sumsound said:
It has been 12+ years since i have dealt with these. a night club i did service calls to had a pair on the dance floor. I they could have been the SDL's? I replaced cones at least every other week. they ran them right to the edge. they really should have had four. whatever model they were, they had breakers in them. and when you would overpower them the breaker would open and the fan would kick in high with a big whiirrrrrrr then when the cooled back down the breaker would kick back on again. they dj's would turn it up until the bass kicked out then back it down a notch. and 5 hours of that bass heavy dance music just beat the crap out of them.
My comment about the power handling was for each. yes they are quite efficient in groups and very low power compression etc. I have known some to try and use them singly when they really needed more.
I have heard comments about them sounding sloppy by many live sound guys. not necessarily my own opinion.
I'm not an authority, but I did buy two of the parts kits and built 'em. They're impressive, but they do suffer from "grumbling" distortion at low powers, so the right way to use them may be to a) cross them over very low, and use normal subwoofers for the next octave and b) noise gate the Contra amp so they only respond to big rumbles.
If you're still interested...
The ghost of the Contrabass Corner at Archive.org
To save a lot of fiddling with making your own, get a used Contrabass or horn like the BT-7. Used and/or broken, they can sell for less than we paid for the parts kits.
I suspect the grumbling is caused by the rubber pads taking a set, because I could feel a "notch" when I turned the rotary-to-linear-converter by hand, without the motor or cones connected. Or it could be the motor.
The belts in the kit were flat and smooth, smelled like neoprene, and were probably reinforced with kevlar.
If you're still interested...
The ghost of the Contrabass Corner at Archive.org
To save a lot of fiddling with making your own, get a used Contrabass or horn like the BT-7. Used and/or broken, they can sell for less than we paid for the parts kits.
I suspect the grumbling is caused by the rubber pads taking a set, because I could feel a "notch" when I turned the rotary-to-linear-converter by hand, without the motor or cones connected. Or it could be the motor.
The belts in the kit were flat and smooth, smelled like neoprene, and were probably reinforced with kevlar.
It is, it's a blend of 'static friction': http://www.reference.com/search?q=Stiction
Anyway, the belt system has a strong centering torque (the 'notch'), so at very low power the motor may not have enough torque to be linear through its zero crossing stroke. I never noticed it with my two, but then I had them in the corners behind baffle boards to create large BPs, so there was fiberglass insulation to quell any turbulence, cavity/whatever generated noise.
BTW, did you stuff the dust caps with wadding? IIRC I read where some folks that didn't do this heard mechanical noise radiated by them, especially if it wasn't XO'd properly to keep from over-driving them with excessive HF. Then there was at least one person that misaligned the belt system WRT the radiators, causing it to bind at low power.
Gees, if used ones can be had that cheap, then maybe a SD TH might be a viable house wrecker.
GM
Anyway, the belt system has a strong centering torque (the 'notch'), so at very low power the motor may not have enough torque to be linear through its zero crossing stroke. I never noticed it with my two, but then I had them in the corners behind baffle boards to create large BPs, so there was fiberglass insulation to quell any turbulence, cavity/whatever generated noise.
BTW, did you stuff the dust caps with wadding? IIRC I read where some folks that didn't do this heard mechanical noise radiated by them, especially if it wasn't XO'd properly to keep from over-driving them with excessive HF. Then there was at least one person that misaligned the belt system WRT the radiators, causing it to bind at low power.
Gees, if used ones can be had that cheap, then maybe a SD TH might be a viable house wrecker.
GM
No, I never even installed the dust caps.
I did wonder if my assembly was at fault, but reviews of the real thing mentioned similar noises. It can't help that I'm using them in a relatively small room, with a sub only 10 feet from the listening position.
As for used stuff.... maybe talking to local sound businesses (sales, rentals, installers) will turn up some leads. Or patiently searching & watching online ads and brokers.
Does anyone have a service manual for the motors?
I did wonder if my assembly was at fault, but reviews of the real thing mentioned similar noises. It can't help that I'm using them in a relatively small room, with a sub only 10 feet from the listening position.
As for used stuff.... maybe talking to local sound businesses (sales, rentals, installers) will turn up some leads. Or patiently searching & watching online ads and brokers.
Does anyone have a service manual for the motors?
Yeah, these weren't designed for small rooms. There was one owner that had a tiny HT with a pair up against the rear wall with the seating right in front and he wasn't overly impressed, so sold them and built an IB.
All I've ever seen is brush replacement instructions in the service manual.
GM
All I've ever seen is brush replacement instructions in the service manual.
GM
Servodrive motor speaker
I know this is an old post but i cant resist posting even if im the only reader. I have been in the servo motor, drive, control, and amplifier business for about 17 years. I have always been a self proclaimed audio enthusiast and dreamt of building a science project. A servo driven subwoofer. Hey what do you know a quick search on the internet and of course some one has done it. I cant help but think of ways to do it better which are pretty apparent to me considering i have seen the solutions to these problems evolve since the id 90's.
Servo 101-
How do you select a driver. Inexperienced engineers determine how much force is required to move an object. Usually works but not when you need that magic ratio that gets you the highest response out of the motor to the load. The inertia of the motor should match the inertia of the load. Use your imagination to apply this to speakers. I can have a big voice coil applying an ounce of force or a little one applying an ounce of force. Which will most quickly reverse when the commanded motion tells it to reverse?
Dc motors have a brush that delivers electricity to the rotating part of the motor. We will refer to this as the rotor henceforth. Its not easy to make an electrical connection to something that is spinning so the brushes are slowly wearing away. The brushes are designed to do this and thats where you get the dust. Cmpounding the trouble is it takes 3 times the current to to start a motor from 0 speed. Speaker motion is a series of high frequency starts and stops. So 3 times the current continuously rubbing the brush on 45 degrees of the rotor things go wrong quick.
So whats the answer. First, get rid of the brushed motor. Use an ac servo motor. More control smotther motion, no wear, less heat. You still have losses from friction and a mechanical device that changes the rotary motion to linear. New high end servo amplifiers pulse the motor at 20 thousand hertz. Sounds like the electrical bandwidth is there but can the mechanics keep up with that. Reality is a rotary motor will fall on its face at about 4000 hz because of physics. Ah but a linear motor. 50g of acceleration and 20000 newtons of liquid cooled low inertia bliss. We will call it the hadron speaker collider. A black hole opening near you soon.
Im kidding but in reality this would be a really cool science project. I would stick with the ac rotary servo motor though. Everyone failed to mention that a servo motor is a closed loop system that allows you to do some very high end math to make the woofer (mechanical motion) represent the electrical command exactly compensating for any changes in environment. Like temperature or even wear. That last statement was only meant for reasonable people but the point is the sound generated is more truely representative over a longer period of time.
One thing that baffles me is that everyone is saying the servo system has more spl for every given watt of input power. A voice coil is given an exact amount of current and you get an exact amount of movement with close to no loss in motion efficiency because you dont have belts and frictions and so on. Linear motor are the exact same thing as a voicecoil and they are much more efficient than rotary. Find a belt, rack and pinion, or screw assembly that doesnt steal 5 to 50 percent of your efficiency.
I know this is an old post but i cant resist posting even if im the only reader. I have been in the servo motor, drive, control, and amplifier business for about 17 years. I have always been a self proclaimed audio enthusiast and dreamt of building a science project. A servo driven subwoofer. Hey what do you know a quick search on the internet and of course some one has done it. I cant help but think of ways to do it better which are pretty apparent to me considering i have seen the solutions to these problems evolve since the id 90's.
Servo 101-
How do you select a driver. Inexperienced engineers determine how much force is required to move an object. Usually works but not when you need that magic ratio that gets you the highest response out of the motor to the load. The inertia of the motor should match the inertia of the load. Use your imagination to apply this to speakers. I can have a big voice coil applying an ounce of force or a little one applying an ounce of force. Which will most quickly reverse when the commanded motion tells it to reverse?
Dc motors have a brush that delivers electricity to the rotating part of the motor. We will refer to this as the rotor henceforth. Its not easy to make an electrical connection to something that is spinning so the brushes are slowly wearing away. The brushes are designed to do this and thats where you get the dust. Cmpounding the trouble is it takes 3 times the current to to start a motor from 0 speed. Speaker motion is a series of high frequency starts and stops. So 3 times the current continuously rubbing the brush on 45 degrees of the rotor things go wrong quick.
So whats the answer. First, get rid of the brushed motor. Use an ac servo motor. More control smotther motion, no wear, less heat. You still have losses from friction and a mechanical device that changes the rotary motion to linear. New high end servo amplifiers pulse the motor at 20 thousand hertz. Sounds like the electrical bandwidth is there but can the mechanics keep up with that. Reality is a rotary motor will fall on its face at about 4000 hz because of physics. Ah but a linear motor. 50g of acceleration and 20000 newtons of liquid cooled low inertia bliss. We will call it the hadron speaker collider. A black hole opening near you soon.
Im kidding but in reality this would be a really cool science project. I would stick with the ac rotary servo motor though. Everyone failed to mention that a servo motor is a closed loop system that allows you to do some very high end math to make the woofer (mechanical motion) represent the electrical command exactly compensating for any changes in environment. Like temperature or even wear. That last statement was only meant for reasonable people but the point is the sound generated is more truely representative over a longer period of time.
One thing that baffles me is that everyone is saying the servo system has more spl for every given watt of input power. A voice coil is given an exact amount of current and you get an exact amount of movement with close to no loss in motion efficiency because you dont have belts and frictions and so on. Linear motor are the exact same thing as a voicecoil and they are much more efficient than rotary. Find a belt, rack and pinion, or screw assembly that doesnt steal 5 to 50 percent of your efficiency.
I'm still using the ContraBass on a daily basis, so I'm interested in alternative motor options. IIRC one of the advantages of the servomotor approach was that for long excursions, the coil never runs out of the gap. Long-excursion woofers usually have a lot of voice coil outside the gap, and the portion outside the gap just heats up. Tom may have experimented with a closed-loop version, but I guess the open-loop behaviour is similar enough to a conventional voice coil subwoofer to be modeled accurately. And implementing feedback control makes it more complicated to install and use. But, maybe feedback could compensate for non-linear effects like the motor stiction or notch in the rubber. DSP hardware is a bit cheaper now than it was 10 or 20 years ago. On the other hand, conventional woofers have improved as well, so for the same cost you could build an array of 12" or 15" long-excursion woofers and not worry about grumbling noises or brush wear or awkward impedances.
Here's a better mirror of the ContraBass Corner:
http://phy6.net/downloads/electronics/Servodrive/DIY/www.mindspring.com/~sdinc/pages/contra.html
Here's a better mirror of the ContraBass Corner:
http://phy6.net/downloads/electronics/Servodrive/DIY/www.mindspring.com/~sdinc/pages/contra.html
Last edited:
I forgot about this thread but still pretty interesting. You can create a model in a servo drive of how the electrical characteristics change with heat over the long run and use a thermister to control the averaging function over time. You can refine the short term reaction to split second current and temperature rise to keep the motor to drive loop extremely tight. If you couple the motor to a conventional speaker that has a voicecoil attached to the cone or moving mass you can confirm that the cone moved how you expected it to move. The voicecoil could be used to generate a voltage like a generator that is fed back into the drive. So the velocity loop is closed on the motor and is refined by the voltage coming back from the voicecoil. This would compensate for wear or envirnmental interference. Actually this would remove any irregularities that come from the surround, spide, or cone inertia. I wonder. It probably takes less energy for a speaker to return to a resting position because of the spring in the system. Ill bet there is a time delay from an electrical perspective when a cone is pushed out to a position to generate a single wave and when it reverses back to start the next wave. Most inductors have this weird thing called hysterises. For example it takes 5 volts or amps to move the cone out and you would think the cone would begin to reverse as soon as the voltage or amperage falls to say 4.9 but it doesnt. Sometimes it takes maybe 4.5 before the reversal starts. I think all the spring in a speaker if designed properly limits how bad this hysteresis is. If you refine a closed loop servo with feedback like an analog voltage from the coild on the mechanical side of the motion you could eliminate the hysteresis. That makes me wonder if you need a suspension at all. You still need to guide the cone so that the moves are linear but the spring isnt needed. It takes a high end servo about 62.5 microseconds to measure a deviation in motion and make a physical correction. If my math is right i could correct 1 wave of a 100 hz signal 1.6 times so active correction would be limited to about 160 hz if you wanted to correct every wave. Hmm. For everything above that you would have to take samples of comanded motion and the feedback from the cone and make a model that averages and does predictive compensations based on a setup run that is stored in the drive. So the drive sees a command for 500 hz looks at the model and knows that it needs to increase or decrese current for 3 waves and the 4th wave brings back a confirmation that advances or retards the table. Ok, that hurt. Too late for all this. Thanks for encouraging me to break my brain. Lol.
I can get the servo system to do this but i need someone who knows the mechanics of a speaker and can help build this frankenstein.
I can get the servo system to do this but i need someone who knows the mechanics of a speaker and can help build this frankenstein.
Tom Danley's Intersonic servodrive subwoofers had an Xmax of 12.5 to 16 mm depending on the unit.
At a time when most woofers had an Xmax of around 4mm, the advantages where huge, and with four units of the BT7 having a sensitivity of 110 dB 1 watt one meter, half space, 28- 125 Hz, they were as efficient as cone speakers. There simply were no cone speakers with the BL product or Xmax anywhere near what the servodrive units had at the time.
Times and technology changes, with off the shelf speakers now available with far more Xmax than the servodrive cones for far much less money (primarily from competition in the car sub industry), and the cost of servomotors having risen, and the belt technology being prone to wearing out, the idea is a time that has come and gone, though there are still many left in service.
The cabinets can be retrofitted with standard cones, a procedure simpler than trying to procure parts for a product that unfortunately is no longer supported.
Most servodrive units seem to be sold in a non-working state, the cabinets often can be purchased for less than the cost of the Baltic Birch plywood they are made from.
For anyone considering building Labsubs or similar large straight horns, retrofitting BT7s could save considerable expense and time.
Art Welter
Last edited:
As Planet10 points out, there is one main thread here, Danley's remarkable servo-motor woofer and with all other kinds of motional feedback lurking in the background.
Bry195 has imaginatively invented a servo-motor with vc motional sensing. But weird to use a servo-motor when you can use a vc to drive the cone!
Which brings me to my point: unless there is fresh magic in servo-motors and their sensing components today, it seems a far-fetched way to move cones, even woofer cones, certainly as compared to voice colls which work down to DC and have nearly 100 years of R&D behind them and lots of commercial examples to play with.
The challenge is deriving suitable feedback to correct the voice coil's motion and then integrating that feedback into the amp that drives the vc. For sure, amp design can handle the feedback frequencies and now-a-days, digital systems can handle the frequencies in stride too.
I wonder what new and wonderful kinds of position sensors are out there to use in a feedback system?
(I have several decades experience, until not long ago, of using motional feedback using the vc in a bridge to get the feedback. I urge folks to try it for super bass.)
Ben
Bry195 has imaginatively invented a servo-motor with vc motional sensing. But weird to use a servo-motor when you can use a vc to drive the cone!
Which brings me to my point: unless there is fresh magic in servo-motors and their sensing components today, it seems a far-fetched way to move cones, even woofer cones, certainly as compared to voice colls which work down to DC and have nearly 100 years of R&D behind them and lots of commercial examples to play with.
The challenge is deriving suitable feedback to correct the voice coil's motion and then integrating that feedback into the amp that drives the vc. For sure, amp design can handle the feedback frequencies and now-a-days, digital systems can handle the frequencies in stride too.
I wonder what new and wonderful kinds of position sensors are out there to use in a feedback system?
(I have several decades experience, until not long ago, of using motional feedback using the vc in a bridge to get the feedback. I urge folks to try it for super bass.)
Ben
Last edited:
The Contrabass was around 200L with 92 db sensitivity to 20 Hz, about 87 db @ 15 Hz. 22mm excursion with dual 15" drivers and a pair of 18" PRs.
200w > 114 db
Quite impressive.
A single TC Sounds LMS 18" in a sealed box will now do the same job, although it does require more power in a box that size.
200w > 114 db
Quite impressive.
A single TC Sounds LMS 18" in a sealed box will now do the same job, although it does require more power in a box that size.
Not to say the Contra Bass was not impressive for it's time, but the dual 15's had a maximum peak to peak excursion of .75 inches, 19mm/2 =9.5mm.
The passive 18" had 1.75" P to P, 44/2 = 22mm.
Those figures would be Xlim, not Xmax.
The TC Sounds LMS 18" has 38.1mm Xmax, just over 3” peak to peak.
Big increase.
@lechuck, the ServoDrive brand of subwoofers used an actual electric motor connected by belt's to two rods that pushed/pulled on woofer cones! Not some type of motion feedback system. actual motor and belts!
The big deal with the servo drives was massive output in a smaller size with less power AND very little power compression. They have an interesting back history. some college was traipsing around the jungle studying the low frequency sounds elephants make, dragging big pro type subwoofer boxes with them trying to reproduce the 5hz sounds. well they were blowing the pro subs up left and right and they hired Intersonics, the company Tom Danley worked for to make some mega subs. and walla, ServoDrive was born!
The big deal with the servo drives was massive output in a smaller size with less power AND very little power compression. They have an interesting back history. some college was traipsing around the jungle studying the low frequency sounds elephants make, dragging big pro type subwoofer boxes with them trying to reproduce the 5hz sounds. well they were blowing the pro subs up left and right and they hired Intersonics, the company Tom Danley worked for to make some mega subs. and walla, ServoDrive was born!
So Mullet, can you give an idea of the outcome you are shooting for?
To make it clear.
Db output @ 1m into your space
Bandwidth eg20hz to 80hz +-3db
efficency 90db/watt
Price in dollars
Power available in watts
how many systemsincluding amp channels are to be used
Size of venue.
type of music played?
do you want to build this device from scratch or buy a complete unit and install it?
does it need to be portable?
External maximum dimension?
Deal breakers? maybe highlight these parameters so we know exactly what parameters will make you abandon a design.
I think this format of posting will help us help one another more efficiently as we seem to pull on a few people alot then bail out on their hard work because it doesnt meet our criteria.
To make it clear.
Db output @ 1m into your space
Bandwidth eg20hz to 80hz +-3db
efficency 90db/watt
Price in dollars
Power available in watts
how many systemsincluding amp channels are to be used
Size of venue.
type of music played?
do you want to build this device from scratch or buy a complete unit and install it?
does it need to be portable?
External maximum dimension?
Deal breakers? maybe highlight these parameters so we know exactly what parameters will make you abandon a design.
I think this format of posting will help us help one another more efficiently as we seem to pull on a few people alot then bail out on their hard work because it doesnt meet our criteria.
Servo systems are, of course, feedback systems. That ensures the motor accurately replicates the signal to the motor. But paradoxically, servo-drive speakers do not have motional feedback as far as I know because there is no feedback from the end-point.
We could argue endlessly about how close the my ear is the feedback loop taken. I favor voice coil feedback because that is a fairly good representation of driver sound and results in bass which is incredibly extended, flat, and clean (AKA HiFi), at least for stiff drivers in sealed boxes. Others may think the a loop that just includes the servo-motor (which drives some cone-like surface, sometimes through a length of rubber band) is close enough to their ear.
Ben
We could argue endlessly about how close the my ear is the feedback loop taken. I favor voice coil feedback because that is a fairly good representation of driver sound and results in bass which is incredibly extended, flat, and clean (AKA HiFi), at least for stiff drivers in sealed boxes. Others may think the a loop that just includes the servo-motor (which drives some cone-like surface, sometimes through a length of rubber band) is close enough to their ear.
Ben
Last edited:
No - that is the other kind of servo. Rythmik have servo subs, both DIY and fully assembled.
In this thread, servo refers to the driver motor based on a voice coil being replaced with servo motors attached to the cone with the motor coupled to the cones via belts.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.