Author Topic: EEVblog #381 - How Do Ultrasonic Delay Lines Work?  (Read 27998 times)

0 Members and 1 Guest are viewing this topic.

Offline nitro2k01Topic starter

  • Frequent Contributor
  • **
  • Posts: 843
  • Country: 00
EEVblog #381 - How Do Ultrasonic Delay Lines Work?
« on: November 08, 2012, 11:05:20 pm »


Come on Dave, you should at least have hooked up the glass one to a pulse generator and probed with a 10M probe. You might have gotten something on the scope.

I'm sure all the stages in the coil delay were actually meant to be the same inductance, but your LCR meter was confused by the common connection sitting in parallel.
Whoa! How the hell did Dave know that Bob is my uncle? Amazing!
 

Offline EEVblog

  • Administrator
  • *****
  • Posts: 37661
  • Country: au
    • EEVblog
Re: EEVblog #381 - How Do Ultrasonic Delay Lines Work?
« Reply #1 on: November 08, 2012, 11:11:30 pm »
Come on Dave, you should at least have hooked up the glass one to a pulse generator and probed with a 10M probe. You might have gotten something on the scope.

I did, I got nothing.

Dave.
 

Offline brainwash

  • Frequent Contributor
  • **
  • Posts: 463
  • Country: de
    • Hack Correlation
Re: EEVblog #381 - How Do Ultrasonic Delay Lines Work?
« Reply #2 on: November 09, 2012, 12:23:54 am »
I did some back of the envelope calculations:

Code: [Select]
air:
330 m 1 s
x m 64 us

x=21120 * 10^-6 = [b]21.12 mm[/b]

quartz:
shear wave
3750 m/s -> 11.36 X multiplier -> [b]1.85 mm[/b]

compression wave
5900 m/s -> 17.87 X -> [b]1.18 mm[/b]

sonic attenuation 11db/m @ 1MHz

Edit: calculation is wrong, see several posts below

I actually found out that it should be a shear wave (probably that's why all the black spots are there) so to induce a delay of 64us the sound needs to be travelling about 1mm. In free air it would be 21mm, which still does not correspond with the convoluted multiple reflection path.

This sounded intriguing and after some seconds of googling I found this root patent:
http://www.google.com/patents/US3286205

It seems that the delay is mostly a fixed constant determined by the crystal cut, electrodes and piezo placements and some RC network.
Fast forward 30 years and the result is this:
http://www.google.com/patents/US5378947
"by changing the transconductance of differential amplifiers which constitute the active filter, the amount of delay of the glass can be changed"

Note: the amplifiers have to match the impedance of the glass, so there has to be some magic there to be able to cut the crystal exactly and/or tweak the active filter. I would assume the impedance gets tweaked on the production line and afterwards they tweak the delay as well. So, somewhere, because of the digital age, several "quartz glass delay ultrasonic impedance matcher and delay adjuster" jobs got lost.
« Last Edit: March 18, 2013, 11:38:22 am by brainwash »
 

Online tom66

  • Super Contributor
  • ***
  • Posts: 6678
  • Country: gb
  • Electronics Hobbyist & FPGA/Embedded Systems EE
Re: EEVblog #381 - How Do Ultrasonic Delay Lines Work?
« Reply #3 on: November 09, 2012, 12:56:15 am »
I used to see these in old VCRs, and never quite understood their function. Cheers Dave
 

Offline mikeselectricstuff

  • Super Contributor
  • ***
  • Posts: 13693
  • Country: gb
    • Mike's Electric Stuff
Re: EEVblog #381 - How Do Ultrasonic Delay Lines Work?
« Reply #4 on: November 09, 2012, 01:29:19 am »
I used to see these in old VCRs, and never quite understood their function. Cheers Dave
In VCRs they were used to conceal tape dropouts. If the signal momentarily drops out, it switches in a delayed signal from the previous line to avoid showing a black or white glitch.

