DiY Tape Delay?

[Jeremy Garber]

Ya, I think you guys do have a point about the break out box. Doing something like that wouldn't bother me at all. That's good advice too- allowing yourself to be flexible and dynamic with ideas.

What was it about the wiring that gave it so much static? Really long wires? Input too close to output? Audio wires crossing power supply wires?

On the plus side I'm almost done with 'Electronic Components'. Just a few more pages, then I'm going to start 'Electronics: Self-Teaching Guide'. The latter goes into depth about setting up certain components, and how to build amp stages and filters and what not. Pretty much taking the last book a bit further. I almost picked up 'Electronics For Inventors', but I was looking for a slightly thinner read.

[nclayton]

I've got to be honest....I think the idea of making a tape delay where the delay part is done with an IC seems kind of crazy. First of all, making an electronic delay is not a super simple thing, and I don't think using one will make the mechanical portion of the job THAT much simpler. In other words, I think using a chip will make the job bigger and harder and more complicated, and you'll end up with something that won't even be a true tape delay, just a digital delay add tape hiss.

If you want to go that route I'd say just get a three head tape deck, leave it in monitor mode and buy a used Boss delay pedal and cascade them. Don't try to DIY it. It won't be worth your time.

On the other hand, I think you CAN make your own true tape delay, I just don't think you can do it if you keep the tape in the cassette. But that's good! Just let yourself go ahead and make a loop. It opens up many more possibilities since you can make the loop as long as you want. Also, you won't have to restrict yourself to finding a 3 head cassette deck, which is totally not suitable for what you want to do anyway. You can just use parts from a couple normal cassette recorders.

Just get 2 cassette recorder transports, mount the parts you need to a piece of aluminum and let the tape loop run between them. Make spring loaded pressure pads that you can pull back with your finger to thread the loop and then, when you let go, permanently press the tape against the heads. Don't worry about disengaging the heads or stopping the tape when the effect isn't on. That stuff will make what you're doing too complicated and too big. Have you disassembled a logic controlled cassette deck? I have. The transport is pretty big! Most three head decks were expensive and had big flywheels and many had 3 motors. On the other hand, the totally mechanical portable decks with 1 motor, crappy rim drives and no flywheels at all have way smaller transports, and if you don't worry about engaging or stopping a tape loop you can probably throw a lot of it away. Plus, if you make a good little head sub-assembly with a head on one side and a sprung pressure pad on the other, you can put it wherever you want and it will work. You can slide it around, or add additional heads, or whatever, and you don't need to add more capstans or keep the head aligned to any mechanical engage/disengage stuff, just screw it to your panel wherever you want. Just let the capstan and roller be after the last head and you'll be fine.

Also keeping it simple will make it more reliable. True, pressure pads will wear out the tape and the heads a LITTLE faster, but I don't think having a great big logic board jammed in some kind of box is going to be what you want either. Just have 2 switches, one to engage the audio, one to start and stop the motor. That makes it cooler anyway, because you can start and stop the motor with the audio running to get speed up/slow down sounds.

That's my opinion.

Ned

[Skipwave]

That last post got me thinking. Two walkman transports mounted on a panel. Cut a slot behind the second (monitor) head, to serve as a channel that it can slide horizontally in. Attach that to slider and you have a variable delay time.

Then add a roller mounted in a vertical channel, spring loaded to maintain tension on your fixed-length tape loop.

Looky, I drawed it on my 'puter:

http://twin-x.com/groupdiy/displayimage ... m=56&pos=1

The cassette outlines are just there so you know what those little blobs are. You might be able to use two cassette shells with the unused side cut off.

Crazy idea?

[neaimepas]

at the time i didn't really know much about electronics in wiring. but for some reason i ended up using a bunch of these really thin rca cables, running right over the power line. Plus i was using trying to use 10 year old pre-written tape

[Teacher's Pet]

I think the tape guts of Hi-Fi VCR's should be considered for this. It's easy to get your hands on one (or more) of these. The half-inch tape is bigger and probably sturdier than eighth-inch cassette tape, and the two-track audio is also superior to regular cassette.

