Lo Fi telephone effect w/feedback looper,will this work?

[Brymus]

As I was recycling stuff yesterday I was removing the spkr/and mic from and old telephone handle.
The thought occured to me to use a LM386 to power the spkr then pick up the sound in the mic and use another LM386 to power it.
Here is the basic concept minus the power supply and rest of the components needed.

I was just curious if this is an idea that would work or just ridiculous?

I was thinking the spkr and mic would be mounted in a small tube or such inside the enclosure perhaps with a lite piece of upholstery foam between them.The pot would control the amount of feedback and cap selection could decide the feedback frequency.
Or even add jacks and add another FX into the feedback loop,like a fuzz or delay,ect.

Perhaps even omitting the feedback loop if it oscillates?

[deadastronaut]

i had a similar thought about getting  feedback...but never got further than thinking...

the thought was splitting the signal from guitar to amp...and another to your kind of setup...

ive done similar using a small battery marshall amp and it worked really well, but i had to put up with a screaming little amp, but the split signal

was recorded really nice...

worth experimenting with imo...

[samhay]

Cool idea. Might work better (be quieter at least, and won't suffer from the questionable frequency response of the mic) if you took the output from the first LM386 though.

[Brymus]

Cool idea. Might work better (be quieter at least, and won't suffer from the questionable frequency response of the mic) if you took the output from the first LM386 though.

That's a really good idea 
I appreciate the feedback (and even welcome critics too).

It could be an option with a switch,my thought was that the telephone spkr and mic would make it sound like you were listening (to whatever is being played) over the phone.You know the really LoFi sound.

Wait it would still have the LoFi sound but wouldn't be degraded twice,nice!

Any other thoughts?
I found a circuit that uses the LM386 to drive a reverb pan.That might be cool to add in the feedback loop too.

[GGBB]

I could be completely off about this, but I have a vague recollection from a college prof who used to work in the telco field that the "lo-fi" sound was not induced by the mic/speaker but instead deliberate filtering in order to avoid distortion at the ends of the spectrum.  If true, the mic and speaker might not be enough to get you the lo-fi sound part of the equation.  Just some food for thought.

[tca]

Here is a similar project from the **The DIYStompboxes 10-Year Anniversary Contest Thread**.

Cheers.

P.S.
Corrected link.

[thelonious]

Here is a similar project from the **The DIYStompboxes 10-Year Anniversary Contest Thread**.

That is exactly what I was thinking about when I saw this thread pop up. That project was epic.

a college prof who used to work in the telco field that the "lo-fi" sound was not induced by the mic/speaker but instead deliberate filtering in order to avoid distortion at the ends of the spectrum.

Yeah, I think you should look into companders (compressor/expanders - see http://www.diystompboxes.com/smfforum/index.php?topic=44194.0) and also heavy bandpass filtering. If you chop off most of the high and low frequencies and then compress the crap out of it, that should get you in the ballpark.

On the other hand, using the speaker/mic technique as in the schematic you posted might get you 85% of the way there with a lot fewer parts... especially if you lowpass filter it, too.

[Brymus]

Here is a similar project from the **The DIYStompboxes 10-Year Anniversary Contest Thread**.

Cheers.

P.S.
Corrected link.

Holy cow! That is really close,the only difference is the lack of feedback loop and he used the 386 at max gain for a distorted sound.
I was aiming at a clean sound, just LoFi telephone sound.

But yeah ,thank you for the link !!! That should really help,It just gave me the idea to have two mics + two spkrs, one with feedback one without and a blend option.IDK

But that confirms it will work!
Now I just need to breadboard and test some configurations.

@ Thelonious,I was thinking more of a band pass to emphasize the mids,or a low and high pass filter if the spkrs don't filter it enough by themselves.

[shawnee]

Another mechanical way might be to, instead of using a mic, use a piezo on some cake tin aluminium foil. That way you have a built in low pass, with the piezo and could even mount the alumium on springs for reverb.

And if I remember rightly there's a telephone mimic circuit at geofex somewheres. I think it even had a white noise maker to add hiss.

[shawnee]

http://www.geofex.com/FX_images/lofi.gif

[Brymus]

http://www.geofex.com/FX_images/lofi.gif

Yeah I knew about RG's design but hadn't looked at it since he first posted it.

I was just thinking of what to do with the spkr and mic I pulled yesterday,when this idea hit me.
Something simple and useful maybe,I'm not sure yet.
The main issue that worried me was making the feedback controllable,or just useful without full on oscillation occuring.

[PRR]

> a college prof

Hmmmmm.

> the "lo-fi" sound was not induced by the mic/speaker but instead deliberate filtering in order to avoid distortion at the ends of the spectrum.  If true, the mic and speaker might not be enough to get you the lo-fi sound part of the equation.  Just some food for thought.

