I'm doing a design process (for myself) that starts at the desired application, choosing the enclosure type, delimiting the max dimensions and setting targets for the resonant frequency and the loudness.
Then I'll set targets for the T/S parameters and build a subwoofer with them in mind, so the obtained parameters will allow me to design an enclosure with the profile I want.
Noting that there are more limitations in respect to the driver: I have a chassis, two magnets and just a few options for the other parts; so I'll be mostly doing coarse estimates to choose the best combination.
But I'm affraid the Qms maybe too hard to estimate.
Without assembling everything I can use a weak glue or even tape to attach cones and spiders, place them in the chassis and estimate the Q factor by adding weights and analyzing the oscillation.
But I'm afraid that value is very different from the one I'll get after being properly assembled, because when assembled the resistance caused by having to force air around the coil increases dampening a lot.
Do you think the change in the Q factor due to this one factor is enough that even coarse estimates may fail?
Thanks!
Then I'll set targets for the T/S parameters and build a subwoofer with them in mind, so the obtained parameters will allow me to design an enclosure with the profile I want.
Noting that there are more limitations in respect to the driver: I have a chassis, two magnets and just a few options for the other parts; so I'll be mostly doing coarse estimates to choose the best combination.
But I'm affraid the Qms maybe too hard to estimate.
Without assembling everything I can use a weak glue or even tape to attach cones and spiders, place them in the chassis and estimate the Q factor by adding weights and analyzing the oscillation.
But I'm afraid that value is very different from the one I'll get after being properly assembled, because when assembled the resistance caused by having to force air around the coil increases dampening a lot.
Do you think the change in the Q factor due to this one factor is enough that even coarse estimates may fail?
Thanks!
Last edited:
Qms will mostly be affected by the material of coil former used. With aluminum it will be much lower. With any nonconductive material it will be much higher.
Well remembered, I forgot how relevant are the forces due to eddy currents are in a aluminum former (which is probably the former I'll use).
But If I account for that, do you think that I can ignore the further decrease of Qms caused by forcing air around the coil?
Last edited:
The damping from the eddy currents in an aluminum coil former are going to swamp any damping forces from air friction in small gaps in the motor.
Wait a minute, I have a question
I forgot that, at least in the coils I have, there's a vertical gap in the aluminium sheet. They don't form a closed ring.
Shouldn't this enormously reduce the eddy currents?
Yes, the slit former design does reduce eddy currents. They would be much higher if the former were a solid tube. More or less no one makes them with a solid tube if they are aluminum.
Well, there are a lot of variables
- The amounts of glue, air load, and moving portion of the suspension need to be accounted for into Mms.
- Qms is inversely proportional to Rms, the unknown you're talking about. I can't naysay Jack Hidley that aluminum would "drag" more than air, but to estimate the actual value of Rms...tough. Maybe look at the parameters of designs similar to yours. If Qms is high compared to Qes, Qes will tend to swamp out the mechanical Q.
- Do you hav access to some kind of FEM simulator to model your magnet? Or something like SpeaD from Redrock Acoustics? BL has a huge effect, which really needs some kind of simulation or prior-prototype data. The field has a varying profile which integrates across the coil, so while I used to calculate B in the gap from magnet curves and an abacus that just gets a rough value.
- There are setups to measure the compliance of a spider and surround-do you have access?
An awful lot of stuff can be simulated these days, but in the end what I don't think has changed is someone finally builds a prototype to nail down all those vaguely known items. A memory floats into my head: when I started as a loudspeaker engineer, my fellows were pretty much cut-and-try guys that would build a prototype, measure, rebuild, remeasure, and so on. I had more theoretical background, so I could short-cut the first few prototypes* but in the end I had to build something to really see where I was with respect to T/S.
*with my boss asking with annoying frequency "So, how's it going?" since he didn't see me building anything. Finally one day he came around 10 am and said they needed a prototype for a meeting with Mazda. Oh it's for several years of production, multimillions of business. No pressure. Oh we need it by 1pm to take on a plane to Japan. After freaking out for a few seconds I got out my slide rule and got to work calculating what magnet and cone and so on, built a prototype, glued it, measured it, gave it to the fellows getting on the plane. YES we did get the business, yay!
P.S. OK there were no abacuses or slide rules involved. But while cleaning a storage area, they DID drop a 135 lb Crest power amp right onto my coworkers desk! WHILE HE WAS SITTING THERE!!! Thank heaven he didn't have his arms on the desk. They later kept him after his normal leaving time-knowing he took the train home-and laid him off, he had to call his brother for a ride. The company president later said they were moving to a new office complex (which it turned out he had a significant hidden share in) upon which the engineering executives-who lived nearby and did NOT want to move-got wildly inflated estimates of the cost to build new anechoic chambers. Engineering didn't move. 😀