The reason for the very short delay line in TV decoders is to delay the luminance signal slightly to compensate for delays in the colour processing, so the two arrive at the same time out of the decoder.
Youtube channel:Taking wierd stuff apart. Very apart.
Mike's Electric Stuff: High voltage, vintage electronics etc.
Day Job: Mostly LEDs
 

Online tom66

  • Super Contributor
  • ***
  • Posts: 6678
  • Country: gb
  • Electronics Hobbyist & FPGA/Embedded Systems EE
Re: EEVblog #381 - How Do Ultrasonic Delay Lines Work?
« Reply #5 on: November 09, 2012, 01:56:29 am »
Hmm, is that why when you have a bad tape, you get bars which seem to stretch the vertical? Never thought about that but it would make sense.
http://plethoralondon.files.wordpress.com/2011/02/vhsd-screen-prt-1.jpg
 

Offline JackOfVA

  • Supporter
  • ****
  • Posts: 350
  • Country: us
Re: EEVblog #381 - How Do Ultrasonic Delay Lines Work?
« Reply #6 on: November 09, 2012, 02:20:16 am »
There's a bit of black magic in the LC delay line.

Notice the spacing of the four inductors - they are grouped with two close spaced, a gap and then two more close spaced. The spacing is designed to yield a certain mutual inductance to improve the delay line performance.

In addition, breaking the delay line into multiple LC segments is necessary to improve the rise time.
 

Offline marinedalek

  • Contributor
  • Posts: 22
  • Country: gb
Re: EEVblog #381 - How Do Ultrasonic Delay Lines Work?
« Reply #7 on: November 09, 2012, 02:59:44 am »
You call that a quartz delay line? THIS is a quartz delay line! https://dl.dropbox.com/u/69960/quartzblock.jpg

As used in the BBC's Advanced Field-Store Standards Converter, used to convert 525-line NTSC footage to 626-line PAL and vice versa. The accuracy of the delay time was maintained by heating the blocks with electric iron elements(!). Here's one of the many BBC engineering documents on their converters: http://downloads.bbc.co.uk/rd/pubs/reports/1964-65.pdf which goes into horrific detail on the operation of the system.
 

Offline KevinGomez

  • Newbie
  • Posts: 1
Re: EEVblog #381 - How Do Ultrasonic Delay Lines Work?
« Reply #8 on: November 09, 2012, 06:15:25 am »
I did some back of the envelope calculations:

air: 330 m   1 s
x m   64 us

x=21120 * 10^-6 = 21.12 mm

quartz:
shear wave
3750 m/s -> 11.36 X multiplier -> 1.85 mm

compression wave
5900 m/s -> 17.87 X -> 1.18 mm


I believe your multiplier should actually multiply instead of divide since the path would need to be longer in the faster material - so 21.12mm x 11.36 ~ 24cm.  Patent US5130679 ( http://www.google.com/patents/US5130679?dq=US5130679&hl=en&sa=X&ei=0Z6cUKTaBMjMyQGU64HICw&ved=0CDQQ6AEwAA )
looks close and it (or a related patent) quotes a propagation speed of 2400m/s - so 64us would need 15.36cm.

 

Offline amyk

  • Super Contributor
  • ***
  • Posts: 8240
Re: EEVblog #381 - How Do Ultrasonic Delay Lines Work?
« Reply #9 on: November 09, 2012, 07:00:23 am »
I used to see these in old VCRs, and never quite understood their function. Cheers Dave
In VCRs they were used to conceal tape dropouts. If the signal momentarily drops out, it switches in a delayed signal from the previous line to avoid showing a black or white glitch.

The reason for the very short delay line in TV decoders is to delay the luminance signal slightly to compensate for delays in the colour processing, so the two arrive at the same time out of the decoder.
Also used in decoding PAL to avoid the Hanover bar effect.
 

Offline mikeselectricstuff

  • Super Contributor
  • ***
  • Posts: 13693
  • Country: gb
    • Mike's Electric Stuff
Re: EEVblog #381 - How Do Ultrasonic Delay Lines Work?
« Reply #10 on: November 09, 2012, 08:05:55 am »
Come on Dave, you should at least have hooked up the glass one to a pulse generator and probed with a 10M probe. You might have gotten something on the scope.