I've been experimenting a little with these and I will try to post some video... It's surprising to see what goes on inside one of these, if you've never peeked under the hood before.

[newfuturevintage]

It's surprising to see what goes on inside one of these, if you've never peeked under the hood before.

The word 'terrifying' as opposed to 'surprising' comes to mind

Question: are the audio channels linear or helical?

[Teacher's Pet]

If the machine is a "hi-fi" VCR, and made after the mid-eighties, I believe the 2 audio tracks are helical. They did a thing where the audio tracks are beneath the video signal on the tape.
It's 2 tracks of high-quality analog audio. Don't let the ADAT association freak you out. It seems like it might be useful in this context because:
- larger tape, easier to work with if you're creating big, weird loops.
- the cassettes are designed to flip open and release a length of tape for playback/record.
- the tapes last for 2 hours or more
- the crappy machine I've been messing with claims "4 heads." Maybe this would make it easier to break apart the record/repro assemblies if you want more space between the 2?

The Wikipedia entry on VHS explains the technology pretty well.

I've been recording band practice on one of these recently, with pleasing results. When I showed it to another guy whose band shares our space, he said "We used to do that for all our old 4-track stuff (mixdown)... That is the best 2-track audio you have in your house."

[Teacher's Pet]

I uploaded a couple of videos to Youtube. Sorry about the poor quality, I'm still figuring this stuff out.

Click these if you want to see what happens inside of a VCR:

http://www.youtube.com/watch?v=aJLqzzV183c

http://www.youtube.com/watch?v=QSb6aYQHYiU

Is it possible/allowed to embed Youtube clips in message board posts?

__

ALSO:
Re: newfuturevintage's earlier question, I think many decks give you the option of using the linear track along the edge of the tape, or the helical track embedded beneath the video signal. My understanding is that the helical one is much higher fidelity. Anybody else care to correct me or explain this better? I am no expert.

[E.Bennett]

if there is a 3 head multitrack cassette machine, there might be an easier solution. i think i did something like this on a 4 track reel to reel machine years ago...

send aux signal into input of trk1 of multitrack
split output of trk 1 repro head to ch. 1 of external mini mixer and input of track 2 of multitrack
split output of trk 2 repro head to ch. 2 of external mini mixer and input of track 3 of multitrack
split output of trk 3 repro head to ch. 3 of external mini mixer and input of track 4 of multitrack
send output of trk 4 repro head to ch. 4 of external mini mixer
output of mini mixer will carry 4 equally spaced repeats of original signal

each delay will be of equal time because it's the same two heads being used to cascade the signal. change delay time by adjusting pitch control. it's been a long time, so i might be forgetting something, but the general idea is there.

[fucdemas]

i love this thread. i've been thinking about doing this for a year or so.

so what's the ebay verdict as far as finding 3 headed pro-cassette machines and just creating a simple tape loop and maybe wiring up the tape speed pot to an external expression pedal?

has anyone found an existing unit that can provide the requisite monitoring needed?

[Jeremy Garber]

nclayton's post has brought a question to my mind. How different is it really to have the delay time controlled by an IC as opposed to a complete analog circuit as per the Echoplex? nclayton says using an IC chip will just be a digital delay add tape hiss. But honestly, how is it different? Do you think you would really hear a difference? Also, I was wanting to keep the actual delay circuit all analog. I really had no intention digitizing the signal for the IC (since digital circuits are beyond my scope of understanding, aside from the simple switches that are used as building blocks), but perhaps I have a misunderstanding of the available chips for such a purpose. I haven't done any additional research for this project because I've been consumed in other parts of my life, including salvaging lots of parts from a bunch of PCs that were donated to me for such a purpose.

In addition to the PCs, I received many printers which have given me step motors and gears- something that could possibly be used towards making my own transport.

