24
\$\begingroup\$

This is from the datasheet for the TDA8217 PAL decoder and video processor. This is the "Applicaton Diagram".

The circled item is unfamiliar to me, I asked at my local electronic components distributor and no-one there recognized it either. It looks like an inductor, but seems to be connected to ground maybe?

TDA8217

\$\endgroup\$

2 Answers 2

48
\$\begingroup\$

Note how it says "330ns", a proper inductor would have a value in Henries.

My guess is that this is an Analog delay line like this:

enter image description here

Note how it has 3 connections of which one is ground which is connected to the tube and the shield.

I actually came across one of these many many years ago, probably in some video equipment I scrapped for parts. At the time didn't realize it was a delay line.

The "330 ns" and a search for "delay line" led me to the Wikipedia page. The delay line in the picture has 450 ns delay so 330 ns looks like that could be achieved with a similar component.

Sidenote: there are also Piezo electrical delay lines (see an example here) used in analog video equipment and I have seen lots of these and most have a 64us delay time which relates to the old PAL television standard (64 us is one horizontal line of the image). But 64 us is much larger than 330 ns so I knew that these were different types.

\$\endgroup\$
6
  • 11
    \$\begingroup\$ I wasn't even aware of these being a discrete component thing! Thanks for this epiphany :D \$\endgroup\$ Jan 7, 2020 at 7:16
  • \$\begingroup\$ @MarcusMüller Same for me. In all I've done with old equipment going back to 1940's radar equipment, I've yet to meet one of these (or, at least, knowingly so.) I learned something today. \$\endgroup\$
    – jonk
    Jan 7, 2020 at 8:23
  • \$\begingroup\$ And I'm a third who has learnt something today. Never heard of these! \$\endgroup\$
    – MCG
    Jan 7, 2020 at 8:25
  • 2
    \$\begingroup\$ Does... Does this just induce a delay because it takes the signal a longer time to propagate through longer lengths of wire? \$\endgroup\$
    – Sidney
    Jan 7, 2020 at 19:22
  • 4
    \$\begingroup\$ @Sidney Generally speaking, yes. Light travels around 30cm per nanosecond, so assuming the speed of light through the wire is around 0.66c, a 330ns delay line would be ~6.5m of wire wound into a small space. Adjusting the length, wire material, and type of insulation used will all adjust the overall delay time. \$\endgroup\$
    – bta
    Jan 7, 2020 at 20:19
21
\$\begingroup\$

This is not an answer, but I'm not allowed to submit a "comment".

As for what this component does, separate processing/filtering of video luminance and chrominance introduce different (group) delays in the two signals. This can become noticeable in the TV picture (see, for example https://www.bbc.co.uk/rd/publications/rdreport_1973_35).

The delay line shown is to counteract this, bringing the two components into sync. (It is separate from the glass delay line usually used in PAL and similar receivers to ameliorate one of the shortcomings found with NTSC).

The TDA8217 datasheet doesn't specify the delay time of this component, since the best value depends on details of the surrounding circuit (basically the delay introduced in the chrominance path).

\$\endgroup\$
1
  • \$\begingroup\$ That was very interesting, thank you! \$\endgroup\$
    – hyarion
    Jan 9, 2020 at 4:56

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.