Which Ceramic 12" driver?

[stevie]

Balcro has just pointed out to me that my figure for Qms doesn't agree with Beyma's. That's because I had my software recalculate the parameters so that they would all match. I should have told it to recalculate Rms but it changed Qms instead. If you change Rms to 2.03 (although you probably won't use it in WinISD), then 5.6 is OK for Qms.
The difference isn't big enough to affect the simulation - but we might as well get it right.

Edited February 17, 2013 by stevie

[Bill Fitzmaurice]

[quote name='Balcro' timestamp='1361103782' post='1981030']


Enter the basic data as per the help file. Perhaps the first 5 paramenters, Qes, Qms, Vas, Mms and Cms
[/quote]Even that is too much. The issue is that WinISD will calculate the parameters to a higher degree of accuracy than the spec sheets do, and when the results don't agree it results in error messages. The only specs you should enter are Qes, Qms, Vas, Fs, Re, Sd, Xmax, Le and Pe.

[Phil Starr]

You're getting some great support here!

I'm rather boring in that I just follow the instructions on WinISD. What is interesting is that it kind of shows what the program is doing, Mms and Cms for example are the weight of the cone and the bendiness of the suspension and this immediately gives the resonance fs. Add in the SD, Bl and Re; cone area, force the magnet exerts on the coil and resistance and you can pretty much calculate how much air the speaker is going to move for a particular input.
The other thing is the 'errors' that are thrown up by WinISD, the usual discrepancies are Vas and the efficiency figures. They almost never match the manufacturers spec. Not this time because they cheat, most manufacturers give measured figures from finished speakers. Vas depends upon atmospheric conditions and is horrible to measure consistently, efficiency is measured over the whole frequency range usually so often midrange humps boost the efficiency a little over the theoretical.


Win ISD's recommended procedure.

[color=#0000cd]The suggested procedure for entering driver parameters is following (check first that "Auto calculate unknowns" option is checked):[/color]
[color=#0000cd]1. Enter [b]Mms[/b] and [b]Cms[/b][/color]
[color=#0000cd]This gives [b]fs[/b]. If either is not available, then enter [b]fs[/b] and other parameter.[/color]
[color=#0000cd]2. Enter [b]Sd[/b], [b]Bl[/b] and [b]Re[/b][/color]
[color=#0000cd]Now, you should get all but [b]Qms[/b] (and [b]Qts[/b]), [b]Vas[/b]. Please note that Vas may not match exactly what is specified by manufacturer, because exact value of Vas depends on environmental parameters. See [/color][url="""][color=#0000cd]FAQ[/color][/url][color=#0000cd].[/color]
[color=#0000cd]3. Enter [b]Rms[/b] or [b]Qms[/b].[/color]
[color=#0000cd]Either one will do, although I tend to prefer [b]Qms[/b] over [b]Rms[/b], because it can usually be measured in driver measurement procedures.[/color]
[color=#0000cd]4. Enter [b]Hc[/b], [b]Hg[/b] and [b]Pe[/b].[/color]
[color=#0000cd]If [b]Hc[/b] or [b]Hg[/b] or either is available, then enter [b]Xmax[/b] and optionally either [b]Hc[/b] or [b]Hg[/b] if available.[/color]

[color=#0000cd]This procedure is most accurate. Also note that it also calculates true SPL (1W/1m) value. So it might not match the marketing SPL value, which is generally somewhat vague. Not in all cases, though.[/color]

[color=#000000]Hc is the height/length of the coil and Hg the size of the magnet gap. This is where the discrepancy between Eminences Xmax and others like Celestion and Beyma comes in. WinISD will calculate Xmax as if the moment the coil leaves the magnet gap controlled movement stops. In practice if you remember your iron filings experiments with magnets the magnetic field will extend quite a way beyond the magnet. Some manufacturers measure Xmax at a certain percentage distortion and others use a fixed proportion of the gap depth. Beyma are using the most conservative method so don't worry about this.[/color]

[Balcro]

[quote name='Bill Fitzmaurice' timestamp='1361107414' post='1981088']
Even that is too much. The issue is that WinISD will calculate the parameters to a higher degree of accuracy than the spec sheets do, and when the results don't agree it results in error messages. The only specs you should enter are Qes, Qms, Vas, Fs, Re, Sd, Xmax, Le and Pe.
[/quote]

You've taken the sentence out of it's context in the paragraph. The context is to help the OP avoid those irritating error messages. By simply entering those 5 parameters you can get winISD to accept your initial data and it will produce a frequency response curve if little else. Please note that I also said recall your data and enter more at a later time. Two or 3 passes rather than 1, but accurate.

