Delay line memory

Followup to EEVBLOG's vid on glass ultrasonic delay lines:
• EEVblog #381 - How Do ...
Including storing 32 bits of data with a TV PAL delay line and a few HCMOS gates - memory 1950s style!

[@joshuaharlow4241]

I have heard an explanation of delay line memory before.... I "got it"... but now I really can "wrap my head around it". Thanks for taking the time to build that up and show us the concept in action. Keep it up!!!

[@elboa8]

Thanks Mike. I worked on these for years and never bothered to find out what was happening inside the "glass" block. Very interesting..

[@gilgameshismist]

I learned more in the first 5 minutes of this video than in the whole 30 minutes of rambling from the other guy.. Thanks Mike!

[@ET_AYY_LMAO]

Thanks for actually showing the delay line in use! Really amazing!

[@jjhamblett]

Thanks Mike. You seem to have a natural ability to explain how things work that anyone can follow.

[@Mark19960]

Excellent Mike. In 20 years servicing consumer electronics where these were used like VCRs and cameras I have actually seen one of these go bad.... the customer dropped their camera and did not obvious damage but they did actually manage to break a delay line.. obviously no fault of the delay line. Other than that they would probably last forever.

[@NanoCottage]

A great explanation, always wondered what they were used for and how they worked.

[@Dibblah1900]

As always, very interesting, well explained, in-depth coverage of a subject! Thank you!

[@DarkInsanePyro]

Thanks Mike for the awesome video. Really enjoyed the attempt at memory implementation at the end.

[@nathantrigg5399]

This is the 3rd video on delay line's l've watch. (The best) Thanks for that.

[@ProtectionDirect]

I have no idea what you're talking about but I've been watching your videos for the last 3 hours. Amazing stuff.

[@whitcwa]

Imagine how hard it was to build the first oscilloscopes!

[@DJignyte]

That is really cool! Thanks for taking the time to do these video's, Mike. They're quite detailed and rather interesting, keep it up!

[@CassetteMaster]

Very interesting. I have seen these in old VCRs I took apart and wondered how they worked and what the heck they were!

[@Derundurel]

I'd wondered about that too. Thanks for demonstrating and making it public.

[@latasha66]

analog error masking... pure genius.

[@ickipoo]

Another excellent exploration, Mike. Thankyou. Two thumbs up.

[@whitcwa]

The epoxy keeps the waves from spreading out too far by dampening the acoustic waves in areas they shouldn't go. I have replaced many which were used in pro video cameras to create a vertical detail circuit. They had cones etched into the glass (on both sides) where the epoxy dots are. They were also used in NTSC TVs for comb filters and in SECAM TVs for line memory.

[@souzzzzza]

It's because of things like this that i love electrical engineering. Thanks for the video.

[@shitheadjohnson2797]

adding the computer logic that is reading and writing from it, is the bit that most people dont bother with give up and get an arduino.

[@PodeCoet]

I just read about delay lines the other day, brilliant video!

[@sysmatt]

Incredible Mike! A really great demo. Thanks

[@azyfloof]

You've basically made an analogue shift register, sending the output back into the input. Bloody clever! :D Mike mike mike mike.

[@Quazzie78]

Fantastic detail and a very interesting use case! Thanks!

[@flatfingertuning727]

Have you tried using a circuit to latch and regenerate pulses? Perhaps use an edge-triggered flop clocked either by the chroma subcarrier or a divided-down version thereof, and feed that through a sequence of four NAND gates, the first two of which take input from a external logic, the third of which accepts chroma subcarrier, and the last of which is simply an inverter? It may be necessary to tweak the timing of the flop to ensure it samples each wave near the peak, but I would think it should be possible, and not overly difficult, to make the circuit so that it could circulate data indefinitely, and it would be interesting to see how many bits it could reliably store. If that worked, it might be interesting to modify the project a little more to implement something like a binary counter or other such logical function.

[@Jyaif]

This is so insanely cool

[@nbsr1]

Yup. As much as I like Dave's videos, Mike is in a different league.

[@TheCrazyInventor]

Good demonstration, thank you for taking the effort to show us. :)

[@erlendse]

Really neat, you do a nice job of showing how it works :) I just hope you take the memory idea a bit forward, possible with many delay-lines in series, including drivers e.t.c.

[@dontbe]

Please remember to tell him its located in a rectangular black plastic box that stands up off the board... you dont see the glass until you open it up!

[@mikeselectricstuff]

You'll find one in old PAL TVs, and also some VCRs

[@gamccoy]

That was a brilliant video. Well done!

[@RamLaska]

Awesome sauce! Fantastic work!!!

