A simple delay circuit

  • Thread starter Thread starter brendandwyer
  • Start date Start date
brendandwyer

brendandwyer

New member
I'd like to get some information on how to build the simplest delay circuit imaginable.

I want a box (this might turn into an 8 channel rack mount module) that has 1 1/4" input jack and a 1/4" output jack. The box will have one knob on it which will adjust the delay time in ms to create a delay between input and output. Maybe if i'm feeling really crazy we could mult the signal and do a wet dry, but for right now, as my first foray into electronics (short of building my own cables, and potentially electrocuting myself, see my thread on dual inputs to single output).

This would be either used on an insert to delay a channel in line, or on a multed channel to create an effect.

Comments?
 
haha. great. i'm an electronics dunce enough as it is. :)
 
Brendan,

> The box will have one knob on it which will adjust the delay time in ms to create a delay between input and output. <

Why do you think this is a "simple" circuit to design and build? :D

Seriously, this is not a simple circuit to design.

--Ethan
 
hmmm. dunno. cause i have no idea what i'm talking about i reckon :)
 
Ethan, so are we talking about a circuit that is difficult to build for a guy like me, but easy for a guy like you? Or is it expensive and difficult for anybody?
 
Ok lets pretend i didn't call it simple at all and instead titled it "a delay circuit". I had conversed at one point with owel from tweaks about the possibility of him building it, but unfortunately he was far too busy. So i guess i'm enlisting someones help in working up a cost/price for this thing. if no one responds, i'll consider it a dead issue....
 
i think the problem is the signal has to be stored for a peroid of time..
The easiest diy circuits seem to be distortion/fuzz type things. There are loads of diy guitasr pedal website's with projects varying in difficulty :)
 
Brendan, considering the amount of time and effort that would go into designing and building a one-off box of who knows what kind of specs, I would imagine it would be cheaper and better to just buy one.

G.
 
Brendan,

> so are we talking about a circuit that is difficult to build for a guy like me, but easy for a guy like you? Or is it expensive and difficult for anybody? <

ROF,L. Well, I'm sure it's easier for me than for you, but it's not a simple task for anyone. Do you understand what's involved? The main problem is you need to store the audio somewhere during the delay period. So this means you need either a mechanical tape or digital memory. You could buy a three-head cassette recorder, but that won't let you vary the delay time. So now you need an A/D convertor, a bunch of computer RAM, a D/A convertor at the output, and a fair amount of other circuitry and programming to make it all happen.

--Ethan
 
cripes. I decidedly do NOT understand whats involved. I thought that it could be accomplished without a recording medium or digital means. I guess i thought that with electrical components, you *stored the current for a period of time and then it resumed it's path. thanks for the response though
 
Well you do store it for a period of time but that's the problem. In what medium do you store it?

It's exactly the same as recording except that you want to store it for a shorter period of time and you don't want the storage to be permanent.
So basically a delay unit is a specialized type of audio recorder.
 
Well ... here's a quick schematic description of way to make a unit that will generate a short delay (adjustable up to 20 or 30 ms without too much trouble) and it's simple ... it doesn't work tremendously well though:

1- Connect input to a reasonably powerful audio power amplifier.
2 - Connect amplifier to fairly large speaker.
3 - Place speaker in a reasonably dead space(1), or maybe outside someplace that's reasonably quiet (or that was reasonably quiet until you started fooling around with this thing)
4 - Place microphone some distance away from speaker.(2)
5 - Connect microphone to microphone preamp.
6 - Output of microphone preamp is your delayed signal.

To make an 8-channel unit would take up a bit of space, though. More than a few rack spaces, I'm afraid.

The obvious (to me, anyway) devices that have been used to create delays (in historical order):

- Tape decks. Any garden-variety three-head reel-to-reel tape deck can be used to create a short single delay by recording, and taking the output of the playback head. You can, of course, make a repeating delay by mixing the output back into the input. You can vary the length of the delay by varying the tape speed, to the extent your unit lets you do this. Or (more difficult) you can rig something up to move the heads relative to one another. Or you can use two recorders, if you can rig up a system for running the tape from one to the other and keeping appropriate tension.