And nclayton, I have yet to fully explore the transports for electronic engaging systems but I'm not that intimidated. I have a great desire to understand these things, and I do want to dabble in mechanical engineering at some point- starting this Christmas when I buy my 4yr old some LEGO Mindstorm sets. I had both my 4 and 2 year olds helping me take apart the PCs tonight. They were both really great at unscrewing and sorting the parts.

Anyways, on to more great ideas...

[The Scum]

How different is it really to have the delay time controlled by an IC as opposed to a complete analog circuit as per the Echoplex? nclayton says using an IC chip will just be a digital delay add tape hiss. But honestly, how is it different? Do you think you would really hear a difference? Also, I was wanting to keep the actual delay circuit all analog. I really had no intention digitizing the signal for the IC

So which IC will you be using?

If you want a simple, one chip (or two) delay, you're looking at a digital one...like the chips from Holtek or Princeton.

There are also bucket brigade chips, which are used in "analog" delays. They're fragile, and require a lot of extra parts to make something really useful out of. They need a carefully designed clock circuit, and certainly aren't hurt if you add encode/decode noise reduction circuitry around them.

They're effectively a very long string of analog sample and hold circuits, tied end to end. Those S&H's still need a sampling clock, so they still employ the Nyquist Fs/2 bandwidth theory, as used in digital sampling. The S&H's are also prone to drift or leak, so the signal output by them may not resemble the input particularly well.

If you want a delay that gets darker and more distorted as the delay times get longer, BBDs are interesting...a number of classic delays use them. But they act nothing like a tape delay. Use each as a tool, for its own merits.

[Teacher's Pet]

How about replacing the print head on one of those old printers with a tape head? It would zip along a page-wide strip of tape, either depositing or retrieving a signal.

Maybe there would be a way to control delay/playback time by "printing" different patterns, faster or slower, etc.

Actually, what I'm describing is probably more like some kind of robo-mellotron. You gotta have dreams.

[Jeremy Garber]

So which IC will you be using?

That I am not sure yet. I haven't looked into what is available. I was under the assumption that there is such a thing as an analog delay (sans tape).

If you want a simple, one chip (or two) delay, you're looking at a digital one...like the chips from Holtek or Princeton.

Ah, so this is probably more of what I had in mind- a simple one chip solution. However, I'm more than willing to dip into the archaic methods of a true analog delay, if for nothing else but a learning experience. I looked at a couple ICs offered by Holtek and Princeton.

There are also bucket brigade chips, which are used in "analog" delays.

Call me crazy, but that sounds fun. I wish I had more time this morning to read up on bucket brigade delays. (throws fist in direction of work)

If you want a delay that gets darker and more distorted as the delay times get longer, BBDs are interesting...a number of classic delays use them. But they act nothing like a tape delay. Use each as a tool, for its own merits.

Understood completely. So maybe I'm not making a true tape delay as recognized by purists. I kind of like the idea of a destructive reproduction of the original. That was pretty much my initial desire to do this project.

[The Scum]

That I am not sure yet. I haven't looked into what is available. I was under the assumption that there is such a thing as an analog delay (sans tape).

The BBD is as close as it comes, I think. And they're finicky.

There's aso the charge-coupled-device (CCD), like digital cameras use, which you can feed an analog signal, and get it back some time later...but the only audio device I know of that ever did that was the Marshall Time Modulator. And Steve Marshall had a very serious "mad scientist" streak (and I mean this in the most flattering way possible...he saw solutions the rest of us just don't grok.).

Understood completely. So maybe I'm not making a true tape delay as recognized by purists. I kind of like the idea of a destructive reproduction of the original. That was pretty much my initial desire to do this project.

If that's all you want, take a step back, and stop worrying about analog/digital/tape.

If you already have a delay unit, try experimenting with stuff in the feedback path. Rather than turning up the feedback control on the unit, turn it down, and return it to an input channel on the console.

Then turn up the send to the delay on that channel, using it to replace the feedback on the delay unit.

Then try messing with stuff on that channel - turn the lows way up, or insert a distortion pedal, or wah, or noise gate, or ring modulator...

There are war-of-the-worlds sounds that are easy to get into with this sort of patch. And it works with any core delay technology.