Searching \ for '[EE]: Analogue delay lines' in subject line. ()
Make payments with PayPal - it's fast, free and secure! Help us get a faster server
FAQ page: www.piclist.com/techref/io/atod.htm?key=analog
Search entire site for: 'Analogue delay lines'.

Exact match. Not showing close matches.
PICList Thread
'[EE]: Analogue delay lines'
2001\01\27@034841 by Pete

flavicon
face
I've come across a VGA to PAL composite converter that calls for a video
'delay line' of 400ns.

I looked through my junk and discovered two such modules. They are branded
with the Matsushita triangle, and have such cryptic part numbers as VLD0014
and EFD-EN645, respectively.

I know what digital delay lines are (and how they can be built cheaply from
slow logic gates), but I am not familiar with analogue delay lines. What's
inside these black boxes (literally, they are black and square) I know that
RC/RL circuits introduce phase shifts, is this how they work. I've noticed
that one such device is represented on the schematic with a tapped inductor
symbol. It's hard to imagine why they are so hard to get hold of.

The ones I have have four legs, but the circuit specifies one with three
connections - Does anybody know anything about these specific devices I
mentioned, and whether they are suitable substitiution for a 3 legged
device?

Thanks

Peter L.

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2001\01\27@085306 by Olin Lathrop

face picon face
> I know what digital delay lines are (and how they can be built cheaply
from
> slow logic gates), but I am not familiar with analogue delay lines. What's
> inside these black boxes (literally, they are black and square) I know
that
> RC/RL circuits introduce phase shifts, is this how they work. I've noticed
> that one such device is represented on the schematic with a tapped
inductor
> symbol. It's hard to imagine why they are so hard to get hold of.

The R/L type of analog delay line is an attempt to approximate a
transmission line with discrete components.  Think of an L in series,
followed by a C accross, followed by another L in series, followed by
another C accross...  Now do the usual math hand waving of making an infinte
number of these, but making each infinitely small (in other words, you
picture the integral instead of a finite summation).  That's a transmission
line.  The total L and C over a fixed length defines the transmission line
impedance.  It also defines the propagation speed.  Series R also enters
into this, but can be ignored for the kinds of purposes we are talking about
here.

A rough rule of thumb is that the propagation speed of realizable
transmission lines is about 1/2 the speed of light in vacuum.  Therefore,
your 400nS of delay can be achieved by about 200 feet of coax.  Some delay
lines are actually built this way.  From you description, however, you have
a box far too small to cram 200 feet of transmission line into.  In that
case they approximated the transmission line with some number of L C L C L C
...  This type of approximation is valid up to some frequency limit.  Above
that frequency you will get varying delay and large attenuation.

Note that when I said "transmission line" above I was using it in the signal
processing sense.  A "transmission line" for delivering 60Hz power to your
local substation is a totally different thing.


*****************************************************************
Olin Lathrop, embedded systems consultant in Devens Massachusetts
(978) 772-3129, spam_OUTolinTakeThisOuTspamembedinc.com, http://www.embedinc.com

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2001\01\27@133603 by rottosen

flavicon
face
Olin Lathrop wrote:
{Quote hidden}

I just looked at an old delay line removed from an oscilloscope. It is a
coil wrapped around about a 4 feet long polyethelene core. Actually it
is 2 coils wrapped around the core at an angle to each other. Why 2
coils? Does the capacitance between the 2 coils act as the distributed
capacitance? This delay line may be used in the oscilloscope
differentially?

I know -- more questions than answers.


-- Rich


{Quote hidden}

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2001\01\27@153316 by Robert Rolf

picon face
What you probably have is 'ultrasonic' delay lines, as commonly used
in VCR's for Y/C separation (comb filter). The video is AM modulated
on a 3 or 4 times color carrier and bounced back and forth on the inside
surfaces of the glass block with has _nearly_ parallel sides, and then
demodulated.
The reason you have 4 pins is that you have two INDEPENDANT piezo
transducers
on the block.

What you are probably looking for is the traditional 'distributed' delay
line commonly found in older TV's. It was a 6" long 1/2" diameter coil
of wire commonly drawn as an inductor on it's side, with a wide flat
line under the symbol. Just look inside and circa 1970-180 televsion to
easily find it ('cause it's so big).

http://www.lionlmb.org/quad/theory.html
has a nice history of television recording methods.
You old broadcast video guys should lick the link to 'Quadruplex park'.
Talk about flashbacks...

Pete wrote:
{Quote hidden}

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2001\01\28@004504 by rottosen

flavicon
face
Those old tube type scopes were real big and long.  ;-)

Actually, the delay line is 2 layers of of about 3 1/2 turns per layer.
The delay line unit is about 2 1/4" wide by 4 1/2" long by 3/8" thick.


-- Rich


Jim Paul wrote:
>
> Richard,
>
> If the core is 4 feet long, I like to see the scope.
>
>                                             Regards,
>
>                                                 Jim
> {Original Message removed}

2001\01\28@004651 by rottosen

flavicon
face
Olin Lathrop wrote:
>
> > I just looked at an old delay line removed from an oscilloscope. It is a
> > coil wrapped around about a 4 feet long polyethelene core. Actually it
> > is 2 coils wrapped around the core at an angle to each other. Why 2
> > coils? Does the capacitance between the 2 coils act as the distributed
> > capacitance? This delay line may be used in the oscilloscope
> > differentially?
>
> I can only speculate here.  Perhaps they only needed a relatively short
> delay that they could achieve with a long wire.


The delay only has to be long enough to get the scopes horizontal sweep
started so you can see the trigger point on the CRT. I went and looked
at the schematic for a Tektronix 465 scope and it has 120ns of vertical
channel delay.


