/ News & Press / Video / About Capacitors ... Part 5 Non-Linear Capacitor by Andy Davies

About Capacitors ... Part 5 Non-Linear Capacitor by Andy Davies

WEBVTT
Kind: captions
Language: en

00:00:00.000
I may remember I said I'll tell you a
00:00:02.690 00:00:02.700 little bit about non linear capacitors
00:00:05.360 00:00:05.370 this capacitor is linear that is the the
00:00:11.209 00:00:11.219 vanes are mounted centrally about the
00:00:14.570 00:00:14.580 shaft so as we turn the shaft so the
00:00:20.109 00:00:20.119 vanes move smoothly and evenly so if we
00:00:26.000 00:00:26.010 have it here the wing say that we've got
00:00:29.890 00:00:29.900 half of the capacitance engaged if we
00:00:33.560 00:00:33.570 have it there we've got all of the
00:00:35.360 00:00:35.370 capacitance engaged and it's
00:00:39.020 00:00:39.030 proportional so as we move the the
00:00:45.470 00:00:45.480 tuning knob not the dial so as a
00:00:48.590 00:00:48.600 capacitor engages further into the fixed
00:00:52.189 00:00:52.199 vanes as the moving vanes engage into
00:00:54.680 00:00:54.690 the fixed vanes so the capacitor
00:00:56.959 00:00:56.969 increases proportionally and that's fine
00:01:00.799 00:01:00.809 that that's good and this capacitor by
00:01:07.219 00:01:07.229 comparison doesn't the if you look at
00:01:13.070 00:01:13.080 the difference between these two the
00:01:17.359 00:01:17.369 shaft is in the center of that span here
00:01:22.039 00:01:22.049 the shaft is off-center a bit difficult
00:01:26.539 00:01:26.549 to see but us even make that a little
00:01:28.940 00:01:28.950 clearer okay I've made two capacitors
00:01:32.420 00:01:32.430 here admittedly they're out of paper so
00:01:35.300 00:01:35.310 I've got a moveable vane they have
00:01:37.429 00:01:37.439 pivoted about the center this represents
00:01:40.580 00:01:40.590 my fixed vane and again my second
00:01:43.460 00:01:43.470 capacitor there it moves off center and
00:01:47.990 00:01:48.000 this is my fixed capacitor plane there
00:01:52.160 00:01:52.170 so if I start off with them both at the
00:01:54.649 00:01:54.659 bottom so we'll call that as zero
00:01:57.679 00:01:57.689 degrees I'm trying to work around the
00:02:00.170 00:02:00.180 tri-party so if we move this one into
00:02:04.490 00:02:04.500 the error 45 degrees
00:02:07.870 00:02:07.880 and put a little line on it that one
00:02:10.389 00:02:10.399 there had 45 degrees immediately you can
00:02:15.940 00:02:15.950 see that that is gone a quarter of its
00:02:18.790 00:02:18.800 way through there there's very little
00:02:21.820 00:02:21.830 engagement as we move through to 90
00:02:25.089 00:02:25.099 degrees and a mark there again clearly
00:02:33.070 00:02:33.080 this one the linear capacitor has moved
00:02:37.630 00:02:37.640 in half of its engagement but this one
00:02:41.680 00:02:41.690 is only on earth I don't know maybe a
00:02:44.890 00:02:44.900 third of its engagement may be less map
00:02:48.900 00:02:48.910 as we go on to 135 degrees on each and
00:03:01.990 00:03:02.000 we mark those and then a 180 clearly the
00:03:05.589 00:03:05.599 whole capacitor is going in both so
00:03:08.350 00:03:08.360 they've both gone from the minimum
00:03:09.910 00:03:09.920 capacitance to the maximum it's just
00:03:12.729 00:03:12.739 that the linear capacitor is gone in in
00:03:15.940 00:03:15.950 four equally divided steps but the
00:03:19.750 00:03:19.760 nonlinear capacitor is gone in for very
00:03:23.500 00:03:23.510 unequal steps and and I'll tell you why
00:03:32.500 00:03:32.510 whenever we use a capacitor like these
00:03:36.380 00:03:36.390 variable capacitors we always do so in
00:03:40.820 00:03:40.830 conjunction with a coil or an inductor
00:03:44.300 00:03:44.310 if you like and there's a strange
00:03:47.780 00:03:47.790 relationship between the coil and the
00:03:51.530 00:03:51.540 capacitor between them they establish
00:03:54.770 00:03:54.780 the resonant frequency or the tuning
00:03:57.140 00:03:57.150 frequency or the tank circuit frequency
00:04:01.220 00:04:01.230 all being the same thing but the
00:04:04.210 00:04:04.220 relationship is nonlinear it's said to
00:04:08.870 00:04:08.880 have a root function and this is the
00:04:13.520 00:04:13.530 formula and it's expressed as F equals 1
00:04:21.650 00:04:21.660 over 2 pie root LC where F is the
00:04:26.120 00:04:26.130 frequency in Hertz and pi is 3.14 one L
00:04:35.630 00:04:35.640 is the inductance in Henry's and C's the
00:04:40.640 00:04:40.650 capacitance in farad's a more practical
00:04:45.320 00:04:45.330 way to express this is to say that F
00:04:49.240 00:04:49.250 equals 10 to the 6 over 2 pie root LC
00:04:54.370 00:04:54.380 and that way frequencies in kilohertz
00:04:59.560 00:04:59.570 inductance is in micro henries and
00:05:02.870 00:05:02.880 capacitance is in Pico farad's so that's
00:05:06.200 00:05:06.210 a more usable of that's a friendlier