- Specially-built tape delay units. They do the same thing, but somebody built the tape deck into a box, with a loop of tape so you don't need to fool with it, and provided you with a simple method for varying the tape speed or the distance between the heads, or both.

- Bucket-brigade delay. Basically, you sample the analog signal, and pass the samples through a series of capacitors, each of which "holds" it briefly, before passing it to the next. Theoretically, I guess you could do this with discrete components, but practically it's done with a chip. This is what's been in your Boss stompboxes since the '80s or so.
http://www.web-ee.com/Electronic-Projects/data/mn3004.pdf

- Full-on DSP. Make your analog signal digital and store it in RAM (in a computer or whatever). Play it back whenever you feel like it.

----------
(1) Or, if you're so inclined, put it in a space that isn't very dead, and you'll create an analog reverb unit. If it's underneath Capitol Records, you can make famous records with it.
(2) Changing distance varies delay time (as does temperature, humidity and altitude/barometric pressure). In ordinary conditions, 1 foot = a little more than a millisecond. Figure 10-3/4" per millisecond, and you'll be close, if not precise.
 
Sj, reamping i get, but that wasn't what i was asking for.

It sure would be great if i could build an 8 channel box that had 8 knobs that controlled delay. And it sure would be great to have a frozen daiquiri right now. **ping** <img src='http://www.gotopuertorico.com/parames/mesones/images/daiquiri.jpg'>

ok so now wheres my 8 channel delay box :)
 
well other than the obvious, it would be used for things like delaying a multed guitar line, or vocals. more of a project than to solve a problem though
 
How long a delay? How about a variable delay line or a hundred of them in series? :D

Alternately, you should be able to design a relatively crude digital delay line with the following:

1. ADC (parallel)
2. DAC (parallel)
3. A static RAM chip of appropriate size---if it's 4-bit SRAM, you'd want 6 for 24-bit audio, 4 for 16-bit.
4. A clock crystal.
5. Two counters wired to wrap to zero at the appropriate point.
6. A counter wired to count to ten or so, then stop.
7. A bunch of small delay lines.
8. A ton of switching transistors.

Here's how you do it (I think):

A. Counter #3 (counting to ten) starts on power on. Upon reaching ten, wire it to close the circuit between the clock source and the other two counters. Each time this clock increments, it should increment the read counter so that it is (mem_size - 10) addresses behind the write pointer.
B. The clock should go to a pair of transistors. One is wired to the base of a transistor between a +5v signal and an output. We'll call that the "clock high" clock output. The second is wired between a separate, resistor-isolated +5v signal and ground. The +5 volt line will get grounded when the clock is high, so that contact will go high when the clock is low. We'll call that output the "clock low" clock output.
C. The clock high output will go four places: the counter #1 trigger input, the clock input on the ADC, the "write address gate" and the "transition to high" delay line.
D. The clock low output will go three places: the counter #2 trigger input, the clock input on the DAC, the "read address/data gates" and to the "transition to low" delay line.
E. The "write address gate" is a series of parallel transistors that control whether the output of counter #1 goes to the address lines on the RAM chip. The "write data inverter" provides inverted data bits from the ADC to write to static RAM. This is a passive component that should need no clock.
F. The "read address gate" is a series of parallel transistors that control whether the output of counter #2 goes to the address lines on the RAM chip. The "read data gate" brings both inverted and non-inverted data lines to +5v when triggered.
G. The "transition to high" delay line will trigger a write operation. Details vary depending on the design of the SRAM. It will also feed another delay line that will disable this operation before the data goes away.
H. The "transition to low" delay line will trigger a read operation. It will also feed another delay line that will trigger the input latch on the DAC while the data is available, and another delay line to pull this line low before the data goes away.

Anybody brave enough to try it? :D

You know, in hindsight, it would probably be a lot easier with a microcontroller. :D
 
You must log in or register to reply here.
Back
Top