Capacitors, especially electrolytics, degrade over time so it is likely a combination of both.
-
WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!
Audio Grade Capacitors?
- Thread starter Fluffy
- Start date
restorer-john
Grand Contributor
A bit of both. Deteriorating capacitors can have varying effects on amplifiers depending on where in the circuit they are and what function they provide.
The only way to determine whether it sounds (or tests) better is to repair the issue causing the failure, test it and then replace/upgrade old components and then test again. No repair operation wanting to actually make money would bother with that.
Glad you have your amp back in service- you've saved another one from landfill and done your bit for the environment.
I, too, am curious about the effect of aging capacitors on performance and sound.
What can be measured and what are the typical sonic symptoms of degraded or bad caps in various parts of audio circuits?
Are some magic audiophile woo caps actually inferior in performance, and as a result able to sometimes create a sonic signature?
From one of SIY's above links:
Is the speed (rate) of charge and discharge of caps variable between types? Does that affect audio performance?
What can be measured and what are the typical sonic symptoms of degraded or bad caps in various parts of audio circuits?
Are some magic audiophile woo caps actually inferior in performance, and as a result able to sometimes create a sonic signature?
From one of SIY's above links:
Is the speed (rate) of charge and discharge of caps variable between types? Does that affect audio performance?
Both capacitance and ESR can be measured with a basic $10 component tester.
Yes, basically the speed of charge/discharge (for a given capacitance) is determined by internal resistance. It can matter in power supplies, but not in the audio path.
What's the expert view on preemptive replacement of power supply caps? Some manufacturers have a nice little earner going with this.
Only replace if clearly not up to spec.
As they age capacitance (charge storage) can decrease while leakage (shunt resistance) and ESR (series resistance) can go up. Used for decoupling in a power supply that can lead to hum and they may eventually leak, short, and destroy themselves and the power supply. Charge storage for sustained high-current peaks is reduced but that is probably not audible. Used as coupling capacitors it varies from almost nothing audible, to higher distortion and changed corner frequencies, to leakage causing "crackling" sounds and upsetting the biasing condition of the stages, to shorting or blowing open (they tend to short) and damaging internal circuits.
Electrolytic capacitors are the most sensitive to time and heat. They are typically delivered with 20% or more higher capacitance than rated (spec) value when new and may be as little as 50% of their rated value at specified end of life. How long that is varies wildly since it depends upon both the capacitor's lifetime spec and the operating environment (high voltage and heat will considerably shorten their lifetime). Film and ceramic capacitors are generally more stable over time and rarely need replacing.
HTH - Don
Electrolytic capacitors are the most sensitive to time and heat. They are typically delivered with 20% or more higher capacitance than rated (spec) value when new and may be as little as 50% of their rated value at specified end of life. How long that is varies wildly since it depends upon both the capacitor's lifetime spec and the operating environment (high voltage and heat will considerably shorten their lifetime). Film and ceramic capacitors are generally more stable over time and rarely need replacing.
HTH - Don
For people without the skills to check, what should they do? Pay someone to check every ten years, some other length of time, or just leave it till something seems off?
Leave it until you have issues (hum, buzz, distortion).
You may never notice "seems off", because it'll happen so gradually. Unfortunately, properly checking capacitor performance means removing them from the circuit, which depending on the equipment can be quite invasive, or as simple as removing some wire connections. It's more likely the former with solid state electronics.
One check you can do yourself is to inspect capacitors and look for bulging or discoloration. If you see anything amiss it's time for a trip to a qualified shop (which probably includes shipping it) or the recycle bin, depending on how the valuable the equipment is and how much you value it. A magnifying glass is useful for this.
Electrolytic capacitor life is highly variable. I've seen 40 year-old equipment that still functions well. I've also seen these caps fail after only a few years on some well-regarded new audio equipment that wasn't abused. Personally, if I had twenty year-old equipment I highly valued and wanted to keep longer term, I'd just have it re-capped at that age. I can't imagine having such an emotional attachment with electronics, except perhaps a reel-to-reel tape deck, but if that's what lights you up be conservative rather than cheap.