[@mikeselectricstuff]

Minmal effect - hard to hit it at 4.43MHz. Fingers do damp it a bit

[@aptsys]

Indeed, this is proper knowledge rather than Dave's scrambling around on the internet for buzzwords and vague descriptions enough to keep most people happy... ;-)

[@lolman2008112]

fascinating video, i now understand delay lines...cheers mike!

[@0MoTheG]

I am sure you could store much more than 64 bit using the proper modulation. Certainly no less than 128 bit assuming only a SNR of 0 dB and 2 MHz usable bandwidth. But likely more than 512 bit.

[@mikeissweet]

Phenomenal!

[@cos6255]

Nice work!

[@truhlikfredy]

MIKE MIKE MIKE MIKE MIKE xD I love it. Can' help every time yours videos are so much more interesting than Daves.

[@KostantisX]

I wonder how stable would be a clock generator based on a delay line and what would be it's tempco and if it is indeed that low if it could be used to correct the tempco of a crystal...

[@Jenny_Digital]

@mikeselectricstuff Excellent demonstration. Just one thing though... Do I detect a little bit of sport between you and Dave Jones?

[@technobabble_]

Could you tell us a bit more about the epoxy they put on the glass and what it does and how?

[@JurekOK]

in the memory case, since you are continously amplify the echo, why does it disapear after several miliseconds? i understand it could get distorted, but why disapear completely? or maybe you are not feeding it back but only reading multiple reflections?

[@aidanppp]

Good stuff Mike!

[@helicoptered]

Yeah, a few... it could be the Mi from Mike and related computer keywords used.. :)

[@aliabedi6282]

Do you know what the maximum frequency these delay lines support is? Probably we have to get rid of the long leads.

[@flubba86]

That was awesome!

[@DIYTAO]

Hi, nice explanation and demo circuilt. Do you have any schematics of the circuilt?

[@SeanBZA]

How about a spring reverb line, I have one knocking around.

[@Ju00Ls]

Very interesting! many thanks

[@wareb321]

You are a wizard!

[@Dustycircuit]

Just found guy on eBay who had a couple. Cheers

[@mikeselectricstuff]

Nope - bored with this now - why don't you try it!

[@jpsousa4]

awesome oscilloscope. what kind is that?

[@olivercoles87]

Great, thanks!

[@Tjousk]

Well explained, thanks (:

[@headsplosive]

eevblog

[@crumplezone1]

Mike have you ever thought about ruling the world ?! :)

[@AntiProtonBoy]

Early calculators also used torsion mode memory, where you have coil of wire, and you send shock waves into the loop (via a torsion force), only to emerge on the other end after some delay. See Monroe Epic 3000 calculators for an example.

[@icesoft1]

I was hoping Dave would have done this, but that seems to be more than he'd prefer to screw with. This seems very similar to the old reverb units that used actual springs to presumably get the delay and then fed back into the input through a resistor to get the decay? Is this correct?

[@douro20]

Old electronic desktop calculators often used wire delay lines, which can store more data than a glass delay line.

[@anthonyj777]

Fascinating stuff! Thanks for the video.

[@Dustycircuit]

Does anyone know a partnr (or link to a store/eBay) for one of the old delaylines Mike was mentioning in the beginning? It would be interesting to do some experiments with. Cheers and thanx for an interesting video Mike!

[@oldevil]

I like your follow up on Daves video, only too bad some viewers are bringing Dave down by saying his videos are boring too long, what are these dad people?I love seeing Daves videos

[@ExStaticBass]

Really cool lad... I would have never thought to use a TV delay line as memory like that or dreamed it would be so stable. I'm glad I subscribed a while ago. I would have missed this...

[@lordchavo]

Actually, truth be told, more interesting than Dave's version :) It's like the continuation, which holds the best part !

[@nrdesign1991]

7:50 what would happen to the data if you hit the delay line? like snapping a finger against it

[@RandomInsano2]

This is great. Thanks again Mike.

[@ZeedijkMike]

It might not have any practical use today but it's still very impressive to see it working.

[@FrankenPC]

THANK YOU! I was just wondering about interference after seeing EEVBLOG's video.

[@1ManWrenching]

Absolutely epic demonstration!

[@HDXFH]

Cool I got a few of them!!

[@jix177]

Very good follow up. Thanks for sharing!

[@aptsys]

Was bored. Now not.

[@jjoster]

bravo

[@helicoptered]

Who is dave userbame please?

[@robstorms]

Fantastic Mike !!!

[@SuperHddf]

fascinating :::)

[@paulbendel]

Cool!

[@jabramow]

Amazing