[i]"The issue is that WinISD will calculate the parameters to a higher degree of accuracy than the spec sheets do, and when the results don't agree it results in error messages".[/i] [b]=[/b] "it's always best to let winISD do the calculations, but don't be surprised if the "Fs" or some other parameter differs from the manufacturers by (in this case 0.56 HZ)".

Balcro

Edited February 17, 2013 by Balcro

[Bill Fitzmaurice]

[quote name='Balcro' timestamp='1361112599' post='1981177']
You've taken the sentence out of it's context in the paragraph.
[/quote]Not at all. Entering the Cms and Mms specs often results in error messages. Entering only the specs I noted will never cause a problem. They're also the specs most likely to be accurate on the data sheets.
BTW, OP, be sure to compare driver data sheet SPL charts. Bergs use Eminence, which have a characteristic rising sensitivity in the midrange. Beyma does not have that. If crossed over at 1.5kHz or lower to a midrange it's of no consequence, but if not it's major.

Edited February 17, 2013 by Bill Fitzmaurice

[JTUK]

[quote name='Bill Fitzmaurice' timestamp='1361118811' post='1981305']
Not at all. Entering the Cms and Mms specs often results in error messages. Entering only the specs I noted will never cause a problem. They're also the specs most likely to be accurate on the data sheets.
BTW, OP, be sure to compare driver data sheet SPL charts. Bergs use Eminence, which have a characteristic rising sensitivity in the midrange. Beyma does not have that. If crossed over at 1.5kHz or lower to a midrange it's of no consequence, but if not it's major.
[/quote]

Pertinent..as the OP has stated they aren't keen on higher extention/horns.

I agree that the help/direction here seems very useful... so it will be interesting to hear how it all turns out.

[Phil Starr]

I'll have a go at the frequency response thing since both Bill and Stevie have brought it up. Since we have started with the Beyma and looked at the Eminence beta I'll look at these primarily and maybe take a sideswipe at the Eminence Delta
[url="http://profesional.beyma.com/pdf/SM-212E.pdf"]http://profesional.b...pdf/SM-212E.pdf[/url]

[url="http://www.eminence.com/pdf/Beta_12A-2.pdf"]http://www.eminence..../Beta_12A-2.pdf[/url]

[url="http://www.eminence.com/pdf/Delta_12A.pdf"]http://www.eminence....f/Delta_12A.pdf[/url]

Now the frequency responses of all speakers up to the point where the wavelength becomes shorter than the diameter of the speaker is better described by the computer descriptions than the measurements as these are affected too much by the measuring technique. Eminence tell you exactly the size of their anechoic chamber and the mic they use for example. Up to a certain point the sound is produced by the whole cone moving as a piston but as the frequency rises the cone will start to do two things, it will start to flex with the middle reacting immediately to the force from the coil whilst the outer edges lag behind. At the same time the cone will start to resonate as sound emanating from the coil cone junction reaches the outer edge of the cone and is partially reflected back across the cone. This all results in an uneven frequency response above this point which is about 360Hz for a 12" speaker. All three plots show a little dip in response at this point and unevenness above.

Now this mess is a good thing for a practical speaker designer as without it the speaker would soon run out of hf response and would radiate high frequencies in an increasingly narrow beam. By controlling the thickness,shape and flexibility of the cone a designer can control the frequency response to a fair degree and even the join between the cone and the surround contributes to the response.