>  The two coils may be an
> attempt allow for longer wire without multiplying the inductance.

In that same Tek 465 schematic, the delay line is driven and received
differentially. So, it would seem that one coil is in each of the
differential paths. I would not want to predict what kind of
interactions there are with this scheme.


>Or they
> could have had "standard" delay modules and needed two of them to get the
> desired delay in this case.

I suspect not. Tektronix tends to make things exactly for what they need
them to do. In other words, if they needed more delay than an existing
part supplied then they would design a new part.


>
> Sometimes you can see two layers of wire wrapped around the same rod but in
> opposite directions, sometimes with a layer of conductive material (but not
> magnetic) in between.  That's an attempt to have the inductances of the two
> coils cancel each other out.  You also see this in some wire wound
> resistors.

Interesting point. Wow, how DOES this thing work?


-- Rich


>
> *****************************************************************
> Olin Lathrop, embedded systems consultant in Devens Massachusetts
> (978) 772-3129, olinspamKILLspamembedinc.com, http://www.embedinc.com

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2001\01\28@135200 by Peter L. Peres

picon face
VLD0014 is a Panasonic (spare) part order number. It is impossible to tell
what kind of line it is without more info.

There are two kinds of analog delay lines in use: glass acoustic delay
lines and allpass-filter types. The picture with the tapped coil
represents an allpass-filter.

The allpass-filter is usually built out of a length of wire wound on an
insulator and surrounded by a shield. It forms a kind of transmission line
whose propagation speed is known. By making it as long as reuired, delays
in the range 5nsec - 2usec (approximately) are obtained. These are usually
squarish in section and much longer than any other dimension. They have
four pins at the ends. You can test them with an ohmmeter to pass DC (abt.
zero ohms) between each pair of two far pins.

The glass delay lines are much taller and thinner than the
allpass-filters, they have three or more terminals, none of which read any
resistance on the ohms scale (between any of them). Sometimes there are
two ground pins and those are obviously shorted. They come in ca. 10usec
to 500usec values. 64 usec is common in TVs and VCRs for color
demodulation, and one of the parts you have bears strikingly similar
numbers...

So, the other one is probably a 330 or 400 nsec allpass delay line from
the same set, and this is probably what you want.

hope this helps,

Peter

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2001\01\30@043217 by Pete

flavicon
face
Thanks Peter and everyone else who replied.

I think I can be sure that the two I have are of the acoustic variety. They
're both black boxes, 1" tall, 1/2 a cm thick. One's 'bout 5cm in length.
The other's twice as long. Both have four pins, none of which are in
continuity with any other.

At the surplus store, I bought a similar looking unit to the smaller of the
black Matsushita ones I have - pt.# Philips DL701. It's pale blue and has a
clip on lid rather than molded plastic shell like the others. I disassembled
it (without damage) - and indeed there was a slab of crystal (looks like a
frosted microscope slide).

I'm afraid these plastic boxes won't be of any use to me.

On the bright side, I opened up a junked 'National' TV and identified a
delay line module that is identical to the shorter of my two black
Matsushita specimens (even the same p# - EN645), BUT there was a second item
of interest, an unpainted tinned inductor-like box, measuring 1cmx1cmx4cm,
standing vertically. It had three legs, and was identified on the PCB as
DL-???. I think this might do the trick.

However, I want to leave the dead TV as a last resort. I also have on hand a
dozen hybrid LC modules (ugly, three legged beasts with lumps all over and
potted in ceramic). They were taken from a circa 1980 VCR, along with the
Matsushita delay lines. I'm mentioning them here because they resemble the
diagram of a delay line in the MC1377 datasheet. Unfortunately their part
numbers turned up nothing on the web. They are all TDK products, labelled
VLF-????.

I think they're mostly LC tuned filters, but I'm wondering if one, perhaps a
couple are actually delay lines. How can I test these little beasts? I have
a nifty 20Mhz function generator and a decent oscilloscope, so I thought I
might be able to use check the delay (phase shift) of a square wave that has
passed through device under test. Any comments?


Pete
.....newmannzKILLspamspam.....ihug.co.nz


{Original Message removed}


'[EE]: Analogue delay lines'
2001\02\01@161241 by Peter L. Peres
picon face
>testing delay lines

Well, the direct method involves a pulse generator (HF range) and a dual
channel scope. Remember to terminate the delay lines in their Z0. That's
anywhere between 220 ohms and 1Kohm for your range of timing and
application. The pulse generator should make a 200nsec wide pulse every
1usec or so. The scope probes go to the input and output of the filter,
sync from the probe on the input side, on '+' slope. The input termination
is series and output is parallel (to GND). Keep the wires very short, or
strange things will creep into the image. Do not drive the lines with more
than 1Vpp (that's 2Vpp out from generator, with 1Vpp dropping on the input
matching resistor. You will have matched impedances when the waveform at
the output is least distorted and the output amplitude is about 1/2 of the
generator output amplitude. Imho use two 1K carbon semivariable resistors
for termination and adjust them together.

imho you are going to discover just exactly how forgiving our (my) eye are
when watching TV upon seeing what that delay line does to the pretty
square wave <g>.

If you do not have a pulse generator you can quickly build one using a few
TTL (LS, HCT) parts on a breadboard. You need a gate oscillator with a
16-20MHz crystal, a binary counter and a NAND gate to gate Q0 and Q1 from
the counter. The output of the NAND is a 1:4 duty cycle pulse proper for
your TDR experiment ;-). This is 5 minutes of soldering, more or less.

good luck,

Peter

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


More... (looser matching)
- Last day of these posts
- In 2001 , 2002 only
- Today
- New search...