the safety and long-life of new caps together with the "worst" characteristics of the old ones that people became used to during the years.
That's it.
Cheers.
Not because you answered my question, but because you underlined me some aspects that, after two decades of moderation in other forums it may have been happened that I reacted the same way without noticing: how key words made us react with preconcepts, and how this influences the perception of other people.
"Vintage", as can be easily understandable by jean-paul's first intervention, readable between the lined of Mooly's post and clearly underlined by bucks bunny's post is clearly one of those, as it forwards us to all the snake oil seen undreds of times at thousands bucks and all the frustration we felt in seeing people convincing others to have invented the chronovisor.
Another aspect is the request to "reinvent the weel", that in this case is a vintage octagon used as a weel chiseled starting from a round weel.
The "newbie" aspect, together with the aggregation (Le Bon's "Psychology of crowds" applies to the web too) to older users are certainly important factor of the recipe too, and often make people believe to be justified to act out of netiquette, or simply out of general etiquette rules that in real life would have been automatically applied.
That said, let me explain the etiology of this thread.
Substituing old caps for friends, I often heard them say "it isn't softer as it was", even in instrument amplification the most common comment after a recap is "it isn't as creamy (you know, we are Italians, we need to refer to food) as before".
That made me think about a possible solutions to substitute old caps with new ones, trying to join the safety and long-life of new caps together with the "worst" characteristics of the old ones that people became used to during the years.
That's it.
Cheers.
In my experience of changing dried out/defective filter caps in good quality and sometimes even excellent 80ties amplifiers the higher ESR and dried out electrolyte make the amplifiers sound different than they did when they were new. This was of course a slow process and no one notices until the amplifier starts behaving strangely or even starts humming. A bit like some only knowing to add engine oil when the red light starts to flicker on the dash board. When replaced for brand new very good ones it was striking that people rarely were satisfied at once as they simply got used to the deteriorated old caps having impact on the overall sound quality very often described as "more warm". As a tech I measured this to be related to sagging rail voltages.
Then after a few days most if not all are very satisfied. I tried to make it a psychological story but decided that they simply have to get used to it and tell them so. Personally I am in this stuff long enough to know that new parts like caps need some time as well (I will be shot for this remark) and I am not afraid to also mention this. "Just leave it playing for a few days and give it some time to settle". Interested in this phenomenon I always ask later if they are now satisfied and in practically all cases I get a positive response that bass has improved, it was worth the money, the device became better after a few days etc. So, if I had to rely on laymen's remarks and wishes combined with direct results I would never have satisfied audio friends 😀
* adding coils in series with perfectly good caps makes it a frequency dependent affair. It would be interesting to know (like already described) how the original caps were behaving when they were brand new. AFAIK these did not have coils in series and the old guys would have a laugh if suggested.
Then after a few days most if not all are very satisfied. I tried to make it a psychological story but decided that they simply have to get used to it and tell them so. Personally I am in this stuff long enough to know that new parts like caps need some time as well (I will be shot for this remark) and I am not afraid to also mention this. "Just leave it playing for a few days and give it some time to settle". Interested in this phenomenon I always ask later if they are now satisfied and in practically all cases I get a positive response that bass has improved, it was worth the money, the device became better after a few days etc. So, if I had to rely on laymen's remarks and wishes combined with direct results I would never have satisfied audio friends 😀
* adding coils in series with perfectly good caps makes it a frequency dependent affair. It would be interesting to know (like already described) how the original caps were behaving when they were brand new. AFAIK these did not have coils in series and the old guys would have a laugh if suggested.
Last edited:
I can't for the life of me see why you would intentionally hinder the performance of a modern quality part on an amplifier meant to reproduce audio. Maybe for guitar or instrument amplification I can see, but if that's the case just buy cheaper capacitors with higher ESL and ESR.
Just get a large can electrolytic. The bigger the more inductive it is, the parasitic inductance is a very large part of the geometry of the part, so a big screw terminal electrolytic will have worse high frequency performance than a small SMD electrolytic even tho they have the same capacitance.
Adding an external bead is idiotic beyond... the parasitic inductance is lumped along the entire capacitor's foils all along the foil plates. Having a discrete bead somewhere on the outside of the cap will just give all kinds of possibilities, including high frequency resosnances or weird behaviour due to currents etc, the inductance is not linear with current, temp, etc...
Adding an external bead is idiotic beyond... the parasitic inductance is lumped along the entire capacitor's foils all along the foil plates. Having a discrete bead somewhere on the outside of the cap will just give all kinds of possibilities, including high frequency resosnances or weird behaviour due to currents etc, the inductance is not linear with current, temp, etc...
Thanks SemperFi(delis), joining your suggestion and JosnSnell's one, a small resistor of few hundreds of mOhm would be easier or even easily switchable.
Add the death cap back in too 😀
The Mrs has a little 1W transistor Roberts radio with WiFi and audio streaming etc.. gets used a lot and has that 'background' thin sound that I equate to old tube radios..
Hi NickKUK, I am intrigued... what is the death cap?
@SemperFi: large can caps often have even lower ESR and higher ripple current than normal radial ones... In fact these are mostly to be preferred when one is in this for optimal performance. The value will likely also be larger than the original one making ripple voltage lower and performance better which apparently is not wanted.
@zintolo: for desired degraded performance I suggest to have large can caps in series to up the ESR, making it possible to lower the value to the equal value of the old caps (modern caps have often higher capacity value in a smaller casing) and such of course with voltage balancing parallel resistor over each cap as good design practice. It would be advisable to have highest possible voltage ratings as these will be lower µF value. One could measure the ripple voltage and try to make the new PSU having as much ripple voltage as the old PSU.
Rap on Replacing Electrolytic Capacitors
@SemperFi: large can caps often have even lower ESR and higher ripple current than normal radial ones... In fact these are mostly to be preferred when one is in this for optimal performance. The value will likely also be larger than the original one making ripple voltage lower and performance better which apparently is not wanted.
@zintolo: for desired degraded performance I suggest to have large can caps in series to up the ESR, making it possible to lower the value to the equal value of the old caps (modern caps have often higher capacity value in a smaller casing) and such of course with voltage balancing parallel resistor over each cap as good design practice. It would be advisable to have highest possible voltage ratings as these will be lower µF value. One could measure the ripple voltage and try to make the new PSU having as much ripple voltage as the old PSU.
Rap on Replacing Electrolytic Capacitors
Last edited:
I don't understand the logic to this at all, personally, i always fit the best quality caps i can into my tube amps from either Panasonic or Nichicon and they are all the better for it.
If you want crappy caps just buy some cheap rubbish from China with a brand name you have never heard of, plenty of them on eBay or AliExpress.
If you want crappy caps just buy some cheap rubbish from China with a brand name you have never heard of, plenty of them on eBay or AliExpress.