On to the three frequency responses then. The Beyma shows a very flat response, on the frequency graph the response stays within the 2.5dB division from below 1kHz to about 4k. It is about 6dB down at 6kHz just about supporting the manufacturers claim of 45-6kHz. This sort of response is typical of a driver designed for PA use with a tweeter and I bought mine for this purpose as it makes crossover design easier. I tried mine with bass because, well I have to try bass through everything and found I liked the sound better than any of my bass speakers.

Moving onto the Beta you can see the response is flat up to 600Hz and then rises up to a 7dB peak at 2kHz and is 6dB down at the 3.8kHz point. There isn't evidence of too sharp a rise in frequency response or a lot of peaks and troughs so the cone breakup is well controlled. This sort of peak is typical of many 'made for instrument' speakers and I think just about all of the Eminence bass speakers. So what does it sound like? Well it is going to give the midrange a lot of extra presence, it will cut through better on stage and I'd describe it as punchy. The trouble is for some people punchy is a slap round the ears and for others it is the gut wrenching sound of your stomach contents vibrating in sympathy with an Ampeg fridge so it is hard to know that you hear what I hear. Bill is probably right about it being part of the Berg sound and the Beyma doesn't have that sound

The Beyma sounds 'cleaner' less cluttered, rather neutral sounding. I practice with headphones and that is the sort of sound you will get. It will go a little deeper than the Beta (and maintain the bass output at higher powers) but without a peak it won't jump out at you. The well controlled frequency response up to 6k is plenty to cover anything the pickups are likely to put out so you won't need a horn but the lack of the usual peak might not immediately appeal, of course you can eq it back in.

Finally my sideswipe at the delta this has a ridiculous peak of 8db between 1.5 and 5kHz with a really sharp cut in and out and this will totally dominate the sound of this speaker. I suspect whatever you play through this you are always going to hear the Delta sound!

Essentially it comes down to taste, if you like a bit of presence at the upper end of your bass and a little mid scoop, a speaker with a character of its own, then you may well get on best with a speaker with the sort of response shown by the beta. If you like something that is more neutral then the beyma is a good bet.

Hope this helps and that the experts will let me off the couple of oversimplifications I've used

Edited February 17, 2013 by Phil Starr

[Phil Starr]

Arghh! where did all those codes come from? Hope I've killed them now

Edited February 17, 2013 by Phil Starr

[Bill Fitzmaurice]

[quote name='JTUK' timestamp='1361120311' post='1981342']
Pertinent..as the OP has stated they aren't keen on higher extention/horns.
[/quote]BTW, it's not all that difficult to figure out what the Berg HD uses. It's a cast frame driver with an 80oz magnet and 350w coil; that means it's a Delta Pro 12 motor. The Berg version might not be exactly the same as the stock version, but it won't be significantly different. Often the main difference between stock and OEM is push button versus spade lug input connectors.

[stevie]

The Delta Pro looks really nice (cast chassis too!), but it's not as good as the Beyma on paper.

[Phil Starr]

The Delta Pro was my original alternative to the Beyma when I built mine. It's a nice speaker which I only rejected because I'd had some failures of Eminence speakers where the aluminium voice coil joined the lead out wires on the earlier models I was replacing.

[6v6]

[quote name='Phil Starr' timestamp='1361110109' post='1981132']
You're getting some great support here!
[/quote]

Agreed! Thanks to all for taking the time to share your knowledge - it's very much appreciated, I'm learning a lot!

[quote name='Phil Starr' timestamp='1361110109' post='1981132']
I'm rather boring in that I just follow the instructions on WinISD. What is interesting is that it kind of shows what the program is doing, Mms and Cms for example are the weight of the cone and the bendiness of the suspension and this immediately gives the resonance fs. Add in the SD, Bl and Re; cone area, force the magnet exerts on the coil and resistance and you can pretty much calculate how much air the speaker is going to move for a particular input.
The other thing is the 'errors' that are thrown up by WinISD, the usual discrepancies are Vas and the efficiency figures. They almost never match the manufacturers spec. Not this time because they cheat, most manufacturers give measured figures from finished speakers. Vas depends upon atmospheric conditions and is horrible to measure consistently, efficiency is measured over the whole frequency range usually so often midrange humps boost the efficiency a little over the theoretical.


