To add: both Self and Cordell would say that the higher the bias of the amp, the lower the (crossover) distortion. At the limit you increase the bias until the amp runs in class A which has, generally speaking, lowest crossover. Still not zero. All competent amp designers would say this; it's basic engineering.
So this has nothing to do with the elcap issues we were discussing.
Jan
Hi Jan,
From experience and empirical measurements, I have to disagree slightly with that.
Distortion drops with increasing bias current in many amplifiers up to a certain point, the "happy spot" for that amplifier. Then it remains essentially constant, or you begin to see power supply artifacts in the output as you increase bias current. My takeaway is that beyond a certain level of bias current, anything added is wasted as heat and reduced lifespan. One amplifier's happy spot was 5 mA of bias current (A Symasym built with matched parts).
For things like these, throw away SPICE and measure real products under normal conditions using real test equipment. For you Jan, I know you are very good at testing real amplifiers.
Nope, it hasn't a thing to do with capacitors at all.
From experience and empirical measurements, I have to disagree slightly with that.
Distortion drops with increasing bias current in many amplifiers up to a certain point, the "happy spot" for that amplifier. Then it remains essentially constant, or you begin to see power supply artifacts in the output as you increase bias current. My takeaway is that beyond a certain level of bias current, anything added is wasted as heat and reduced lifespan. One amplifier's happy spot was 5 mA of bias current (A Symasym built with matched parts).
For things like these, throw away SPICE and measure real products under normal conditions using real test equipment. For you Jan, I know you are very good at testing real amplifiers.
Nope, it hasn't a thing to do with capacitors at all.
This fact was the reason for many hard to find faults on Philips devices - mostly cd players.
What made matters worse was that these faults were often only observed temporarily.
A friend of me send me an email with suppliers of top quality electrolytic caps, also in an axial outline - one of them is in Paris:
https://www.ebay.com/sch/i.html?_nkw=sic+safco&_sacat=0&_from=R40&_trksid=p2334524.m570.l1313
https://www.ebay.de/str/fesasparis?_trksid=p4429486.m3561.l170197
for an axial outline e. g.
https://www.ebay.de/itm/155921398135
the genuine web-address of the last ebay supplier is here
https://fecomponents.shop/capacitors/?q=Sic+Safco
Thank you for this advice. I'll consider the film caps for the low value electrolytics. I've seen posts on this forum by @jean-paul suggesting the same - can't remember where tho.
A couple of days ago I have ordered Vishay (BC Components) caps in the same form factor (axial electrolytic and radial electrolytic) as the originals, indeed upping the voltage to at least 40V up to 63V. The axials are even blue, so that's pretty great for visuals at least. All identical values though. As BC Components was formerly Philips owned, this choice gives me the idea that there is some sense of continuity there - probably wholly imaginary at this point.
@anatech: please help me out - I'm not a native English speaker and my OED has no entry for "gribblets". What are they? Top to bottom connectors? (Easily misread for "giblets" and then making no sense whatsoever - see another recent thread here where you advised to resolder them.)
I have a question that maybe are anwered before and is a little stupid.
Lats say a modern low ESR 50V 220uF cap for input and output DC protection
Can the 1 A current capacity destroy components conected to it? Lets say Op amp or 100 ohm output resistors.
Take 20 Volt DC that is suddenly set to 0 on input or utput or a case where the cap gets charged by the cirquit because of DC in cirquit.
Lats say a modern low ESR 50V 220uF cap for input and output DC protection
Can the 1 A current capacity destroy components conected to it? Lets say Op amp or 100 ohm output resistors.
Take 20 Volt DC that is suddenly set to 0 on input or utput or a case where the cap gets charged by the cirquit because of DC in cirquit.
Hi kunlun121,
As you correctly guessed, they are top to bottom PCB connections. Normally a better PCB substrate is used that allows for plated through holes. Heck, Sony did this with phenolic PCBs that were single sided. A real pain! Philips used a light copper or brass rivet (I have never looked closely) that was crimped into place and wave soldered. These didn't heat up as quickly, and probably surface prep didn't allow them to wet properly. The end result is many connection that would go resistive or open ranging from a year to several. You have to remove the old solder to access the surface of the rivet, and heat it up sufficiently to allow the solder to wet. Since they have oxides formed by now, liquid solder flux is highly recommended. Non-corrosive designed for electronics work. You may get air bubbling up, but a little solder with the edges smooth on both sides means a good connection. They also used some component leads for bottom to top connections and I have found some of those bad as well.
Capacitors. Replace like with like. Film capacitors tend to be a little larger and they don't buy you anything. Axial capacitors are being discontinued these days, better order some for future.
Translators aren't prefect as we all know.
As you correctly guessed, they are top to bottom PCB connections. Normally a better PCB substrate is used that allows for plated through holes. Heck, Sony did this with phenolic PCBs that were single sided. A real pain! Philips used a light copper or brass rivet (I have never looked closely) that was crimped into place and wave soldered. These didn't heat up as quickly, and probably surface prep didn't allow them to wet properly. The end result is many connection that would go resistive or open ranging from a year to several. You have to remove the old solder to access the surface of the rivet, and heat it up sufficiently to allow the solder to wet. Since they have oxides formed by now, liquid solder flux is highly recommended. Non-corrosive designed for electronics work. You may get air bubbling up, but a little solder with the edges smooth on both sides means a good connection. They also used some component leads for bottom to top connections and I have found some of those bad as well.
Capacitors. Replace like with like. Film capacitors tend to be a little larger and they don't buy you anything. Axial capacitors are being discontinued these days, better order some for future.