With many of them you can actually get a pretty good picture even measuring them in-circuit with a modern ESR tester.
I am imagining at least six scenarios when capacitors might influence the output sound or the operation of the electronic equipment, although not sure if it's easy to prove if upgrading a capacitor really improves the output of the audio signal:
- Input DC-blocking capacitors: usually a cheap MKP cap should do very well its DC-blocking job without any need to upgrade it ever.
- Output DC-blocking capacitors: usually "audio-grade" capacitors, but AFAIK MKP caps are doing this job better, just the price for a 100...1000uF MKP cap would be extreme, so upgrading the output caps with a DC-servo and/or a DC-protection circuit might be cheaper and safer.
- Low-Pass/High-Pass Filters, no matter if passive or active: any regular MKS cap should the job very well, no need for further upgrade.
- Bypassing operational amplifiers: any cheap 0.01...0.1 MKS cap paralleled with a 47...100uF low-ESR electrolytic (or 5...10uF tantalum) should do, unless opamp's manufacturer states a different "recipe". Placing several caps of different values might create post-ringing that might do more damage than good. Spending lot of money of "audiophile" caps here makes no sense, same for 1000uF or higher cap values.
- "Reservoir caps" from the power supplies of amplifiers: any low-ESR electrolytic able to sustain a big current should do. "Upgrading" an existing Nichicon PW with a Nichicon Fine-Gold might not improve anything because the ripple current that Fine-Gold can sustain and combat is lower. Also, doubling or tripling the capacitance of the reservoir caps might destroy the diodes from the bridge rectifier due to the huge in-rush current that the higher capacitance will need when powering ON the device. If the diodes will not break then the PCB traces will or cold-joint might appear on the PCB soldering around the capacitors. Some good reading here: https://sound-au.com/power-supplies.htm.
- Speakers filters: I will let somebody else, with speakers building experience, to explain here pro and cons of the caps. From what I know paper oil caps are preferred, or at least aluminium foil, but I see that there are cheap MKP caps that could do the job very well: http://www.humblehomemadehifi.com/Cap.html.
Last edited:
My first post after just enjoying the reviews and reading...
I recently saw a deal on a new home theater receiver - so i went for it, replacing my 14 year old Denon.
For my main speakers I drive them with a Threshold T200 ( Class A ) amp.
I have had the amp for about 25 years - I use it a lot, it gets very warm when used...
Well with the new receiver I had a hum, I really did not have a noticeable one with the old receiver, but it was also a more expensive unit. Thought I may have had a bad new receiver. So I tried a different amp - no hum..
A couple years ago I replaced the large power supply caps in the amp, no noticeable sound difference but when I switch the amp off the LED stayed on longer, like when I first bought it.
Long intro but now the point of this post...
I replaced the electrolytic caps on each of the boards that drive each channel - and the other two small caps on the power supply board. Problem solved.
Originally they were Elna caps, most (44 of them ) were 470uf/100V - the caps I removed were reading around 380 to 410 uf on my capacitance meter. I replaced them with Nichicon KW series caps ( $1.53 each from Digikey) because I needed them in 16mm diameter. I might have replaced them with a FG series or a Muse but they did not make them in this value/size so... KW it is. I did replace the four 47uf / 50V caps with Fine Gold / FG series caps. They were $.57 each. In researching the older Elna, the 4 were a similar "audio" spec cap.
As someone mentioned above, the Muse is a cool green color.. if I could have easily gotten them I would have bought them.. not because I thought they would dramatically alter the sound but because they were a cool green color...
When I returned the amp to operation I did think it sounded better then it had previously - not better than when I purchased it ( although it is hard to remember exactly what it sounded like 25 years ago ) so the new caps returned it to working as designed, but not better than original.
So my take is this - if you have a old amp you want to restore, or that has been serving you faithfully for 20+ years a capacitor replacement might be in order. The better grade of capacitor is usually more expensive but we are not talking $400 USB cables here.. each cap might be $.10 more so why not go for the slightly more expensive / upgraded caps and get another 20 years out of your investment...