From the dawn through the 1930s, the receiver (earpiece) was just plain awful. Narrow resonance at 1KC and tiny peaks above. An improved type got some bass and a little treble, 600cps-2KC. The 500 set got a good 300Hz-3KHz with a reasonable presence boost.

Carbon mikes varied a lot, but were often better than the receivers.

I don't recall much deliberate effort to restrict frequency response; it was hard enough getting the little you got.

Talking-movie systems were much more likely to delibrately filter to avoid nastiness in reproduction systems.

While telephone sets were improving, there was a radical change in "wires". Simple open wires have good frequency response but declining with distance and frequency. "Loading" greatly improves transmission up to 2KC or 3KC, clobbers anything higher. Even fat wire and heavy loading wasn't enough to get across the USA. Amplifiers were developed. But now that you have electronics in the path, it starts to make sense (took a lot of work) to put one conversation at 0-4KC, another at 4KC-8KC, another 8KC-12KC. As amplifiers improved they went for 50KC, *twelve* conversations on one wire pair. Without that, only Big Business Men could afford Long Distance; with stacking, normal people could afford to call for Mothers Day, Xmas, or to coordinate visits.

The result of this is incredibly steep filtering before 4KHz. (3KHz for cross-Atlantic cables.)

However.... ring armature receivers have not been made in decades. I don't know what "old" means in the Mohave. Since the end of company-owned telephones, nearly ALL telephones use modified speaker/headphone drivers. These are potentially flat, but there are advantages in power and clarity if the curve is bumped-around a bit; and no small cheap driver is ever really flat.

Try it. It will mangle sound some.

If not mangled enough, use a paper-towel tube.

Yes, it will feed-back, and more abruptly than a PA system in a room.

If you have a sick garden-hose and some small funnels, you can build a delay line. These were never popular because they color the sound with driver and hose resonances. (Cooper 'Time Cube' came closest to common.) Yes you can apply feedback and increase the color and tail-time just before it squeals uncontrollably.

[Strategy]

Not quite the same but similar inspiration. Friend of mine used to buy kids' toy walkie talkies and solder cables to them. Not sure where he was kludging into the circuit. The result was the coolest gated lo fi distortion ever. Like boxy and middy in sound, but totally controllable because of the gating. He never figured out bypass for it, there's probably a way to box it all up as a pedal. Your phone hack reminded me of this

Strategy

[Brymus]

>
However.... ring armature receivers have not been made in decades. I don't know what "old" means in the Mohave. Since the end of company-owned telephones, nearly ALL telephones use modified speaker/headphone drivers. These are potentially flat, but there are advantages in power and clarity if the curve is bumped-around a bit; and no small cheap driver is ever really flat.

Try it. It will mangle sound some.

If not mangled enough, use a paper-towel tube.

Yes, it will feed-back, and more abruptly than a PA system in a room.

If you have a sick garden-hose and some small funnels, you can build a delay line. These were never popular because they color the sound with driver and hose resonances. (Cooper 'Time Cube' came closest to common.) Yes you can apply feedback and increase the color and tail-time just before it squeals uncontrollably.

Wow Thanks PRR,
I spent a few hours perusing "time cube" on the web,that's some cool stuff.
Cooper used the transducers from SM57 mics and had a stereo option at 14ms,and 16ms.Or series of the 2 channels for a 30ms delay.
It seems you get about .86ms-1ms of delay per foot of tubing IIRC

And the handset wasn't a "Ma Bell" version, I wish.
I miss those indestructable phones actually.
It was from a 15+ yr old fax/copier.

So now I have about 20' of clear PVC tubing (still need a 3 way fitting for filtering) and dug out what's left my little cheap spkrs and breadboard,this I have to try it's too simple not to.

Sadly when I moved I threw out a decent sized box of pulled spkrs I had used for my LM386 musings years ago,it also had several small electret and condensor mics in it too
You always find a "need" for stuff after you throw it away...

And thanks Strategy,yeah same inspiration.I had hoped I discovered something new...
At least it is all new to me,and I have the work of others to aid me in my quest.

[PRR]

> It seems you get about .86ms-1ms of delay per foot of tubing IIRC

Sound travels 1 foot in 1mS.

(Actually closer to 1,130 feet per second... ah, that's where you got 0.86mS/ft. But close-enuff for any practical purpose.)

And light travels 1 foot per nanoSecond. (That's how my father explained why computers would only ever go so-fast. You couldn't cram a useful computer into much less than 10 feet as the wires run, so a hundred MHz was tops, and even that was fanciful. He also sent the new-boys out for a One Farad capacitor.)

[Brymus]

PRR
Unknowingly sent me on a quest/inspired me.
I cobbled up a DIY "Timecube" last night,the first results were better than I expected after reading other DIY attempts that were never finished.
I will start another thread for it after I get some pics and some more progress.As I still plan to do the original LoFi telephone feedback effect too.
I actually incorporated both together last night.
The exception is that it wasn't LoFi like I expected,the drivers I used actually sound fairly full range through the tubing.
They sound better mounted in the tubing that they do used as intended/freestanding actually.

I believe I have 25' of tubing,it's the deluxe "aquarium cleaner/water bad filler" I bought from Wal-Mart 10 years ago for 30$

Anyway I used a "bullet" usb rechargeable self powered spkr for I-phone/MP3 use,it's input driven by a LM386.
The Bullet fits right into the large tube you vac/clean your aquarium with.
I used the tele earpiece (180Ohm) as the driver at the other end feeding a second 386 then onto the output and feedback loop.
It (earpiece spkr) actually attached to the fitting that hooks to your hose spicket really nice.
Pretty cool delay,sounds like your playing in a large cistern or in a well.
The feedback circuit needs to be filtered just right,I like it when the sustaining notes bloom into self oscillation,that cancels when you hit the next one.
With it set right you can adjust the feedback with the guitar volume.

Problems I need to fix:
*I need to isolate the tubing and drivers,the delay is very noticable compared to the sound coming from the back side of the amped up   bullet.

*There was a low frequency resonance or feedback,may be caused by the mic picking up vibrations from the first driver as it was laying on my table with both drivers in fairly close proximity.Or it could be an issue with the 386's oscillating from too low a gain(pins 1+8 open) or the tubing picking up vibrations.

* If the resonance is from the tubing I need to use a high pass filter to remove the "rumble"
(what I read from the "Cooper TimeCube" is that it boosted the high end a lot on the input)

*I need to get the filter for the feedback loop dialed in so its not so hard to adjust between usable/musical and run away oscillation/squeal.
I am thinking somewhere in the mid/upper-middle  /lower- high frequencies.( I used a couple of different low value caps last night)
(1n and 4n7 IIRC)

*I need to route a Dry/Wet mixer so you can hear the delayed signal over the clean one.I did a quick one using a pot last night I think a mixing op amp would be better IDK yet.

* I still need to add an FX loop to the feedback loop so the feedback can be affected with "whatever" pedal I choose.
That's it for now more testing and some sound clips to come when I get time 

In another forum someone was building one and quit because of the sound quality,a poster suggested shifting the input up several octaves then shifting the output back down several octaves.(cool idea)

This made me think if you used that approach, you can make the physical size much smaller.You could use fishtank air bubble tubing and two small electret mics.Mount it all in a cheap kids metal lunchbox then fill the whole thing with that expanding foam for damping.Even the electronics and jacks would fit so it was all self contained.40'-50' of that small diameter tubing and several electrets placed at different points would fit easily and you could then switch between delay times or run a improvised stereo effect.(which the Timecube did as well)

All in all I think this is a very cool mechanical/electrical FX with lots of un tapped potential.

[Brymus]

Another idea just hit me:
I could take 3" tubing And suspend it from the ceiling about an inch, my garage/man cave it is 22'x25'.
By wrapping around the entire ceiling it would give me almost 90' or up to 90ms of delay.
Once the drivers/mics and wiring are installed and tested.Spray around the whole tube with spray foam then drywall it in so I would end up with approx a 4.5" x 4.5" square around the ceiling some trim and it looks fancy and the whole delay chamber is hidden in plain sight.

I can't afford to do this now,but what a cool idea.

[PRR]

> sound better mounted in the tubing that they do used ...freestanding

Small speaker can't make bass waves in open air. Open-back speaker cancels its own bass. These things are awful out in the open. In a tube, both effects go away. Small pistonophone can calibrate a waveguide down to 2hz.

> *
> *
> *
> *
> *
> *

For a 1-night build, six bullet-points is a fair haul.

> boosted the high end a lot

If the tube is small, "air drag" is significant, and IIRC moreso for highs.

If the tube is large, small (treble) waves get cross-ways in it and cancel out.

If the tube is soft (limp plastic), that sucks on the waves.

There's a Lost Art. People in large houses, and on ships, used a LOT of "speaking tubes" before telephony became common. Some were quite long, and evidently reasonably efficient. There must be some rules-o-thumb to select good proportions and materials. That knowledge doesn't seem to be around any more. Perhaps nobody ever wrote it down, not interesting enough.

Lead was probably popular but can't get that today. Brass too, and copper tubing is readily available, though not free. Probably some terne (iron with lead/tin coat) but you have to knock it up from sheet and nobody knocks tin these days.

For voice (guitar) under 50 feet, I *think* 1/2" PEX tubing on minimum 15" radius could be a good start, and quite cheap. Diameter should increase with length, but as you go over 1" the upper voice tones will be muffled. (Note that cost rises at-least with square of length/delay-time.)