00:05:11.110 00:05:11.120 formula if you're involved with radio if
00:05:14.870 00:05:14.880 you're involved with other things then
00:05:16.670 00:05:16.680 you may not want to work in there
00:05:18.650 00:05:18.660 kilohertz and pick a Ference but both
00:05:23.420 00:05:23.430 formulas are correct they just work in
00:05:26.690 00:05:26.700 different values so the sizes of the
00:05:30.230 00:05:30.240 numbers change but the principle is
00:05:33.020 00:05:33.030 exactly the same okay so what does it
00:05:36.500 00:05:36.510 all mean what does it matter
00:05:38.530 00:05:38.540 it's worth remembering that the radio
00:05:42.370 00:05:42.380 frequencies say for the a.m. bands were
00:05:46.720 00:05:46.730 issued on a nine kilohertz or 10
00:05:49.750 00:05:49.760 kilohertz basis so the stations are
00:05:53.110 00:05:53.120 evenly distributed on the wave band and
00:05:57.790 00:05:57.800 they don't interfere with one another so
00:06:01.330 00:06:01.340 it's very desirable to have the wave
00:06:05.560 00:06:05.570 band divided in nice linear steps and
00:06:10.590 00:06:10.600 this satisfies the engineers obsession
00:06:14.590 00:06:14.600 with linearity and just whilst I'm
00:06:18.190 00:06:18.200 talking about radio dials I love the Art
00:06:21.460 00:06:21.470 Deco style used on some of these old
00:06:25.630 00:06:25.640 vintage radios I find that very
00:06:28.120 00:06:28.130 attractive for mechanical reasons it's
00:06:32.620 00:06:32.630 very desirable to have the linear
00:06:38.790 00:06:38.800 rotation of the capacitor associated
00:06:43.210 00:06:43.220 with the linear movement of the point
00:06:48.670 00:06:48.680 across the radio dial but because of
00:06:52.960 00:06:52.970 that route function on the bottom of the
00:06:56.650 00:06:56.660 formula it gives us the problem that
00:06:59.860 00:06:59.870 I've been talking about but fortunately
00:07:03.250 00:07:03.260 the capacitor manufacturer is taking
00:07:06.220 00:07:06.230 care of it for us by arranging for
00:07:11.220 00:07:11.230 progressively larger amounts of
00:07:13.750 00:07:13.760 capacitance to be engaged for a given
00:07:19.170 00:07:19.180 number of degrees of rotation of the
00:07:23.590 00:07:23.600 shaft
00:07:25.920 00:07:25.930 as compared with the shaft that he and
00:07:30.610 00:07:30.620 the linear capacitor where for each
00:07:34.380 00:07:34.390 proportion of rotation we get a
00:07:37.620 00:07:37.630 proportional amount of engagement
00:07:48.430 00:07:48.440 just one little point that I'll mention
00:07:51.970 00:07:51.980 on some equipment may be a good
00:07:58.380 00:07:58.390 communications receiver or an oscillator
00:08:02.500 00:08:02.510 a test oscillator you may see that these
00:08:05.890 00:08:05.900 out of fingers have been been tamed
00:08:08.890 00:08:08.900 they've been moved out slightly and that
00:08:12.790 00:08:12.800 is to find trim the frequency and to
00:08:18.190 00:08:18.200 find trim that non-linearity I say you
00:08:22.810 00:08:22.820 will see these sometimes bent out it's
00:08:25.750 00:08:25.760 only on the outer veins in in each case
00:08:30.000 00:08:30.010 if you do see that don't mess with it
00:08:32.860 00:08:32.870 because the chances of you improving
00:08:35.680 00:08:35.690 things are very remote and the chances
00:08:37.870 00:08:37.880 have you freaking it are probably very
00:08:40.990 00:08:41.000 high if you look at the spacing of the
00:08:47.160 00:08:47.170 graduations on this dial you'll see that
00:08:51.520 00:08:51.530 they're all bunched together at the
00:08:54.430 00:08:54.440 bottom of the dial and yet they are
00:08:58.410 00:08:58.420 spaced very far apart at the top of the
00:09:01.810 00:09:01.820 dial and that is the typical frequency
00:09:05.260 00:09:05.270 response that an LC circuit would have
00:09:09.400 00:09:09.410 if a linear capacitor is used so it's
00:09:16.360 00:09:16.370 not being compensated for so the
00:09:19.110 00:09:19.120 irregularity shows up on the dial rather
00:09:22.750 00:09:22.760 than in the shape of the capacitor plate
00:09:26.700 00:09:26.710 I'll put a couple of worked out examples
00:09:30.250 00:09:30.260 of that formula that I showed you
00:09:33.510 00:09:33.520 for those interested okay I hope you
00:09:37.500 00:09:37.510 found that interesting thanks for
00:09:39.870 00:09:39.880 watching
00:09:40.500 00:09:40.510 buh-bye okay his couple of worked
00:09:44.360 00:09:44.370 examples and that you see the first set
00:09:48.120 00:09:48.130 of calculations are in sort of radio
00:09:51.390 00:09:51.400 values and then the next set of
00:09:55.910 00:09:55.920 calculations are in Henry's and ferrets
00:10:01.710 00:10:01.720 but I figure you can stop and look at
00:10:04.110 00:10:04.120 this in your own time so thanks for
00:10:07.170 00:10:07.180 watching have fun with you Matt's bye

Sales phone