Help ID: Audio Amplifier Capacitor with strange Bracket marking next to uf rating?

[Dr_PC_Repair]

I am repairing a high end Denon receiver. It appears that a # of capacitors on the amp board are bulging. I ordered replacements I thought matched. But when I looked at them they were still in circuit and I could not see them very well (or measure them well) and also missed that there is a small submark next to the uf which is something I don’t think I’ve seen before. It’s listed like this 100uf(m) and I am not sure if this is some special form of capacitor? (Like maybe for “M=music/audio”?)

The ones I ordered are much smaller seen in photo as RED (despite actually measuring the one on the right before ordering in circuit) Anyways I chose the RED ones (as seen here: 860240675007 Würth Elektronik | Capacitors | DigiKey ) based on lifespan being the longest (since it’s on top of the heatsink they will be exposed to high heat for long periods of time) but now I am worried they won’t give the same acoustic/sound results as the originals because of the (m) and size difference? (Also would going with aluminium polymer be better?)

Anyone have any feedback on this?

Also I’ve included a photo of the amp board on top of the chimney style blower heatsink.
And I wonder why the green ceramic capacitors have a white powder residue on them? (they are like this throughout the entire receiver?)

[Dr_PC_Repair]

I found this info on them: But it still does not explain the (m) these claim they are tailor made for their acoustic characteristics : And so I am not sure the ones I ordered will work/match the original intended audio/sound… again, any help would be greatly appreciated!

And their PDF here: https://www.hificollective.co.uk/sites/default/files/elna_silmic.pdf
(from what I can tell they look like these are 12.5x20mm so 380ma ripple current) vs the RED ones I ordered at 250-500ma does that seem acceptable?

[APDahlen]

Hello Dr_PC_Repair,

Best guess is that the M is a tolerance rating - like the 4th color band on a 4-color resistor.

The datasheet you linked to shows this part numbering system:

The “M” designation indicates a 20% tolerance. This is typical for aluminum electrolytic capacitors.

WRT the Würth Elektronik capacitors, this looks like a perfectly reasonable replacement - one with a 105°C rating. Note that DigiKey does offer a few audio rated capacitors such as the UFW series. However, as you pointed out, this series is only rated to 85°C which may be insufficient your application.

Kindly keep us in the loop as your repair your Denon. Additional pictures are appreciated.

Sincerely,

Aaron

[David_1768]

Hello Dr_PC_Repair,

One thing, is you want to pick a capacitor that is Audio grade, for the best performance.
The closest alternate I found to the size, was 493-10877-1-ND.
This is obsolete, because the manufacture has stopped making it, but we remaining stock.
This is an Audio grade capacitor, 100uf, 80V (max), 20%, 12.5x22, 270ma of Ripple current.
I went back to search for a higher ripple current, and I found 493-15515-ND. This is Audio grade, 100uf, 100V, 20%, 5mm lead spacing, 10x20mm, and 350mA ripple current.
I think this would be the best choice overall. These are “Not For New Design” as they will probably be going obsolete in the future, but we have stock.
Please check the datasheet.

[David_1768]

I would not use the red ones you purchased, as they are not audio grade, and are a lot lower ripple current.

[David_1528]

Personally, I’m skeptical about the actual value of using “Audio” specified caps. The fact that most of the ones in our system seem to be either EOL or obsolete kind of lends to this theory. There may be something to it, but most of what one hears about such things come from the audiophile rags where they seem to hear things based on the cost of the device rather than any objective analysis.

I think that picking parts with lower internal impedance and good ripple current capability, and in the cases where heat is a real issue, higher temperature tolerance, should be the focus.

Assuming you have 5mm lead spacing, I would look at these standard aluminum caps or these aluminum polymer caps. My guess is that the aluminum poly caps will, in general do better.

[Dr_PC_Repair]

Oh you want pics?
Here is how I spotted the blown caps (slight bulge) but also leaked fluids!

Here is what the PCB looked like with the caps removed (luckily the electrolytic fluid that leaked did not short anything!)