I recently saw a deal on a new home theater receiver - so i went for it, replacing my 14 year old Denon.
For my main speakers I drive them with a Threshold T200 ( Class A ) amp.
I have had the amp for about 25 years - I use it a lot, it gets very warm when used...
Well with the new receiver I had a hum, I really did not have a noticeable one with the old receiver, but it was also a more expensive unit. Thought I may have had a bad new receiver. So I tried a different amp - no hum..
A couple years ago I replaced the large power supply caps in the amp, no noticeable sound difference but when I switch the amp off the LED stayed on longer, like when I first bought it.
Long intro but now the point of this post...
I replaced the electrolytic caps on each of the boards that drive each channel - and the other two small caps on the power supply board. Problem solved.
Originally they were Elna caps, most (44 of them ) were 470uf/100V - the caps I removed were reading around 380 to 410 uf on my capacitance meter. I replaced them with Nichicon KW series caps ( $1.53 each from Digikey) because I needed them in 16mm diameter. I might have replaced them with a FG series or a Muse but they did not make them in this value/size so... KW it is. I did replace the four 47uf / 50V caps with Fine Gold / FG series caps. They were $.57 each. In researching the older Elna, the 4 were a similar "audio" spec cap.
As someone mentioned above, the Muse is a cool green color.. if I could have easily gotten them I would have bought them.. not because I thought they would dramatically alter the sound but because they were a cool green color...
When I returned the amp to operation I did think it sounded better then it had previously - not better than when I purchased it ( although it is hard to remember exactly what it sounded like 25 years ago ) so the new caps returned it to working as designed, but not better than original.
So my take is this - if you have a old amp you want to restore, or that has been serving you faithfully for 20+ years a capacitor replacement might be in order. The better grade of capacitor is usually more expensive but we are not talking $400 USB cables here.. each cap might be $.10 more so why not go for the slightly more expensive / upgraded caps and get another 20 years out of your investment...
Last edited:
Totally agree. However, this being so, it also indicates that replacing new capacitors with better quality new caps can also make a system sound different, even better. Just as a deteriorated cap will measure and sound different in some instance, so too new caps with different construction and quality, will also measure and may also sound different.
Last edited:
No doubt that replacing badly deteriorated caps will make the amp once again measure to factory spec or better. The reality of the amp actually sounding better will be highly dependent on how bad the old caps actually were and how much improvement the new ones made in tech performance.
What is a serious influence here is bias. If you spend the $ and do all the work to get that amp percolating like new, your expectation bias and (near) human brain is going to make you hear a huge improvement in the sound no matter the reality.
Win Win Win.Even just cleaning my 25 yo Adcom's and tweaking the bias and offset made them sound even better than the Benchmark ABH2 amps I heard playing at the Tampa show in February.
YMMV
What is a serious influence here is bias. If you spend the $ and do all the work to get that amp percolating like new, your expectation bias and (near) human brain is going to make you hear a huge improvement in the sound no matter the reality.
How about when I replaced some coupling caps with a different type, hoping to prove it makes no difference, but instead, got a big surprise? The change was quite noticeable. Whether it was an "improvement" only measurement would show (is the response flatter?) or personal preference may dictate.
Happens all the time even if there is no change. Your mind works differently when you attempt to listen carefully for differences. In doing so, it hears more detail, air, etc. These were always there but ignored.
Do the test blind and you will fail to hear the same difference because then, you don't know if something has, or has not changed. Make sure you test 10 times and see if you get 8 right. If it is big surprise, you better get 10 out of 10 right....
Yes, I agree, but my reply was intended to refute that positive expectation bias is always the case. My expectation was for zero change. Instead I not only heard a change, but I liked the change. My expectation, remember, was that it would sound the same. The outcome, probably imagined, was the opposite of my expectations.
btw. I don't really need to do a blind test. I would be completely convinced if measurements showed nothing had changed.
Similar threads
