Final Speaker design before driver purchase - please comment!!

[AndrewUK1990]

Hello again everyone, i have completed yet another simulated crossover design using FRD tools. Ive followed the rjb tutorial very closely and after several weeks of faffing about with learning how to use all the software and getting my head around crossover design i have finally come out with this, which i hope to be my actual design which ill use to build my loudspeakers. I have designed the cabinets and im happy with their design, but i would just like some last advice and/or comments from anyone concerning my crossover design, just so i know what people think and also if there are any possible flaws with my design. I will be going ahead with this project as long as no one spots anything majorly wrong with the design here so it's quite important that ive got this right.

Also i am well aware of the fact this is only a simulation and the real world results will differ to some extent but i will use this as a solid base to work from prior to obtaining proper measuring equipment etc.

Anyway, enough jabbering on, here are some quick snaps of the crossover design and frequency response / phase diagrams as well as impedance. Minimum impedance is about 4.5Ohms also. Note the drivers im using here are: 2x Scanspeak 18W/4531G00 bass, Morel EM428 Midrange, Scanspeak Classic D2904/950000 Tweeter.

Thanks for reading! Any advice/comments will be greatly appreciated as always!

CROSSOVER DESIGN:

IMPEDANCE GRAPH:

FREQUENCY RESPONSE GRAPH:

[PeteMcK]

I'm wondering why the mid has attenuation when the data sheets show it to be less sensitive than the woofers? something to do with BSC?

[AndrewUK1990]

Hey PeteMcK, I was surprised actually at that myself but I think it is to do with BSC. In the crossover I have actually started the lowpass filter in the 200hz region to eliminate the rise from baffle step which occures in those upper bass frequencies. I have also modelled the drivers on my baffle design in BDS, so would I be right in assuming this program takes into account things like BSC yes? Also I used wwMMT setting in BDS as both bass drivers which are adjacent to one another cover the same frequency range. The Mid and Tweeter are seperate. Also my predicted frequency response for the scanspeak drivers shows that their sensitivity will be about 88db in low frequencies and rise with a constant gradient up to about 8khz or so where it peaks at around 94db. This would show why they dont actually seem more sensitive than the morel midrange because the frequency range they cover in this design is only up to around 300hz or so.

Also note that only one line is present on the graph of frequency response for the bass units because they both follow a very similar frequency response obviously so i didnt see any point in showing both. However, both drivers are used as can be seen clearly in the crossover design and also in the FR graph which shows a higher summed response of both drivers. Also the bass units are wired in series to keep their sensitivity down as they are already 90dB each being the 4ohm version of the 18W Revelators.

[AndrewUK1990]

Hey all, ive altered my crossover slightly after taking into account driver offset. However it is still quite similar and i am now satisfied with the results so im going to build this project now !

First though, there is one thing i have overlooked here! What are the advantages/disadvantages of using air cored or ferrite cored inductors? I know the physical differences and how each achieves their goal but is one type better than the other and if so please can someone explain? Thanks in advance!

Ill be buying these components and building the cabinets ive designed some time in april so i am just looking to understand the differences in the various components with regards to quality.

[benb]

With air core inductors you are guaranteed they won't saturate.

You can also make/wind air core inductors yourself easily and calculate inductance and resistance based on well-known formulas based on coil dimensions (for inductance) and wire length and diameter (for resistance). For some neat info, google brooks coil. Air core coils do put out changing magnetic fields into freesspace, which means you don't want the speaker next to a CRT or magnetic tapes, but you already knew that because of the drivers' large permanent magnets. Air cores also pick up changing magnetic fields, so such inductors should be mounted so their mutual interaction is minimized (generally placed at 90 degrees to one another and with as much spacing as is practical, rather than being stacked on top of one another).

Ferrite and other cored inductors are physically smaller and use less wire, so can have smaller series resistance. With the reduction in wire used, they might even weigh less, even counting the weight of the core, but these are not generally considerations in speaker crossovers - the main positive here might be the lower resistance. There's the possibility of core saturation at high current, resulting in an instantaneous decrease in inductance and thus distortion, but saturation current should be in the specs of a commercially made inductor, and that spec should be greater than any PEAK current the inductor could possibly see.

I'm not familiar with what's available in ferrite and other-core inductors, but long story short, I think air cores are great for crossovers, and if you've got an accurate way to measure inductance (for tweaking - the winding formulas will take you 95 percent of the way) you can wind your own with the appropriate size magnet wire.

[gfiandy]

Air cored inductors do not suffer with the saturation problems that ferrite cored ones do. They are usually considered to sound better.

Disadvatage is you need much more wire to mae an air cored inductor so they are more expensive.

You need to look at the resistance as well as the inductance as this can become significant in your design and affect the response.

Even if you take the resistance into account in your response the bass / mid speaker will tend to be less well damped if the resistance of the inductor is high as the amplifer has to control the back emf of the coil through this resistance.

It would be worth looking at the resistances of some off the shelf inductors of about the same size and type you intend to use and add this to your simulation to see the effect. The resistance appears in series with the inductor.

Regards,
Andrew

[AndrewUK1990]

Great stuff, cheers for the info guys. I assume its ok to mix and match the inductors to some extent. I mean i would likely use the air cored inductors for most of the inductors, but for the 6mH inductor i would perhaps use a ferrite cored one to keep the resistance down, as well as the cost of the components and i dont really want a huge inductor either.

[benb]

Yes, you can mix, and furthermore, to amplify on Andrew's response, in your schematic you should add a resistor in series with each inductor to model each inductor's resistance. Typical values might be between 0.05 and 1 ohms, but you can put in exact values (both inductance and resistance) from the specs of the inductors you're considering buying and see how they affect response.

[AndrewUK1990]

Thanks for the advice there benb, i have already inputted resistance for each inductor. Speaker Workshop allows you to do this in the individual component properties so i dont need to add a resistor for every conductor (but i do understand what you mean).

Cheers for pointing it out though

[AndrewUK1990]

Thanks very much everyone for the help so far, im almost settled now on my design and im just running though checking the last few things before i go ahead.

I was just after peoples opinions about which tuning frequency would be best. Currently i have the whole system design (Fb) tuned at 30hz. This frequency gives me an (F3) -3dB freq of about 39hz but further extension into the bass region. I have chosen this because the scan drivers are so sensitive that the small amount of SPL they lose from extending this low is nothing to worry about. Also i should mention the drivers dont come anywhere near their huge Xmax excursion distance with this amount of extension so it should be quite safe.

The box btw is 60L, if i try a tuning (Fb) of 35hz, the F3 -3dB freq moves down to 35hz but extension is slightly reduced. SPL is increased slightly also in the bass region relative to a tuning of 30hz.


So basically, i just would like to know peoples opinions on the matter as to which would be the better option to go for, 30hz or 35hz Fb. This is obviously quite subjective based on music type and preference etc but im just after some opinions. FYI i listen to a lot of metal/rock, Iron Maiden and such like, but i also like stuff like Mark Knopfler or Eric Clapton and 'unplugged' music when im not up for an ear bashing!