Here is the board in the receiver…

I had to remove the entire heatsink to gain access to the solder joints to these capacitors… but the amp IC’s were literally silicone glued to the heatsink (not paste) so they would not come off… I had to remove all the solder connected to the top PCB where the blown caps were.

Finally getting to the bottom:

I hope these failed simply because of hours running at warm temps… if so this repair should be all that’s required. I really hope so especially since this is a high end Denon AVR-5700 which still sells for around $600

Which is a 140w @ 8Ohms 170w @ 6ohms per channel receiver.

[Dr_PC_Repair]

Nice choices, I always tend towards the 5,000 - 10,000 hour rated caps because I don’t want them failing again and needing replacing… to me on a small 5 re-cap job like this, it’s always worth the extra few $.

And yes these are just above the heatsink - although it is a blower style so there are fans pushing the hot air out the back. I still assume the top does warm up.
And judging by these being leaked I assume it was heat over time (because there is no signs of them actually exploding/bursting) there is just what appears to be gradual swelling until the fluids leaked out and the unit failed to produce audio. (Because it did still turn on)

In this photo you can see the PCB with the leaked caps on the right on the heatsink (which is the top of heatsink)

As you see here that PCB is only 1/3 of an inch off the heatsink (and I have no idea if the fan comes on when the unit is powered on or if it only comes on at certain temps - I’d assume this one to avoid fan noise at low volumes)

[Dr_PC_Repair]

As for your comment on audio specific caps. They still seem to sell many on Ebay and AliExpress from Japan/China. (Down side is 2weeks+ wait)
So perhaps it’s just they no longer want to distribute them overseas to retailers making people searching for them buy direct?

Also I imagine it would make much greater difference in higher end audio equipment like this unit I am repairing. (But I have no actual idea in that I do not understand the entire theory of how audio amps work with capacitance and specifically how the ripple currents and ESR etc affect them.)

*As a side comment so many TV’s these days seem to use the 1000h rated caps, as I continue to get TV’s with blown caps typically around or above the heatsinks… like this TV I got yesterday:

As you can see the caps are directly above the heatsink (top is up when installed) notice the ones to the left are fine (since heat rises - tho they could be higher rated hours)

And here they are on an angle to see the bulge (replaced them with the caps I got today from Digikey and TV is fine now)

[APDahlen]

Thank you Dr_PC_Repair,

Appreciate the pictures and educational responses.

Let us know if we can assist in the future.

Sincerely,

Aaron

[David_1528]

Hi Dr_PC_Repair,

Regarding ESR and ripple current, they are inversely related. ESR is “equivalent series resistance”, which essentially is the internal apparent resistance of the cap. So, since resistance impedes current flow, lower ESR generally means higher ripple current is available to the load (which should help with the high current demand of the cannons firing in Tchaikovsky’s “1812 Overture”, or the clean base notes in Pink Floyd’s “Money”).

Also, since current squared times resistance equals power dissipated across the resistance, this means that for a given ripple current, more power will be dissipated as heat inside the cap, making the caps run hotter. I think the theory behind some of the “audio” caps is that if they use special materials inside (silk, for instance), that the sound will be “softer” or “smoother” or such, based on the mechanical vibrations of the materials as current passes through. There may be something to this - I’ve never experimented with it - but having caps run cooler and being able to supply a bunch of current quickly seems like the more important aspects to me.

[Dr_PC_Repair]

Excellent way of explaining this!
Thanks that helped a lot!

[Dr_PC_Repair]

Is there any sort of capacitor socket like there is with IC sockets?

I’d like to do some testing where I swap out capacitors in the amp, but to get to the solder side of the PCB I have to tear apart the entire unit to swap them. It would be much easier if there was a capacitor socket where I could just pop in & out caps to test before I decide which ones sound best and solder them in permanent.