I did, I got nothing.

Dave.

I got something - the trick is to use a burst of 4.43MHz as these devices are narrowband :
Youtube channel:Taking wierd stuff apart. Very apart.
Mike's Electric Stuff: High voltage, vintage electronics etc.
Day Job: Mostly LEDs
 

Offline brainwash

  • Frequent Contributor
  • **
  • Posts: 463
  • Country: de
    • Hack Correlation
Re: EEVblog #381 - How Do Ultrasonic Delay Lines Work?
« Reply #11 on: November 09, 2012, 08:26:23 am »
I believe your multiplier should actually multiply instead of divide since the path would need to be longer in the faster material - so 21.12mm x 11.36 ~ 24cm.  Patent US5130679 ( http://www.google.com/patents/US5130679?dq=US5130679&hl=en&sa=X&ei=0Z6cUKTaBMjMyQGU64HICw&ved=0CDQQ6AEwAA )
looks close and it (or a related patent) quotes a propagation speed of 2400m/s - so 64us would need 15.36cm.

You are right, sorry! Thanks for pointing out the error.

But then I don't understand how on one side the delay is variable by changing some parameters (patent description) but on the other side it is EXACTLY 64us.
Someone care to explain?
 

Offline TVguy

  • Newbie
  • Posts: 2
Re: EEVblog #381 - How Do Ultrasonic Delay Lines Work?
« Reply #12 on: November 09, 2012, 10:06:07 am »
I used to see these in old VCRs, and never quite understood their function. Cheers Dave
In VCRs they were used to conceal tape dropouts. If the signal momentarily drops out, it switches in a delayed signal from the previous line to avoid showing a black or white glitch.

The reason for the very short delay line in TV decoders is to delay the luminance signal slightly to compensate for delays in the colour processing, so the two arrive at the same time out of the decoder.

The 1H delay line was used for luminance dropout compensation in VCRs during playback whereby a previous line was inserted when a dropout was detected. The 2H delay which was disected is a 128 microsecond delay line used for chroma crosstalk cancellation. Consumer video tape recorder systems used FM for luminance processing and a hetrodyning system for the chroma where the 4.433 MHz subcarrier was downconverted to an approximate frequency of 627 KHz (still QAM). Video head azimuth offset gave the luminance signal great crosstalk separation but the "colour under" chroma signal used a 90 degree phase rotational method per horizontal line of recording. During playback the 2H delay line would average out crosstalk errors. BTW from what I remember from my teaching days, the original material used for delay lines is known isopaustic glass.
 

Offline EEVblog

  • Administrator
  • *****
  • Posts: 37661
  • Country: au
    • EEVblog
Re: EEVblog #381 - How Do Ultrasonic Delay Lines Work?
« Reply #13 on: November 09, 2012, 11:41:28 am »
I got something - the trick is to use a burst of 4.43MHz as these devices are narrowband :

Yeah, I only did a quick test as a sweep around that range and expected to see it pop up. I guess it's much narrower than I expected.
Damn, I wanted to try the memory aspect of it as well but you beat me to it!
Nice work.

Dave.
 

Offline mikeselectricstuff

  • Super Contributor
  • ***
  • Posts: 13693
  • Country: gb
    • Mike's Electric Stuff
Re: EEVblog #381 - How Do Ultrasonic Delay Lines Work?
« Reply #14 on: November 09, 2012, 12:15:16 pm »
I got something - the trick is to use a burst of 4.43MHz as these devices are narrowband :

Yeah, I only did a quick test as a sweep around that range and expected to see it pop up. I guess it's much narrower than I expected.
Damn, I wanted to try the memory aspect of it as well but you beat me to it!
Nice work.

Dave.

It would be interesting to use a better amplifier & pulse reconstruction to see how long you could bounce the signal around, and how much bandwidth you could get.
A while ago someone made a Magnetic core memory Arduino shield  - at the time I though a mercury delay line USB stick would be fun...


