Paralleling KSG Silver Mica Caps

RE: Paralleling KSG Silver Mica Caps, posted on July 20, 2017 at 17:04:28
gusser Audiophile Posts: 3649 Location: So. California Joined: September 6, 2006	Lead inductance - especially if this is an RF circuit. Could even be a problem at audio bandwidths.

RE: Paralleling KSG Silver Mica Caps, posted on July 20, 2017 at 17:10:18
Zedguy Audiophile Posts: 142 Joined: September 14, 2007	Thanks. I thought perhaps there would be issues.

lead inductance, posted on July 20, 2017 at 18:36:07
RayP Audiophile Posts: 727 Location: Maryland Joined: June 30, 2005	Not being familiar with this term, I looked it up and it made me wonder if lead inductance might be desirable in AF circuits where presumably we don't hear above 20 khz, or far less for many of us. Which leads to the question. Is it better to go for maximum bandwidth or sufficient bandwidth? I have no opinion one way or the other. Just curious. ray

Which leads to the question. Is it better to go for maximum bandwidth or sufficient bandwidth?..., posted on July 21, 2017 at 02:47:04
Cleantimestream Audiophile Posts: 7551 Location: Kentucky Joined: June 30, 2005	Exactly, as for my hearing @ 55 above 12K drops off like err, A cliff, the point is exactly the same as the one arguing has on his/her head. Just opinion here. The Mind has No Firewall~ U.S. Army War College.

RE: lead inductance, posted on July 21, 2017 at 06:37:44

gusser
Audiophile

Posts: 3649
Location: So. California
Joined: September 6, 2006

If you mean counting on the component lead as a deliberate inductor, I would say that's bad engineering in an audio circuit. If you need a small inductor, then put one in. Of course sometimes component artifacts do work in your favor. And in cheap consumer gear, tricks like this were pulled all the time. Look at an AA5 radio as the king of corner cutting circuit design.

In RF circuits, the lead inductance is a design nuisance and often must be worked around. That's why it is often recommended here never to recap tuner circuits. Mica and ceramic capacitors rarely fail and just moving them in old point to point wired circuits can change RF alignment.

A lot of people also think todays high speed digital circuits are the product of superior semiconductor engineering. And that would be only partially true. But another factor is surface mount technology and FPGA's. Here, the lead length is much smaller or even microscopic in an FPGA. This means much higher clock rates can be run without the problems of component lead length and their associated inductance.

To a point., posted on July 21, 2017 at 06:45:24

gusser
Audiophile

Posts: 3649
Location: So. California
Joined: September 6, 2006

General rule of thumb for measuring equipment is 5x the measured bandwidth. So to accurately measure a 20khz square wave, your scope needs at least 100khz of bandwidth.

And I would say today, an audio amplifier that goes out to 100khz is good but also the reasonable limit. Going beyond the 5x rule is just wasting money and additionally inviting stability issues you don't need.

A popular example is the HK Citation power amps with their video pentodes in the driver section. But there was a very good reason for this. Remember that the Citation uses nested feedback loops. Stu Hegman needed video bandwidth to get the gain he required and keep phase shift low at the same time. But as you can see these were somewhat complicated audio amplifier circuits for the day. The complexity is in the math of selecting those component values.

RE: Paralleling KSG Silver Mica Caps, posted on July 21, 2017 at 08:03:59
Stuben Audiophile Posts: 669 Location: Guber Ohio Joined: December 30, 2005	Try and have a listen. Don't take our word for it.

RE: To a point., posted on July 21, 2017 at 08:12:32

Triode_Kingdom
Audiophile

Posts: 10049
Location: Central Texas
Joined: September 24, 2006

"General rule of thumb for measuring equipment is 5x the measured bandwidth. So to accurately measure a 20khz square wave, your scope needs at least 100khz of bandwidth."

Not to nitpick your comment, but for accurate visual analysis of square waves, 10X is a better rule. We have scopes where I work with bandwidths of 1.5 gHz, and we make (among other things) devices that output TTL-level square waves to beyond 200 MHz. I've tested 80 MHz square waves (one of our "standard" frequencies) with both a 400 MHz scope and our 1.5 gHz scope. There's a marked improvement with the higher frequency instrument.

RE: To a point., posted on July 21, 2017 at 08:27:00

PakProtector
Audiophile

Posts: 12366
Joined: May 14, 2002

I think the video pentode was just an easy choice driven by output impedance, gain and the required plate dissipation rating required to get the first two. For gain, gm and load resistance rule, and with output impedance, the load is set...which drives plate dissipation requirements well beyond the usual signal pentodes.
cheers,
Douglas

Friend, I would not hurt thee for the world...but thou art standing where I am about to shoot.

RE: Paralleling KSG Silver Mica Caps, posted on July 21, 2017 at 10:47:54
Zedguy Audiophile Posts: 142 Joined: September 14, 2007	I am having my tube preamp recapped and wanted to try the KSG as reviews say they are excellent. The values required are 1uf and 2uf and considered paralleling many 0.1uf KSG and FT-3 Teflon. I don't really want to spend the money on 60 caps in the event the experiment fails. I suppose I should just stick with poly caps of the single value and save myself grief.

RE: To a point., posted on July 22, 2017 at 01:42:39

Cleantimestream
Audiophile

Posts: 7551
Location: Kentucky
Joined: June 30, 2005

My bad.

I was thinking almost exclusively of speakers and the desire to go far beyond 20k. This IS DIY Tube.

Am well aware the 'usefulness' of wide bandwidth in output transformers and agree 100k is excellent and in some cases {yours mentioned} required. Writing of output transformers, Silk and Hashimoto are the equal {better?} of many highly revered ones.
The Mind has No Firewall~ U.S. Army War College.

Tube DIY Asylum

Paralleling KSG Silver Mica Caps

76.67.127.83