Maybe I could just use this and cut out 2 pins 25-0513-10 Aries Electronics | Connectors, Interconnects | DigiKey

[Kristof_2649]

Possibly a terminal like the following could help with swapping out caps:
4119-0-15-15-47-27-04-0 Mill-Max Manufacturing Corp. | Connectors, Interconnects | DigiKey

[Dr_PC_Repair]

Nice! Thanks

[Dr_PC_Repair]

So the ripple current is not a “leakage” of signal… but rather how much current the capacitor can supply at 120Hrz (their base example).

I was reading up on Silk capacitors and the way it was explaining, made it sound almost as if silk made less “mechanical vibrations” which is what I was was assuming “ripple current” meant. Ripples being like waves of vibrations. There for less ripples (or lower the ripple current) the less “noise/distortion” and there for the more clean and accurate the sound would be…

BUT, the way you’ve explained it is much better because you’ve used “cause and effect” by giving me an example of what the “cause” ESR/Ripple current affects, and now I get it! Thanks!

[Dr_PC_Repair]

I’ve read a few place that state things like this:
For analog audio circuits and audio amplifiers, a lower ripple coefficient is ideal, which is generally required to be below the acceptable level, usually a few percent (such as 0.1%). It is important to note that specific ripple factor standards may vary depending on the criticality of the application, performance requirements and industry standards.
And others claim ripple current adds noise…so in theory you want the lowest ripple current capacitors.

First the wording on this is confusing… are they saying “a lower ripple coefficient is ideal” meaning a lower rated ripple current capacitor is ideal? Because that goes against how you’ve explaining it? And lower ripple current rating = less noise?
Others claim ripple current is only generated when in a AC to DC conversion the AC causes ripples from the 50/60 cycles per second it runs. Which implies then that in a DC load there is no ripple and just find the longest life span caps instead.

[Dr_PC_Repair]

Ok so since I’ve never gotten an EXACT answer only “sceince and math” answers everywhere I’ve asked coupled with conflicting “real world” answers… I went ahead and asked an AI about how ripple current affects audio.
And it’s answered me, the way I had originally assumed that:
Lower ripple current = cleaner audio
Higher ripple current = more artifacts and distortions.
(From what it’s saying below it sounds to me like the lower the rated ripple current the smoother/tighter more true to the original signal it will be because it has tighter more precise power kept from wavering all over - where as higher ripple current is less precise and causes less accurate playback since the signal is fluctuating all over)
But what you’ve said also makes sense with more current available it can reproduce larger sounds easier… so which is it?
Is higher ripple current a rating like “I can provide more current when needed for bright loud notes in music” OR is it “Lower ripple current means less noise gets generated inside the capacitor generated by the power coming in” - this latter one is how I assumed “ripple current” to mean… that lower rated ripple means less “noise leakage” (aka less tolerance to fluctuations… like tuning a radio the lower the ripple current the cleaner the station comes in - the higher the ripple current the more it allows “other noise” in such as static etc)

Anyhow here is what the AI answered to my question about:
“What does ripple current of a capacitor in an audio circuit affect in real world terms *not answered in math or science theory”

AI: In the realm of audio, the ripple current of a capacitor can have several real-world effects. Here’s a non-technical explanation of how it can impact audio quality:

When it comes to audio signals, we want them to be as clean and pure as possible. Ripple current refers to the fluctuating current that flows through a capacitor due to the alternating nature of the audio signal. If a capacitor with a low ripple current rating is used, it means that the capacitor can handle smaller variations in the current more effectively.

Using a capacitor with a low ripple current rating can help minimize the introduction of unwanted noise or distortion into the audio signal. This is because if the capacitor is not able to handle the fluctuating current adequately, it may introduce additional artifacts or irregularities into the audio signal. These artifacts can manifest as audible noise, distortion, or other undesirable effects.

On the other hand, if a capacitor with a high ripple current rating is used, it may struggle to cope with the varying current demands of the audio signal. This can result in poorer audio quality, as the capacitor may introduce more noise or distortion into the signal.

By selecting a capacitor with a low ripple current rating, audio engineers aim to ensure that the audio signal remains as faithful to the original source as possible. This helps to maintain clarity, detail, and overall fidelity in the reproduced sound.

In summary, using a capacitor with a low ripple current rating in audio applications can contribute to a cleaner and more accurate representation of the audio signal, resulting in improved sound quality.

So I am really stumped as I keep getting conflicting answers everywhere I read up on the concept!

[APDahlen]

Hello Dr_PC_Repair,

Not to confuse the issue further, but audio is not one thing. Also, feedback works great to cover many faults right up to the point when it doesn’t.

We need to recognize that capacitors have a response that varies with frequency. For example, the large aluminum electrolytic devices can store large amounts of energy. When short circuited, they can deliver very high currents. However, consider what happens when we “short circuit” the capacitor to a high frequency AC source. We find that the effective capacitance is reduce significantly as the frequency increases.

We know this to be true as aluminum electrolytic capacitors are rarely found in isolation. Instead, they are typically bypasses with a smaller film capacitor located near the associated load. A perfect example is the 0.1 uF bypass capacitor located next to a digital chip. The problem is the same for audio: a large capacitor to store bulk energy with stiff (low reactance) for low audio frequency. The small film capacitors then provide low reactance for the high frequency.

That brings us to feedback.

As a rule, amplifier feedback works really well at low frequencies. However, the open loop gain of the amplifier is reduced at high frequencies. You may recognize this as the op amp gain bandwidth product where amplifier gain drops to units at some high frequency. This high frequency roll off is necessary for amplifier stability with similar circuit in audio and op amps. This high frequency roll of is typically accomplished using a small Miller capacitor (collector to base) in the final voltage amplification stage.

TL;DR This is a system optimization problem with many interactive elements that must work together to provide good sound with guaranteed amplifier stability for all conditions.

Best Wishes,

APDhalen

P.S. For more information about audio and DIY audio projects may I recommend Elliott Sound Products - The Audio Pages (Main Index) (sound-au.com). Another interesting read is PassDiy.

[David_1528]

Sigh. I think that part of the problem is the conflation of current with voltage.

In a power supply, you want the voltage to remain as stable as possible, with as little ripple as possible - with “ripple” being the fluctuation of voltage above or below the set point. The purpose of placing a capacitor between the input and ground and another between the output and ground is to try to keep the voltage ripple magnitude as small as reasonably possible.

If the current going into the supply and the current coming out of the power supply is steady, as in no fluctuations or changes in magnitude over time, then this is trivial, and you wouldn’t even need capacitors. However, in a real-world application, this is never the case.

On the input side, if you are using an AC supply, the voltage will be a sine wave at 50Hz or 60Hz, depending on where you are in the world, and it is passed through a bridge rectifier which converts it into a series of positive-going voltage “humps” at 100Hz or 120Hz (see the red line in the image below). If you place a capacitor on the output of the bridge rectifier, it smooths out the humps to a degree (see the blue line below).

Image from EETimes article:

What the capacitor is doing is absorbing charge (which is the definition of current) during the positive portions of each hump of voltage. The current changes as the voltage changes, and this is “ripple current”. The higher the capacitance, the higher the ripple current, and the smoother those voltage humps will be. The ESR of the capacitor affects how much heat will be generated when that current is drawn in or released. Lower ESR allows higher ripple current without excessive heating within the cap. The diagram below shows the internal real-world model of a capacitor. It consists of a parallel resistance (which is so large that it can usually be ignored) a series inductance (which is important only at higher frequencies) and the ESR, which is relevant at any frequency.

From Wikipedia:

Where ripple current really comes into play is on the output side. As I said earlier, if the load current is steady, there’s not much of a problem. However, for dynamic loads (as in rapidly changing current demands, such as a motor turning on and off, or a loud deep base note in an audio amp circuit), having a capacitor with high ripple current capability allows large amounts of current to flow with less heat build-up, while also maintaining the voltage level better than one with higher ESR. Notice that with the internal series resistance, as high current passes through it, not only does it dissipate power, but there will be a voltage drop across it. This causes distortion in the output voltage.

Therefore, a capacitor with higher ripple current/lower ESR, all other things being equal, should be the preferred product.