Buying filtering cap.

[bob p]

I think the CDE app guides make for a good read:

Application Guides | Cornell Dubilier Electronics, Inc.

Here is the one for electrolytics. Lots of good info here:

http://www.cde.com/catalogs/AEappGUIDE.pdf

[Alan0354]

For any and all interested there is an extensive presentation about Aluminum Electrolytic Capacitors at http://ww.chemi-con.com/chemi-con.co...ionmay2012.pdf .
It includes a lot of information about actual expected life based on the many factors that affect that life.
There are other sources for this information of course but the one I referenced is apparently a presentation format that gives a lot of information without the need to read a lot of detailed text. The downside is that you need to have some previous background knowledge to understand some of the information.
Cheers,
Tom

I read like half, I learn those before, it's really the loss tangent. All in all, I was going to respond back here. It is very common for wide band application that you parallel a small cap with the big cap to eliminate the ESR problem. OR using two of the half value cap in parallel to lower the parasitic series resistance!!! I am not going to pay $10 for a cap if I can use two $1 cap in parallel to get the same result!!!

[Alan0354]

That's 2000 hours at 105C. At lower temps they last much, much longer. I recently was looking at MIEC caps for ripple current ratings (Full River Cap co in China) and felt they were an excellent choice taking into account voltage ratings, ripple current, life and price.

I have been looking up MIEC, they have 600V 22uF and even 100uF. Dissipation factor is 0.24 and 0.2 respectively. I know 600V is good. I don't need to stack two of the 350V 100uF for the power amp stage. That got to be better.

What kind of dissipation factor I should look for other than 0?

[nickb]

That's hilarious.
2000 hours is NOT long life. But for "Illinois" i suppose it is ! hahahahahahahahahahahahahhahahahahahahhaah!

10,000 hours IS long life. 105C IS GOOD! high ripple current is GOOD! Low ESR is GOOD!

KNOW the difference between cheap and GOOD.

I buy the best ones. They only cost a little more. And worth every penny.

No, I wasn't thinking about traditional "long "life caps". I'm was referring to life at the actual working temperature. The fact is that most modern caps are very good.

[nickb]

I have been looking up MIEC, they have 600V 22uF and even 100uF. Dissipation factor is 0.24 and 0.2 respectively. I know 600V is good. I don't need to stack two of the 350V 100uF for the power amp stage. That got to be better. How's the dissipation factor sounds?

Dissipation factor only has meaning at the measurement frequency, as do most parameters, but especially DF. I find ESR easier to visualize. However, these are seldom given over a frequency range in the data sheets for most electrolytic types. I know I had a tough time in getting the ripple data for the MIEC caps despite FRC being as helpful as they could. For bypass caps at audio frequencies the ESR defines the limit of attenuation at higher frequencies. The attenuation also depends on the load and the series impedance of the feeding circuit.

Be a little cautious on current since for power supplies it's ripple current that matters and that can be much higher than the output current due to the horrific waveform distortion of the rectifiers. Suggested rules of thumb for ripple current = DC x 3.5 for full wave and DC x 6.5 for half wave.

The things that will really reduce the life of you caps are
(1) Excess voltage - consider the surge you can get before the tube warm up.
(2) Excess ripple current - this will cook the caps from the inside.
(3) Excess heat - locate them in cool place

Oh, and DF of 0.25 is fine.

[Alan0354]

Thanks, I still cannot rap my head around about these filter caps can affect sound. To me, look at the hv on the scope, if there is no noise voltage, just droop went hit the guitar hard, it's all good. Do you know anything about it?

[Enzo]

Start there. They sag. Do all caps sag the same in a given circuit? Maybe one type recovers from transients differently from another. After the first node, the majority of filter cap function is decoupling. And if freq response differs at all, it can have a selective effect on decoupling, and that might not be enough to cause trouble, but could affect response of the stage. We play amps hard. Those caps spend their audio time partly sagged as a rule, and peaks might expose ripple in some cap differently from another, and that affects the sound.

Will you or I hear any difference between IC and F&T caps in the power supply of some Pro reverb? Maybe yes, maybe no. Depends on the caps, the amp, and what we are listening to/for. But even if you hear nothing, I don;t think it is a good idea to decide a priori that no effect can be had. because all circuits are not alike.

There are causes and effects all around us. Someone might ask why we should care if an audio system can reproduce ultrasonics - we can;t hear them after all. But ultrasonics are part of our natural environment, and they also make sum and difference artifacts we maybe CAN hear. That means they can affect our sound. Cap artifacts we cannot readily measure in our little shops can similarly have the potential to affect what we hear leaving an amp.

[Alan0354]

I agree that the effectiveness of the electrolytic cap depend on the ESR. The impedance of the cap drop as the frequency rise to the minimum and then rise as the loss take over. The impedance curve vs frequency is like a parabola that reach a minimum and rise. so yes, it loss it's effectiveness as a filter cap ( decoupling as you said) at high frequency.

I just review loss tangent( Dissipating factor) and ESR. It is frequency related. As I posted, that can be fixed very easily by paralleling a small value cap like 0.1uF ceramic cap that take over the impedance curve at high frequency. This practice has be common in the modern electronics where frequency extends to GHz.

So the question is back to whether you need to buy expensive filter caps or just cheap caps with parallel small caps. Those German caps can be like 5 times the cost!!!!

[Enzo]

I just looked at AES (Antique Electronic Supply) and the common 22uf 500v axial is $3.50 for Illinois Capacitor, and $4.45 for F&T. That is about a 30% premium. If you have to pay 5 times cost, find a different store. Of course, I don;t know where you are.

[J M Fahey]

If Leo was still alive, I doubt that he would be using Chinese Capacitors.

If he were alive *today* I'm *sure* he would use Chinese capacitors (and tons of other Chinese parts), Russian or Chinese tubes, etc.
If not actually having a factory in China,that is

FWIW he always used generic widely available reasonable price parts.
That meant USA made parts .... up to the Seventies, ......... but definitely not any more today.

FWIW he *did* use cheapest available workforce, which way back then meant Mexican or Mexican descent, just look at all those Lupe, Carlos and Dolores signed assembly check labels glued everywhere inside Fender chassis and boards.
Today it's either Mexican on the other side of the border or straight going to China.

Remember the ton of Japanese Fenders from the late 70's and along 80's?
Very reliable by the way, most are still alive or easily repairable.

They also sold tons of very good "Fender" guitars .... made by FujiGen Gakki, Tōkai and Dyna Gakki.

So "good value for a low production cost" really has been a constant along 60 years of Fender Company life.

So yes, he would definitely use Chinese today .

He never was a "boutique" amp maker, far from it, but a "working musician" amp maker.

[Alan0354]

If he were alive *today* I'm *sure* he would use Chinese capacitors (and tons of other Chinese parts), Russian or Chinese tubes, etc.
If not actually having a factory in China,that is

FWIW he always used generic widely available reasonable price parts.
That meant USA made parts .... up to the Seventies, ......... but definitely not any more today.

FWIW he *did* use cheapest available workforce, which way back then meant Mexican or Mexican descent, just look at all those Lupe, Carlos and Dolores signed assembly check labels glued everywhere inside Fender chassis and boards.
Today it's either Mexican on the other side of the border or straight going to China.

Remember the ton of Japanese Fenders from the late 70's and along 80's?
Very reliable by the way, most are still alive or easily repairable.

They also sold tons of very good "Fender" guitars .... made by FujiGen Gakki, Tōkai and Dyna Gakki.

So "good value for a low production cost" really has been a constant along 60 years of Fender Company life.

So yes, he would definitely use Chinese today .

He never was a "boutique" amp maker, far from it, but a "working musician" amp maker.

+1000.

F T cap Enzo showed is good price though. BUT, at this point of time, I won't exactly choose German cap by default. These kind of things change, if I can find Japanese or ever Korea made, I think I'll take the chance with them. If you look at reliability of BMW, Volkswagen, Audi and some Mercedes, they are not confidence building. The world moved on.

BTW, my 73 Pro Reverb stamped "Castillo"!!!

I personally have no issue of American or even Japanese brand that made in China. As long as quality control by Japan or US, I think it's all good.

[Alan0354]

I just looked at AES (Antique Electronic Supply) and the common 22uf 500v axial is $3.50 for Illinois Capacitor, and $4.45 for F&T. That is about a 30% premium. If you have to pay 5 times cost, find a different store. Of course, I don;t know where you are.

Thanks for this, $4.45 is definitely reasonable. I can put it on my list of possibilities. Ebay has them much higher.

[Enzo]

Looking further, try Tube depot, I see both F&T and IC both selling for the same $3.95, your choice.

https://tubedepot.com/t/diy-central/...s/electrolytic

[Mick Bailey]

Illinois caps are fitted to quite a few Fender amps and often they're the only reason for the amp needing attention. But, those amps have been used for a number of years before the caps fail. I doubt there is a cap on the market that will outlast the silver-can Mallory's from the Silverface era, though.

I buy mainly F&Ts, but to me the thin leads are the weak point.

[nickb]

So the question is back to whether you need to buy expensive filter caps or just cheap caps with parallel small caps. Those German caps can be like 5 times the cost!!!!

Yes, adding a parallel HF cap flattens the Z over a wider frequency range. That matters if that frequency range is of interest to you. That is the case in digital circuits and switching power supplies for example, but in an analog audio tube amplifier the band of interest is usually less that 20kHz or so. Thus, it's unlikely to be of benefit.

Also, adding a cap in parallel doesn't turn an unreliable cap into a reliable one, nor indeed does it make an inexpensive one into an expensive one. The price you pay has more to do with the kudos attached to the part than the actual performance. The price reflects what the market will accept.

Also the choice is often driven by mechanical size, availability and sometimes customer demand.