Solen Lytics

[Leo_Gnardo]

Film caps in cathode bypass duty up to 2.2uF at 5% tolerance or better as this is where all important low end shelving equalization is set. -20%+50% on electrolytics in these positions can make or break low to low mid response.
Also, large variations in cathode bypass caps between two otherwise almost identical amps will result in distinct sonic differences between them.

Right on!

As for polyprop capacitors in ripple smoothing duty (B+ filtering by another name). You guys are opening a can of worms.
Has anyone asked a question why WIMA, EPCOS etc high capacitance high voltage caps are radically more expensive than Solens and some other audiophile" grade caps?
One reason is to be found in what these are called. DC-link capacitors, developed specifically for ripple filtering duty. Solens are not since they don't say so explicitly.

Solen 22uF/630V---ca 12$CAD
EPCOS 20uF/630V--ca 15$US

The difference? EPCOS specify which application these are made for, Solen doesn't. Check the Epcos data sheet, then try to find similar info on Solens.

http://www.epcos.com/inf/20/20/db/fc_20 ... 74_678.pdf
https://www.solen.ca/pdf/solen/fastcap.pdf

About this, I'm confused. I've found Solen's work perfectly well for power supply duty - ripple filtering. Haven't tried the EPCOS.

So, about specmanship. 2 milliohms or 0.5 milliohm in ESR specs.
Right. Cool. What is the resistance of 1" of the lead? and then another inch or 3 of connecting wire to the rectifier?
Makes one wonder why EPCOS package theirs in boxes with short 0.2" pins.

So, "speed" and transient response. How fast a capacitor charges up OR DISCHARGES depends on capacitance, ESR and the resistive part of driving source impedance and its current sourcing % sinking capacity.
In tube amps and their linear power supplies all this is lost in wiring resistances and inductances. Only advantage is the extremely long lifetime. But hey, you can achieve close to that by correctly degrading high quality electrolytics.
And always remember, there's this "wishful hearing" phenomenon will all fall victim to.

Thanks VV "the voice of reason."

[talbany]

So, "speed" and transient response. How fast a capacitor charges up OR DISCHARGES depends on capacitance, ESR and the resistive part of driving source impedance and its current sourcing % sinking capacity.
In tube amps and their linear power supplies all this is lost in wiring resistances and inductances. Only advantage is the extremely long lifetime. But hey, you can achieve close to that by correctly degrading high quality electrolytics.
And always remember, there's this "wishful hearing" phenomenon will all fall victim to.

When measuring caps you want to look for ESR, Q and Dissipation. Your typical PS series and Mallory's will have a Q of around 250 300 range.In a polyprop cap it will be in the 4k to 6k range..Big difference..This is IMO why these caps in the signal path can sound brighter and edgier and especially true with metalized poly prop...In the power supply it's the lack of ESR really kills the feel, bloom and organic qualities.. If a person wants to experiment with PP caps, I would suggest a 1 to 10 ohm resistor in the ground leg to simulate a little ESR, (why I said this cap is too good to be used in the power supply of a guitar amp!)
As far as cathode bypass go? if you want to try and smooth out say an OD stage in a guitar amp Solens or Daytons might be something you might like there..Personally I have a hard time telling the difference between a PP and a polyester there..YMMV

Tony

[VacuumVoodoo]

One reason is to be found in what these are called. DC-link capacitors, developed specifically for ripple filtering duty. Solens are not since they don't say so explicitly.

Solen 22uF/630V---ca 12$CAD
EPCOS 20uF/630V--ca 15$US

The difference? EPCOS specify which application these are made for, Solen doesn't. Check the Epcos data sheet, then try to find similar info on Solens.
http://www.epcos.com/inf/20/20/db/fc_20 ... 74_678.pdf
https://www.solen.ca/pdf/solen/fastcap.pdf

About this, I'm confused. I've found Solen's work perfectly well for power supply duty - ripple filtering. Haven't tried the EPCOS.

OK, How about this:
EPCOS; WIMA; VISHAY and other major industrial suppliers' caps dedicated to things such as HF inverters, need to be, shall we say, on top of the game on all parameters all the time. Product slightly off specs can be classified as "good enough for audio"

When measuring caps you want to look for ESR, Q and Dissipation. Your typical PS series and Mallory's will have a Q of around 250 300 range.In a polyprop cap it will be in the 4k to 6k range..Big difference.
Tony

Tony,
Without telling at what frequency this complex capacitor has a resonant peak, Q is just a meaningless symbol without context. It certainly is nowhere near acoustic spectrum including bat sonar or ultrasonic vaporizers etc. Same applies to dissipation factor and ESR, however for ESR we can safely assume it is quoted at 60/120Hz or 50/100Hz, then 1, 10, 100kHz and higher in some extreme cases.
A resistance in series with ripple smoothing cap, yes, it works to soften the PSU by increasing B+ source resistance. In a different context I would say it degrades the PSU performance. Which in context of guitar amps can be a good thing.
How about a 20 Ohm or higher wire wound pot? Or MOSFET as a variable resistor? Try it, works like a charm. Have fun!

[rp]

Black Gate used to make a 22uF 6.3V non-polar film cap in the premium NX line.

David, this is beyond my pay grade, but I don't think BGs were film but some proprietary and exotic carbon powder & paper construction. Not typical wet or dry but still an electrolytic. They sure were all the rage for a while. If so successful I find it odd that they closed up and no one is carrying on the technology. Even if expensive, when has there not been a market for expensive things?

I found this and sped read through it:

http://www.audiotuning.de/english/index ... -tech.html

very fine conductive particles of super-high-purity graphite and other materials are distributed with no contact and applied in the separator made of thin capacitor paper.

Black Gate BG-NX non-polarized capacitors, conductive separator particles function as gate electrodes in stabilizing the zero potential of alternating current. Non-polarized Black Gate capacitors offer enhanced performance. They beat out every other kind of capacitor. The fine graphite particles play an important role here, as well...

...The BLACK carbon particles functioning as a GATE electrode is where the name "Black Gate" comes from.

[talbany]

Without telling at what frequency this complex capacitor has a resonant peak, Q is just a meaningless symbol without context.

Sure, same really could be said for impedance (AC resistance) in guitar amps..We generally discuss these types of things in general guitar amp context...If I read your post correctly?..
Context.
Dissipation and Q are related to one another by (D= 1/Q)..

Have Fun as Well!
Tony.

[VacuumVoodoo]

Without telling at what frequency this complex capacitor has a resonant peak, Q is just a meaningless symbol without context.

Sure, same really could be said for impedance (AC resistance) in guitar amps..We generally discuss these types of things in general guitar amp context...If I read your post correctly?..
Context..
Dissipation and Q are related to one another (D= 1/Q)..

Have Fun Back
Tony.

D = 1/Q = ESR/Xc where Xc is capacitors reactance thus all vary with frequency.
QED.
More fun to you

[Jana]

DQ? Dairy Queen? Alright, now we are talking my language--I'm all over that--let's go!

(Aleksander, Diary Queen is an ice cream/greasy bacon burger chain in the US. If you're not afraid of cholesterol, this is the place to go!)

[VacuumVoodoo]

Q.E.D. is an initialism of the Latin phrase quod erat demonstrandum,
meaning "which had to be demonstrated".
BTW I don't mix dairy products and meat so I'll pass on the DQ, but thanks for info and warning

[gingertube]

Blackgates were not film. Their value in tube HiFi as cathode bypasses etc was in their 40dB+ superior noise floor.
I have used Blackgate N and NX and PK in tube amps and Standard, N and NX in SS Power Amps. In all cases fitting the Blackgates in lieu of ordinary electrolytics was like lifting a "veil" to reveal an whole new level of micro detail which was being masked by the low level broadband noise from those garden variety electrolytics. That mask was not really apparent until it was not there anymore.

On one of tube HiFi Amps I built, I used all polyproylene (60uF/900V) B+ caps. It does have a gorgous loping pace but it was primarily done for the same reason I would use them in a guitar amp - to give reliability and lifetime.

While electrolytics of recent manufacture are FAR superior to the ones being made 20+ years ago (due to demands of switch mode power supplies on recent electrolytic design) they will still have a limited lifetime and eventually the amp will need a recap. By comparison film caps should last forever.

Understanding the ripple current and choosing electrolytic capacitors to suit will go a long way to address this. Basically electrolytic life is determined by its self heating. That is determined by the power dissipated internally which is Ripple Current Squared x Equivalent Series Resistance (ESR). Choosing a capacitor with 2 x the ripple current rating will give you 2 squared = 4 times the life ( 3 times the rating => 9 times the life etc.)

There is no reason to use electrolytics for cathode bypasses in a guitar amp. Cathode bypass caps beyond 4u7 are just not required (in any sensible design) and that is well within the range of values you can get in a plastic film cap. I would cetainly be insisting on film cathode bypass caps for the input stage, purely on a noise basis. Similarly I would use film B+ bypass cap for the input stgae.

Cheers,
Ian

[Leo_Gnardo]

Q.E.D. is an initialism of the Latin phrase quod erat demonstrandum,
meaning "which had to be demonstrated".
BTW I don't mix dairy products and meat so I'll pass on the DQ, but thanks for info and warning

HA! We used to translate Q.E.D. as Quite Easily Derived ... but we knew what it really meant. kollege students we were.

DQ will make you a nice burger without cheese nor bacon, still mighty good. The one around the corner employs cute hi school girls as servers. Anything to attract the customers.