With a 220 uF coupling cap this seems unlikely. A 100 Ohm resistor and 20V still is a maximum current of 200 mA worst case. In reality other (current limiting) resistance and active current limiting in the opamp are involved combined with the very short time and not full fail voltage occuring.
Last edited:
Thank you for going into the details of this. Are they easy to identify? They haven’t stood out when previously I inspected the boards for bad solder joints. Are they fully covered in solder or does the solder job typically still show the rivets? Will try to identify some and post a picture.
Depending on the side of the PCB you're looking at, it's a metallic circle, or a four petal "flower" on a circular pad. Once you see one, you can pick them up easier. There are many of these things, look closely to get them all. They are all the colour of solder.
Sikorel 125
The 125° is for max temp apparently.
I was looking for something exceptionally reliable for PSU's recently and discovered these,
quite an impressive set of characteristics.
Those look like anti resonant cans, like the Vishay 021ASM.
TDK have a 'line' labelled B40600 series on this Sikorel 125 too.
Found Kemet PEG124, but haven't played with these yet.
Kemet PHH223
The 125° is for max temp apparently.
I was looking for something exceptionally reliable for PSU's recently and discovered these,
quite an impressive set of characteristics.
Those look like anti resonant cans, like the Vishay 021ASM.
TDK have a 'line' labelled B40600 series on this Sikorel 125 too.
Found Kemet PEG124, but haven't played with these yet.
Kemet PHH223
Last edited:
What parameters are those that give it 'exceptional reliability'?
Is it just the temp spec or is there more?
I'd be really interesting in those, it's an important metric for me too.
Jan
Is it just the temp spec or is there more?
I'd be really interesting in those, it's an important metric for me too.
Jan
The showed Sikorel 125 (B41684) are according several old guys from service-centers for industrial stuff one of the best available elcaps - even still in good condition after a long time (over 10 years) of operating.
Important it is a reforming process after a long time without operating (in general with NOS versions).
Often cheap offers are to find, e. g.
https://www.ebay.de/itm/235972140053
Also S+M Elcaps (Siemens+Matsushita) from the 90s - go to
https://eshop.eca.ir/خازن-الکترولیت...-الکترولیتی-صوتی-150uf-385v-مارک-siemens.html
had an excellent reputation among technicians.
It would be interesting to know where currently manufactured versions are tested for quality in terms of reliability and service life expectancy.
Companies that manufacture electronic assemblies for aerospace applications need to be fully aware of this.
Jan...
Not having a bench full of test gear means metrics would not be something I can communicate effectively.
Having been at the front end of Rock and Roll, by god does one learn about endurance and how equipment bears up or doesn't.
True Hifi has always been of interest as it's the other end of music.
The importance of good power supplies cannot be overstated, many folks fascination of battery powered hifi is a valid one.
Off the top of my head as a passing thought from my experiences :
High ripple current handling is one. Which flags up Polymer caps and their various hybrid developments, but size restrictions are what, 1000µF and rising?
Another basic one is: Larger cans ( not related to a recent humorous post ) seem to last longer ( non etched folis) they're not trying to squeeze everything in to the smallest package possible.
Plate connection: Those T-power BHC Aerovox ALN20S1108DF caps have impressive speed of discharge, thats more performance related. But the quality and quantity of connection to the plate is important, very few manufacturers share this, unless its a selling point like some mentioned post 3190.
Anti vibration design. In audio, what device is not exposed to this?
Self healing, more a tantalum thing
And of course higher voltage handling, but you dont always want that.
The majority of electrolytics last around 3000 hours, if they last longer its because they're either using different materials and/or some of the above points.
There's the TDK/Epcos cap B41505.
This was something pulled for me to look at later, it might as well be put here:
Not having a bench full of test gear means metrics would not be something I can communicate effectively.
Having been at the front end of Rock and Roll, by god does one learn about endurance and how equipment bears up or doesn't.
True Hifi has always been of interest as it's the other end of music.
The importance of good power supplies cannot be overstated, many folks fascination of battery powered hifi is a valid one.
Off the top of my head as a passing thought from my experiences :
High ripple current handling is one. Which flags up Polymer caps and their various hybrid developments, but size restrictions are what, 1000µF and rising?
Another basic one is: Larger cans ( not related to a recent humorous post ) seem to last longer ( non etched folis) they're not trying to squeeze everything in to the smallest package possible.
Plate connection: Those T-power BHC Aerovox ALN20S1108DF caps have impressive speed of discharge, thats more performance related. But the quality and quantity of connection to the plate is important, very few manufacturers share this, unless its a selling point like some mentioned post 3190.
Anti vibration design. In audio, what device is not exposed to this?
Self healing, more a tantalum thing
And of course higher voltage handling, but you dont always want that.
The majority of electrolytics last around 3000 hours, if they last longer its because they're either using different materials and/or some of the above points.
There's the TDK/Epcos cap B41505.
This was something pulled for me to look at later, it might as well be put here:
Last edited:
Special electrolytic caps for heavy industrial/high reliability purposes are produced specifically for that. The misconception is that these are equally suitable for use in the extreme field of home audio. And that using such parts will improve sonical results. Unfortunately this is an assumption not based on facts.
As I had access to such parts I of course tried things out. The parts I used to test this only rarely were good performers in audio. Some series by Sic Safco were the exception.
What lives longest in harsh circumstances just does not necessarily sounds best. What sounds best also often does not need to live longest 🙂
As I had access to such parts I of course tried things out. The parts I used to test this only rarely were good performers in audio. Some series by Sic Safco were the exception.
What lives longest in harsh circumstances just does not necessarily sounds best. What sounds best also often does not need to live longest 🙂
Last edited: