Why do they use a Gimmick Capacitor?

[o***@tubes.com]

I was just making a list of capacitors in my Hallicrafters SX-99, so I
know what to buy for a recap job. I wrote down all the paper and
electrolytic caps, but will leave the ceramic caps. In the list of caps
on the schematic, they list a 2mmfd (pf) Gimmick Capacitor. I recall
wondering what that was years ago, long before the internet existed. I
just googled it and found this on Wikipedia. (very short article).

[quote]

From Wikipedia, the free encyclopedia

A gimmick capacitor is a capacitor made by twisting two pieces of
insulated wire together. The capacitance may be varied by loosening or
tightening the winding. The capacitance can also be reduced by
shortening the twisted pair by cutting. The available capacitance is on
the order of 1pF/inch (0.4 pF/cm). [1]

[end quote]

Yep, that's the whole article.....

Ok, so now I know what it is.....

But why do they do it this way? Why not just use a 2pf capacitor?

And I also have to ask, how thick the wire insulation should be. I'd
think that would make significant difference.

I dont know if they still use this in modern radios, but this was one of
the stranger things they did in the old radios...

If this can be explained in more detail, I'd recommend updating the
article on Wikipedia. It's real lacking in detail, and purpose.

[Peter Wieck]

Post by o***@tubes.com
But why do they do it this way? Why not just use a 2pf capacitor?
And I also have to ask, how thick the wire insulation should be. I'd
think that would make significant difference.
I dont know if they still use this in modern radios, but this was one of
the stranger things they did in the old radios...
If this can be explained in more detail, I'd recommend updating the
article on Wikipedia. It's real lacking in detail, and purpose.

A few of things:

a) the biggest hint is right in the article - the amount of 'twist' will vary the amount of capacitance - a fixed cap will not do this.
b) Gimmicks are typical of high-frequency devices (SW & FM) - and most specific to fine adjustments when aligning. But, as it happens, there are a few FM tuners out there that use gimmicks that may, or may not, remain in-circuit at the end of the alignment process (c.f. Dynaco FM3).
c) some gimmicks are 'air-bleed' for lack of a better term. They are often single bits of wire connected at only one end (as above).

If you want an exaggerated effect - consider what happens when you touch a grid cap. In some cases, the radio goes silent, in some cases, you get a massive hum, in some cases you act as an antenna... a gimmick performs the same function at a microscopic level.

Enamel wire works best.
Phone wire is used very typically as there is no danger of insulation failure, similarly, kynar wire.

Here is link to a how-to:

Peter Wieck
Melrose Park, PA

[Foxs Mercantile]

Post by o***@tubes.com
But why do they do it this way? Why not just use a 2pf capacitor?
And I also have to ask, how thick the wire insulation should be. I'd
think that would make significant difference.

First off, it's cheaper than a fixed capacitor, secondly it's
adjustable.

Speaking of SX-99s I use a gimmick capacitor for the BFO coupling to
the product detector so I can adjust the level correctly.
Same audio output for both product detector and diode detector.

[o***@tubes.com]

Post by Foxs Mercantile

Post by o***@tubes.com
But why do they do it this way? Why not just use a 2pf capacitor?
And I also have to ask, how thick the wire insulation should be. I'd
think that would make significant difference.

First off, it's cheaper than a fixed capacitor, secondly it's
adjustable.
Speaking of SX-99s I use a gimmick capacitor for the BFO coupling to
the product detector so I can adjust the level correctly.
Same audio output for both product detector and diode detector.

OK, this makes sense. I figured it would be cheaper, and I can see it
being "adjustable". What puzzles me, is how do you adjust the number of
turns and length while the radio is operating, or do you have to keep
turning the radio off and change the wires using trial and error, until
you get it right?
Although more costly, it seems a trimmer cap would be easier to adjust.

Since phone wire was mentioned, I got to thinking about the twisted
pairs used in phone lines. This is off the topic, but there must be a
lot of capacitance in those phone lines, when the twisted pair wires
travel long distances. Obviously it works, because it's been done that
way for decades, but I've always wondered if that capatance had any
effect on the phone at the end of those wires. I have thought about this
long before this thread, while I was wiring my own phones.

One last thing, regarding the SX-99. All the paper caps are rated at
600v or less. Almost all the sellers of replacement caps are selling
caps rated at 600 to 650 volts. These should be suitable for almost all
caps in old tube radios. However, this SX-99 schematic (bama download),
shows that C47 is a .0022 cap rated at 1000 volts. (and is listed as a
paper cap). This cap goes from the plate on the 6K6GT audio output tube
to ground. I'm wondering why that particular cap is rated so high
(voltage) when there is no voltage that high in the radio?

Of course I wont replace it with anything lower than the original, so
where can I buy a cap rated at 1000V?

(I guess for now, I'll have to just leave the old cap until I can find a
suitable replacement).

[Foxs Mercantile]

Post by o***@tubes.com
OK, this makes sense. I figured it would be cheaper, and I can see it
being "adjustable". What puzzles me, is how do you adjust the number of
turns and length while the radio is operating, or do you have to keep
turning the radio off and change the wires using trial and error, until
you get it right?
Although more costly, it seems a trimmer cap would be easier to adjust.

Surprisingly, MOST of the applications of a gimmick were NOT adjustable.
Or I should say, did NOT need to be adjustable. They were just a small
capacitor value and that's all it needed to be.

Post by o***@tubes.com
Since phone wire was mentioned, I got to thinking about the twisted
pairs used in phone lines. This is off the topic, but there must be a
lot of capacitance in those phone lines, when the twisted pair wires
travel long distances. Obviously it works, because it's been done that
way for decades, but I've always wondered if that capatance had any
effect on the phone at the end of those wires. I have thought about this
long before this thread, while I was wiring my own phones.

Two separate questions.
Usually they are made with solid conductor hookup wire. With enough
insulation on the to be adequate for the voltage across them.

Phone line pairs. "Transmission line" Either "open wire" or "twisted
pair." The capacitance between the wires is NOT an issue when they are
used as a transmission line with the correct impedances. Source and
load.

Post by o***@tubes.com
One last thing, regarding the SX-99. All the paper caps are rated at
600v or less. Almost all the sellers of replacement caps are selling
caps rated at 600 to 650 volts. These should be suitable for almost all
caps in old tube radios. However, this SX-99 schematic (bama download),
shows that C47 is a .0022 cap rated at 1000 volts. (and is listed as a
paper cap). This cap goes from the plate on the 6K6GT audio output tube
to ground. I'm wondering why that particular cap is rated so high
(voltage) when there is no voltage that high in the radio?

It's there across the plate of the audio output tube.

Post by o***@tubes.com
Of course I wont replace it with anything lower than the original, so
where can I buy a cap rated at 1000V?

Use a ceramic disc capacitor.

[o***@tubes.com]

Post by Foxs Mercantile
It's there across the plate of the audio output tube.

Post by o***@tubes.com
Of course I wont replace it with anything lower than the original, so
where can I buy a cap rated at 1000V?

Use a ceramic disc capacitor.

That solves that !!!!
A ceramic disc it will be....

Thanks

[Michael Black]

Post by Foxs Mercantile

Post by o***@tubes.com
OK, this makes sense. I figured it would be cheaper, and I can see it
being "adjustable". What puzzles me, is how do you adjust the number of
turns and length while the radio is operating, or do you have to keep
turning the radio off and change the wires using trial and error, until
you get it right?
Although more costly, it seems a trimmer cap would be easier to adjust.

Surprisingly, MOST of the applications of a gimmick were NOT adjustable.
Or I should say, did NOT need to be adjustable. They were just a small
capacitor value and that's all it needed to be.

And it could have just as easily be "bring this wire close to this other
wire", and I think there are instances of that. But a "gimmick" is more
formal, it looks less like a wire that goes nowhere (which might cause
someone to move it without realizing why it's there, and then wonder what
that loose wire was for). There's probably a level of strength to two
twisted wires versus leaving a wire hanging close to some other point.

Michael

[Jim Mueller]

On Thu, 13 Apr 2017 14:29:43 -0400, oldschool wrote:

snip

Post by o***@tubes.com

Post by o***@tubes.com
And I also have to ask, how thick the wire insulation should be. I'd
think that would make significant difference.

They designed it to use whatever type of wire the manufacturer had. If
the insulation were thicker, they added more turns. That doesn't help
much if it needs to be replaced, but most of the time the capacitance
wasn't all that critical to begin with.
snip

Post by o***@tubes.com
Although more costly, it seems a trimmer cap would be easier to adjust.

True, but 2pF trimmers are hard to come by; most are much higher value.

Post by o***@tubes.com
Since phone wire was mentioned, I got to thinking about the twisted
pairs used in phone lines. This is off the topic, but there must be a
lot of capacitance in those phone lines, when the twisted pair wires
travel long distances. Obviously it works, because it's been done that
way for decades, but I've always wondered if that capatance had any
effect on the phone at the end of those wires. I have thought about this
long before this thread, while I was wiring my own phones.

The reason for twisting the wires is to prevent crosstalk from other
pairs connecting other customers and to prevent other hum or noise
pickup. Since the wires are twisted, each gets exposed to the same
electric and magnetic fields so the effects cancel. Different pairs are
twisted a different number of turns per foot so that the same two wires
in adjacent pairs don't always match up with each other. As others have
explained, operating the pair as a matched transmission line cancels the
effect of the capacitance and inductance of the wires. This is the same
theory as the transmission lines used to connect transmitting antennas.
Consult any book on the subject, such as the ARRL Handbook, for an
explanation.

Post by o***@tubes.com
One last thing, regarding the SX-99. All the paper caps are rated at
600v or less. Almost all the sellers of replacement caps are selling
caps rated at 600 to 650 volts. These should be suitable for almost all
caps in old tube radios. However, this SX-99 schematic (bama download),
shows that C47 is a .0022 cap rated at 1000 volts. (and is listed as a
paper cap). This cap goes from the plate on the 6K6GT audio output tube
to ground. I'm wondering why that particular cap is rated so high
(voltage) when there is no voltage that high in the radio?

Ah, but there are voltages that high. The output tube is fed from the
full B+ voltage. This is the voltage on the plate with no signal. Now,
when there is a signal, it has to develop a voltage across the primary of
the output transformer in order for anything to reach the speaker. When
the signal at the grid goes positive, the tube draws more current and the
plate voltage drops; no conceptual problem here. However, when the grid
goes negative, the plate current drops and the voltage rises. Since the
no signal value is the B+ voltage, in order to develop a signal, the
voltage has to swing ABOVE the B+ value. The inductance of the
transformer and speaker make that happen. How high the voltage rises is
determined by how loud the set is playing, how fast the plate current
drops, and the value of the capacitor connected there. If there is a
loud noise spike, like a lightning crash, the voltage can get very high
indeed. Or if the set is operated with no load on the transformer, high
voltages can occur even at reasonable volume settings and with no
lightning. I have seen tube sockets that have been burned out by arcing
caused by these effects. It could just as well take out the output
transformer or a capacitor with insufficient rating.

Post by o***@tubes.com
snip
(I guess for now, I'll have to just leave the old cap until I can find a
suitable replacement).

Bad idea. If that cap shorts, it can burn out the output transformer,
rectifier, power transformer, or anything else in the current path.