Youtube channel:Taking wierd stuff apart. Very apart.
Mike's Electric Stuff: High voltage, vintage electronics etc.
Day Job: Mostly LEDs
 

Offline robrenz

  • Super Contributor
  • ***
  • Posts: 3035
  • Country: us
  • Real Machinist, Wannabe EE
Re: EEVblog #381 - How Do Ultrasonic Delay Lines Work?
« Reply #15 on: November 09, 2012, 01:49:56 pm »
Very educational Mike, loved  it!

Offline SeanB

  • Super Contributor
  • ***
  • Posts: 16272
  • Country: za
Re: EEVblog #381 - How Do Ultrasonic Delay Lines Work?
« Reply #16 on: November 09, 2012, 03:22:05 pm »
Those delay lines also need the right input and output terminating resistance. most common is 330R, and still you have a high loss, but still usable.
 

Offline Zad

  • Super Contributor
  • ***
  • Posts: 1013
  • Country: gb
    • Digital Wizardry, Analogue Alchemy, Software Sorcery
Re: EEVblog #381 - How Do Ultrasonic Delay Lines Work?
« Reply #17 on: November 10, 2012, 01:20:13 am »
Here is something that may interest you.

Many years ago my dad was a field repair techy, and covered all sorts of consumer electronics. Back then the level of training and education was seriously impressive in comparison to most of today's technicians. Much more worthy of the title Engineer. Anyway, here are a couple of delay lines he extracted from old valve / transistor era TVs.



It seems that these were Mark 2 delay lines. The Mark 1s were even bigger! The one on the left measures 90mm long, by 42 wide and 6.5mm high. The one on the right is considerably thicker at 10mm, but 39mm wide and 80mm long. Originally they would have been in plastic housings. Both are pretty weighty lumps of glass.

The reason for one being 10mm shorter than the other, and also the reason for their being out of their original plastic cases, is just visible in the centre. One of my dad's hobbies is gem cutting! Old delay lines provided a good source of high quality material to practise a variety of gem cuts. The photo really doesn't do it justice, it really sparkles and glitters.



You can just about see the moulded ripples in the side, and the moulded-in feet underneath. The piezoelectric transducers are clearly visible.

Offline NiHaoMike

  • Super Contributor
  • ***
  • Posts: 8972
  • Country: us
  • "Don't turn it on - Take it apart!"
    • Facebook Page
Re: EEVblog #381 - How Do Ultrasonic Delay Lines Work?
« Reply #18 on: November 10, 2012, 02:07:27 am »
I used to see these in old VCRs, and never quite understood their function. Cheers Dave
In VCRs they were used to conceal tape dropouts. If the signal momentarily drops out, it switches in a delayed signal from the previous line to avoid showing a black or white glitch.

The reason for the very short delay line in TV decoders is to delay the luminance signal slightly to compensate for delays in the colour processing, so the two arrive at the same time out of the decoder.
Nowadays, when digitizing tapes, maybe the delay line can be removed so the PC can more easily recognize the dropouts and use more advanced algorithms to restore the original? Or better yet, wire in a digitization system (maybe a FPGA board with some ADCs) into the VCR and then process the raw data using software.
Cryptocurrency has taught me to love math and at the same time be baffled by it.

Cryptocurrency lesson 0: Altcoins and Bitcoin are not the same thing.
 

Offline T4P

  • Super Contributor
  • ***
  • Posts: 3697
  • Country: sg
    • T4P
Re: EEVblog #381 - How Do Ultrasonic Delay Lines Work?
« Reply #19 on: November 10, 2012, 07:45:31 pm »
Here is something that may interest you.

Many years ago my dad was a field repair techy, and covered all sorts of consumer electronics. Back then the level of training and education was seriously impressive in comparison to most of today's technicians. Much more worthy of the title Engineer. Anyway, here are a couple of delay lines he extracted from old valve / transistor era TVs.
"modern tech"
Writing overly bloated code while waiting for the mega-fast processor to come out and hopefully make it work - then sell it  :-DD
That's what happens to many of the consumer electronics. Just chock full of badly written code and the codemonkeys writing it are actually called software engineers!
I call them software babboons!  ^-^
I like the look of the delay lines though
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf