Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Constant signal delay circuit

Status
Not open for further replies.

Robsta

New Member
Hi there!

I'm new here and, to be honest, a bit out of touch with circuit building- about 10 years out of touch.

To that end i am wanting some help (a simple circuit diagram or instructions should suffice) with a project i am building.

I need to design a circuit that will apply a delay to an RF signal. This delay could be variable but this is not essential, but the ideal delay required would be 30 seconds. The circuit needs to maintain the input signal quality as much as possible, or self-compensate for any loss created, as this will be used in proffesional video applications. The idea in practice will be that i can externally split an RF feed and run one half through this device, with the output signal constantly (and uninterupted) 30 seconds behind the untouched feed.

I believe the circuit will be a fairly simple one, and possible based around the 555 timer? As i say, i am a little out of touch and might be over-complicating things or even over-simplifying them!

Hope this makes sense, and look forward to hearing any advice and suggestions.

Cheers
 
It's certainly possible (depending on the frequency) but it won't be cheap or simple.

The signal needs to be converted to a digital representation which can be temporarilly stored in a memory device which as a shift register or RAM.

What frequency is the RF signal?

The problem is, it will take a huge amount of memory to store 30s of an RF signal. The sample rate needs to be at least double the maximum frequency of interest and you need enough bits per sample to store all the data.

To put this into perspective, assuming the signal is 100MHz and 8-bit/sample you'll need 6GB of memory.
 
I believe the circuit will be a fairly simple one, and possible based around the 555 timer? As i say, i am a little out of touch and might be over-complicating things or even over-simplifying them!

Hope this makes sense, and look forward to hearing any advice and suggestions.

Cheers

hi,
Its not simple, the easiest way is with a PC with an TV/RF input pcb and record the RF data modulation onto a hard drive then remodulate an RF carrier with the delayed modulation signal.

When you say RF, whats the frequency spec etc, whats the signal.?
 
Many thanks for your reply!

I haven't got all the facts and figures at this spoint because it is in very early stages- too higher cost may even make it a non starter.

Could the same be achieved without storing the signal to RAM, by means of a resistance path? Depending on the signal drop-off this may be acceptable as the input signal should be fairly strong. Delay time could be dropped to 20 seconds or even 10 seconds at this stage if this helps.

I'm sorry for sounding vague- i think where i am at this stage is trying to build a prototype to demonstrate a concept, which could then be further developed if the theory worked.

Many thanks again for your help
 
an alternative option would be to use the device at the oposite end of the equipment chain, once the signals have been processed. At this point it would be accepting a combined AV feed from a SCART or HDMI (or both) connection, and need to output via the same.
 
Such devices already exist, they are HDD recorders - but most are true PVR's (and only record direct digital signals) - so you would need an 'analogue' one, which probably records to DVD as well as HDD.

There's nothing simple, it's a VERY complicated thing to do, and has only fairly recently become reasonably cheap to do.
 
Could the same be achieved without storing the signal to RAM, by means of a resistance path? Depending on the signal drop-off this may be acceptable as the input signal should be fairly strong. Delay time could be dropped to 20 seconds or even 10 seconds at this stage if this helps.
There's no practical way to delay an analog RF signal for that length of time. Typical analog delay lines provide microseconds, not seconds of delay. Any usable scheme would require converting the data to digital, storing it in memory (solid-state or hard drive), reading it out after the desired delay, and then converting it back to analog.

One analog scheme they used years ago for millisecond delays (such as for the rear speakers at the Hollywood Bowl) was to record the data on tape and then read the tape at a different point on the tape for the delay. But storing seconds of high frequency data would require a huge distance between the record head and the readout head.
 
hi,
Its not simple, the easiest way is with a PC with an TV/RF input pcb and record the RF data modulation onto a hard drive then remodulate an RF carrier with the delayed modulation signal.

Of course, you heterodyne it so you only sample the bandwidth of they signal you're interested in.

I know, you could bounce the signal off the moon, or up into space, off a few satellites around the world a few times and back again.:D
 
Of course, you heterodyne it so you only sample the bandwidth of they signal you're interested in.

I know, you could bounce the signal off the moon, or up into space, off a few satellites around the world a few times and back again.:D

Im sure if the OP knew how to hetrodyne the signal or bounce it off a few satellites a few times he wouldnt need to ask the question on a forum.:p

The technology already exists if he wants to spend a few a pounds, but I would like to see someone do it with a 555.:rolleyes:
 
Last edited:
New Question!

Hmmmm.....

Right then, i've had a rethink.

Is it possible to alter the transfer rate of a realtime digital signal without buffering? As an example, could an effect be achieved similar to playing a DVD, putting it into slow-motion for 30 seconds, then resuming original speed?
This might sound bizarre, but i basically need to create this delay without physically storing the information (ie it must stay constantly moving), so buffering to RAM would, i believe, be out of the question (even though it would only be stored on a temporary basis).

The signal would be a processed AV feed over HDMI.

I apologise for my ignorance- my electronics experience was limited, up to now, to audio effects circuits.

Many thanks again for all your replies.
 
Last edited:
As an example, could an effect be achieved similar to playing a DVD, putting it into slow-motion for 30 seconds, the resuming original speed? This might sound bizarre, but i basically need to create this delay without physically storing the information, so buffering to RAM would, i believe, be out of the question (even though it would only be stored on a temporary basis).
That's not possible, in order to create a delay, you need to temporarily store information.
 
Damn!

Right, last question then before i give up on this project!

Would it be possible- in theory- to take the input signal and constantly transfer it back and forth between two RAM sets for the required time, at the original transfer rate, then release it to the output stage after the designated delay period? Or even send the data round in a continuous loop for the required delay period?

I know this is getting more and more complex, but if it is possible i believe it is worth pursuing.

Many thanks
 
Whatever you do, you can't get away from the fact that you need to store the amount of information transmitted before the delay. For example, if you want to delay a video signal be 1 second, you'll always need to store 1 second of information, whether it be in RAM, as radio waves travelling through space or on a cassette loop.

Slowing down a signal, requires even more storage space because as the receiver gets further behind the transmitter more and more information needs to be stored.

EDIT:
The easiest way of doing this is to use a computer and software.
 
Last edited:
ok. well, thats sorted that then!

Well, im glad i asked the questions before i put too much time, effort and expense into it!

Thank you all for your advice.
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top