Win ISD's recommended procedure.

[color=#0000cd]The suggested procedure for entering driver parameters is following (check first that "Auto calculate unknowns" option is checked):[/color]
[color=#0000cd]1. Enter [b]Mms[/b] and [b]Cms[/b][/color]
[color=#0000cd]This gives [b]fs[/b]. If either is not available, then enter [b]fs[/b] and other parameter.[/color]
[color=#0000cd]2. Enter [b]Sd[/b], [b]Bl[/b] and [b]Re[/b][/color]
[color=#0000cd]Now, you should get all but [b]Qms[/b] (and [b]Qts[/b]), [b]Vas[/b]. Please note that Vas may not match exactly what is specified by manufacturer, because exact value of Vas depends on environmental parameters. See [/color][color=#0000cd]FAQ[/color][color=#0000cd].[/color]
[color=#0000cd]3. Enter [b]Rms[/b] or [b]Qms[/b].[/color]
[color=#0000cd]Either one will do, although I tend to prefer [b]Qms[/b] over [b]Rms[/b], because it can usually be measured in driver measurement procedures.[/color]
[color=#0000cd]4. Enter [b]Hc[/b], [b]Hg[/b] and [b]Pe[/b].[/color]
[color=#0000cd]If [b]Hc[/b] or [b]Hg[/b] or either is available, then enter [b]Xmax[/b] and optionally either [b]Hc[/b] or [b]Hg[/b] if available.[/color]

[color=#0000cd]This procedure is most accurate. Also note that it also calculates true SPL (1W/1m) value. So it might not match the marketing SPL value, which is generally somewhat vague. Not in all cases, though.[/color]

[color=#000000]Hc is the height/length of the coil and Hg the size of the magnet gap. This is where the discrepancy between Eminences Xmax and others like Celestion and Beyma comes in. WinISD will calculate Xmax as if the moment the coil leaves the magnet gap controlled movement stops. In practice if you remember your iron filings experiments with magnets the magnetic field will extend quite a way beyond the magnet. Some manufacturers measure Xmax at a certain percentage distortion and others use a fixed proportion of the gap depth. Beyma are using the most conservative method so don't worry about this.[/color]
[/quote]

Ok, thanks - I think I've cracked it now - I tried entering the data in a different order to that in the WinISD procedure (which always gave me the errors) - I used the parameter list as mentioned by Bill Fitzmaurice, and kept saving after every couple of parameters as suggested by Balcro:

It turns out that the calculated values of BL factor (12.83897) differs from the datasheet (rounded down to 12.8) and the calculated value of Vd (385 cm^3 vs 382 on the datasheet)

Here's what I ended up with - the only parameter I'm not sure on is Pe - is this the Power Capacity stated on the datasheet (350w AES)?

[Phil Starr]

Yep Pe is 350W.

Now you have the data entered you can start modelling straight away, it only takes seconds now for each model. I usually let WinISD calculate the cab size for the first type of vented enclosure it comes up with, usually an SBB4 or a QB3 and then run a second project using the same data so it comes up with an identical cab. Then click on 'plot' and change the colour. You can now start changing box sizes and tuning frequency and see what the effect is on the speakers low frequency performance. WinISD will display both plots on the same graph so by keeping one 'ideal' plot unchanged you can see how you are doing and what you lose by making the cab 'undersized'

Have fun!

[6v6]

[quote name='Phil Starr' timestamp='1361126682' post='1981485']
<snip>
The Beyma sounds 'cleaner' less cluttered, rather neutral sounding. I practice with headphones and that is the sort of sound you will get. It will go a little deeper than the Beta (and maintain the bass output at higher powers) but without a peak it won't jump out at you. The well controlled frequency response up to 6k is plenty to cover anything the pickups are likely to put out so you won't need a horn but the lack of the usual peak might not immediately appeal, of course you can eq it back in.

Finally my sideswipe at the delta this has a ridiculous peak of 8db between 1.5 and 5kHz with a really sharp cut in and out and this will totally dominate the sound of this speaker. I suspect whatever you play through this you are always going to hear the Delta sound!

Essentially it comes down to taste, if you like a bit of presence at the upper end of your bass and a little mid scoop, a speaker with a character of its own, then you may well get on best with a speaker with the sort of response shown by the beta. If you like something that is more neutral then the beyma is a good bet.

Hope this helps and that the experts will let me off the couple of oversimplifications I've used
[/quote]

Thanks, this does help a lot - I'm definitely starting to see how the data on these datasheets ties together and getting a much clearer idea how to make an informed driver choice.

As an exercise (and because it's mentioned elsewhere in the thread), I took a look at the Eminence Delta Pro 12A - I'd assumed it would be similar to the non-pro Delta, but it actually looks a fair bit flatter?

http://www.eminence.com/pdf/Delta_Pro_12A.pdf

It seems like this frequency response chart indicates a gradual rise in the HF response, with a bit of a bump around 2.5KHz, but nowhere near the huge peak of the non-pro Delta, ie it's not as flat as the Beyma, but should be a lot flatter than the Beta/Delta you described?

Given that the Delta Pro seems a fair bit more expensive, and has a lower xmax, it seems to me that the Beyma is still going to the best choice?

As mentioned elsewhere I do realize I'm not going to end up with a Berg sound, I've pretty much decided I'm better just trying to make a nice sounding cab and treating it as a learning exercise rather than attempting to clone some existing manufacturers product.

[6v6]

[quote name='Bill Fitzmaurice' timestamp='1361128296' post='1981528']
BTW, it's not all that difficult to figure out what the Berg HD uses. It's a cast frame driver with an 80oz magnet and 350w coil; that means it's a Delta Pro 12 motor. The Berg version might not be exactly the same as the stock version, but it won't be significantly different. Often the main difference between stock and OEM is push button versus spade lug input connectors.
[/quote]

Lol - that's pretty much all the information I was after when I started this thread!

However it's great that the discussion has turned out so much more detailed (although I admit I found some of the more obtuse responses frustrating initially) - I'm certainly learning far more than I would've done by just buying a Delta Pro and sticking it in a Berg-sized-box without really understanding if/how/why it would work.

Thanks again to all who've taken the time to contribute, I'll be sure to start a build thread when I get started actually making this thing

[Mr. Foxen]

Have you sued the max power handling graph in WinISD? That's fairly informative.

[6v6]

[quote name='Mr. Foxen' timestamp='1361198224' post='1982577']
Have you sued the max power handling graph in WinISD? That's fairly informative.
[/quote]

No I hadn't (so far mainly been experimenting with frequency response plots of box-size/tuning frequency), but I just looked and plotted the Beyma and Delta Pro:

Am I correct in thinking this means that despite having higher peak power handling, the Delta Pro actually has much much lower average power-handling than they Beyma (limited by the Xmax?)

[Bill Fitzmaurice]

[quote name='6v6' timestamp='1361202190' post='1982657']
Am I correct in thinking this means that despite having higher peak power handling, the Delta Pro actually has much much lower average power-handling than they Beyma (limited by the Xmax?)[/quote]+1. And the average OEM driver has less xmax than the Delta Pro 12.

[Phil Starr]

That's the surprise isn't it? It's why so many of the experts bang on about Xmax. Limited to 60W or below 120W for most of the bottom octave is significant. Of course you are using Xmax the undistorted output as the limit and you could use Xlimit which is how far the cone will go before it gets damaged but you will still hear distortion and the inevitable power compression. There are much more extreme examples than the Delta Pro but it is why I look for at least 4mm Xmax.

Edited February 18, 2013 by Phil Starr

[6v6]

Thanks to everyone for your help and knowledge-sharing - I've ordered a Beyma SM212, and started a build-diary thread, where I'll document the box design before I build it, and then the actual build:

http://basschat.co.uk/topic/200152-1x12-diy-cab-build/

I'd very much welcome input from the experts who contributed to this thread to point out my inevitable errors and oversights as I work my way through the process

[JTUK]

This build had better be good, after all this :lol; :lol;