itty bitty rotating speaker

[jaapie]

I cobbled this little thing together over the holidays. I've got a bunch of those little $1 mylar speakers from tayda lying around and I'd been fooling around building small enclosures out of hobby plywood. Like pretty much everyone else around here, I've always been fascinated by leslies and their ilk and figured that a little rotating speaker would be fun to build and play around with. Here's a quick demo video: http://youtu.be/6OoB68_T1yM.

The gearmotor was ripped out of a cheap old RC car with a dead-simple PWM speed control using a 555. The swivel/slipring is just a telephone detangler attachment (like this) jammed into a hole in the top of the rotating cabinet. Only one speaker is active in the video, but since telephone cables have 4 conductors the two speakers can be used independently. The amp is just half an opamp and a 386 i threw onto a breadboard just to test it, although i don't think it ended up sounding too terrible. I thought it might be interesting to invert the signal to one speaker to sort of make a crappy mechanical 4-quadrant multiplier, but it didnt seem to make a lot of difference at first listen and I havent really fooled around with that idea much since then.

It definitely works, but it's more subtle than a 'real' leslie. It's been a month or so and I still haven't gotten around to finishing it up (the motor controller and amp never made it off the breadboard). Anyway, i thought some folks here might get a kick out of it. Let me know what you think!

[Mark Hammer]

1) My hat is tipped, and I bow graciously and deeply to your diligence and industriousness.

2) The doppler effect produced by moving sound sources, depends fundamentally on the distance covered.  It is the place where a sound wave peak (a moving compression of air particles) occurs, and the velocity with which the sound source that produced that peak (but has now moved) is travelling, that produces the cancellations/notches where it does.

Put more bluntly, for cancellations in the audio range to be produced, a speaker and cab that small would need to be rotating at a very fast speed; indeed, faster than we would actually appreciate.

Put more practically: the richness of the doppler effect (the plentifulness of notches in the audible spectrum) will increase with the radius of the rotating sound source.

So, the speaker itself doesn't have to be large, but the place where the sound comes out needs to be farther from the center of rotation.

Why?  Because sound has a fixed speed, and the doppler notches are produced when the sound source is moving near that speed.  If I have a small radius, then the circumference is fairly small.  That, in turn, means that if the speaker rotates at 1hz, the sound source hasn't really travelled very far in that one second.  If the radius is larger, then by doing one rotation in a second, the sound source will have travelled much farther in the same amount of time: a higher velocity.

Anyone adept at the search function here may probably embarrass me by noting some posts I may have made years back, proposing something similar to what jaapie has made.  I knew less then than I do now.

[bool]

That looks very very nice.

[Brisance]

Mark, isn't doppler effect the stretching out of soundwaves due to moving source as opposed to filtering effects?

[Morocotopo]

Extremely cool!!!

[Mark Hammer]

Mark, isn't doppler effect the stretching out of soundwaves due to moving source as opposed to filtering effects?

It's both.  As the source moves, the compressions and rarefactions it produces in the air (that ARE the "sound waves") as it is moving can line up in a way that results in cancellation; what we call "notches".  In the same way that a bucket brigade chip that "swept" between 10 and 200usec would produce simply inaudible flanging, if the velocity of the moving sound source, compared to the speed of sound, is such that any phase-related cancellations occur either bwlow or above human hearing range, there will still be a doppler effect, of sorts, but we just won't hear it.  That's why jaapie says it is "more subtle than a real Leslie". 

If the very same speaker were mounted in a cabinet that was more rectangular than cube-like, such that the speaker had a longer radius from the center of rotation (which obviously requires a redesign of the bracket holding everything in place), the doppler effect would be "thicker", because the speaker would be travelling along a longer/larger circumference in the same amount of time: i.e., a faster velocity.

If you harken back to the days of vinyl (or have recently joined them), you will note that the "hit single" was often the first track on the album side.  Keep in mind that the album spins at a constant rate, such that as you get closer to the inner tracks, the actual velocity  (distance covered per unit of time) decreases.  And, as velocity decreases, and more information is packed into a smaller space, so does audio quality (which is why tape recordings made at 7-1/2 or 15 inches per second sound MUCH better than those made at 1-7/8 or 3-3/4).  If you wanted a single to sound better and be more appealing, you stuck it where the velocity would be highest - closest to the perimeter of the album side.

In this digital era, we can easily forget about the role that distance plays in some of the things we like.

[karbomusic]

Mark, isn't doppler effect the stretching out of soundwaves due to moving source as opposed to filtering effects?

It should be both, and it should result additionally in pitch changes since that is the primary component of the Doppler effect, such as a train approaching and passing while blowing its horn. Knowing this, the time spent moving toward or away from you (aka affected by the circumference of the spin) is going to come into play.

However, some leslies simply have a opening that rotates around the bigger (bottom) speaker and the two horns on top have some distance. Then there are the volume changes that occur because the speaker isn't always pointing directly at you. Goes to show why attempts to reliably recreate them electronically have been more complex that not.

[PRR]

>> more subtle than a 'real' leslie.
> the place where the sound comes out needs to be farther from the center of rotation.

+1

My guess is that around a foot radius is a good compromise between useful rate (RPM) and depth.

> some leslies simply have a opening that rotates around the ....speaker

Speaker mounted a foot off the axis is very unbalanced. Especially the heavy woofer. The woofer is fairly omni, so we can instead spin a light-weight sound deflector. Sine the extreme radius is an opening, the deflector can be almost perfectly balanced.

The 2-horn idea is also balanced but blares twice per turn. For the same trem rate it must run half the RPM. That also suggests twice the radius to keep the same depth. The actual product uses horns as big as will spin in a cabinet that passes through a door-way.

The telephone cord swivel connector is a very clever hack. I'd worry about long-term reliability if I sold it to someone else, but as a for-self DIY expedient it is sweet.

> volume changes that occur because the speaker isn't always pointing directly at you.

Not just Volume. The horns are omni at their lower limit (1KHz?) and very beamy above 6KHz. So there is a varying low-pass. But also the two omni sources spaced a varying distance apart will comb-filter.

I do not think it is practical to emulate all this in a box. There's very real room effects that can not be duplicated any other way. OTOH, I'm not sure the audience cares after the 2nd drink. If the music is good any wobbulator is fine; if the music sucks who cares if it comes out of a vintage Leslie?

> i thought some folks here might get a kick out of it.

Real kick. Thanks for posting.

[Jdansti]

Cool experiment!!! 

[Buzz]

That's great. I really love how you did it out of parts you had on hand. True DIY champion.

[KazooMan]

That looks like a really fun project.  In addition to what the others have suggested, I wonder if you might get an improvement in the effect by having more of an enclosure around the unit.  There is vibrato and then there is tremolo.  The other suggestions were directed at the variation in pitch from the Doppler effect.  Enclosing the rear and sides, especially with some sound absorbent material should give you a more noticeable variation in amplitude of the signal.  I think that is a part of what is happening with a fixed speaker with a rotating baffle.

[Tony Forestiere]

...I wonder if you might get an improvement in the effect by having more of an enclosure around the unit. 

Take a cheap, large-ish trash can and flip it over your spinner. Experiment with cutting some slots in the sides of the can to hear what differs. (Duct tape the failures.)

[tubegeek]

The 2-horn idea is also balanced but blares twice per turn. For the same trem rate it must run half the RPM. That also suggests twice the radius to keep the same depth. The actual product uses horns as big as will spin in a cabinet that passes through a door-way.

The full-sized Leslies I've worked with (122, e.g.) have one "real" horn coupled to a HF driver, and one dummy horn that is there only to counterbalance the rotating mass. So typically there is only one spinning sound source.

Also: depends on the doorway

[PRR]

> dummy horn

I thought you were making fun of the brass section.

You are correct, my memory sucks.

[tubegeek]

You are correct, my memory sucks.

If it weren't for the flaws in your memory I'd never have an opportunity to correct you - think of it as a public service you're making towards my self-esteem.

[petemoore]

I was surprized how easy it was to make good 'spinning contacts' on an old phonograph turntable, the speakers [2] had to be small, all made from common parts.
The fixed connector was a phone plug plugged through the bottom of the turntable [where the record centering pin was].
One rotating connector was screwed to the turntable and clamped the sleeve of the plug [a set of points for automotive distritubor].
The other connector was on an old style nylon spoon, used to spring it firmly down on the center of the plug tip, the contact made of a brass pellet [tailpiece stop from an old electric guitar string] was melted into the spoon shovel.
No static, worked great, to our amazement...we'd expected static.

[jdub]

Cheers to you, friend.  That is totally cool.  Well done!

[DougH]

Hey, that is really cool. Sounds really good! Well done!

[tranceracer]

diy w/ recycled parts and it sounds good to!  Awesome!