1x12" Cab Design Diary

[tks.se]

[quote name='Phil Starr' timestamp='1412163364' post='2566270']
Thanks

That's great information, do you have a link?

I suppose I'm designing for worst case, because someone somewhere is going to think they have a 350W speaker that can reproduce anything they throw at it, add 12dB of bass boost from a traditional tone control and the wattage distribution would look very different and my suspicion is that someone, somewhere will do this. I suppose the possibility of designing this hazard out was worth a look, in the end the compromises weren't worth the gains and I'd advise anyone with a fiver who wants to drive at extreme sound levels uses at least two speakers and preferably a filter like a Thumpinator.
[/quote]

Only in Swedish unfortunately. Here's a promise to have it translated to English, but it could take a while, much to do!

Designing for the worst possible scenario is always a good idea, but measurements like the one above puts a little more perspective on things. It's hard to get all the energy out through the fundamental.

We've done some workshop experiments focusing on "worst case". That means using a speaker with a small xlim value in a cabinet we know how it affects the excursion (size/tuning), and experimenting on what tone that in the real world gives you the most trouble, by adjusting EQ and volume until repeated "voice coil former meets back plate" occurs (does not destroy the speaker instantly, but is of course not recommended at all). This particular testing setup had the most excursion (above the tuning frequency) calculated around 69 Hz.

When using the B-string, all EQ settings att min, except bass, which was on max (+12 dB) the problem centered around C# on the B-string. Lower or higher made it gradually go away. The fundamental of that tone is around 34 Hz. Double that (go in to the harmonic range) and you're awfully close to the 69 Hz of calculated maximum excursion. When going lower, the problem went away, detuning to A meant no problems, even though the calculated excursion around 34 Hz and lower should be higher than that around 69 Hz, which means that less power goes through the fundamental than the harmonics, even at those settings.

Still, it's up to the user to have a bit of common sense and not actively try to mistreat the equipment - an octaver would probably make things even worse in the above case. It's easy to list, say car analogies of what not to try with your car or expect a car to manage without damage that would mean comparable circumstances.

[tks.se]

Just a comment about the testing procedure - in the workshop a lot of things start to sound like distortion when turning the volume way up. Voice coil former hitting back plate gives off a noise you can't mistake, so in that environment it's a much faster way to examine where the highest excursion occurs than other methods. Not a recommended method though! The speaker used was a cheap 10" speaker, it's much harder to push a high quality, high volume displacement speaker to those limits.

Edited October 1, 2014 by tks.se

[tks.se]

[quote name='tks.se' timestamp='1412160565' post='2566219']
When playing a low B, most of the tone consists of harmonics to the fundamental.
[/quote]

A quick correction to this as well, all tones on the bass consist of fundamental/harmonics, with more energy in the harmonics, not just the low B. Not brilliantly expressed there, just commenting on it to avoid any misunderstanding.

[Chienmortbb]

Phil, Do not beat yourself up. It was important that the three of you disagreed about some things as it makes sure that each idea or theory is fully tested. There was no right or wrong so be proud of the end result. One of the reasons I stopped designing my own cab was that is was clear that in this case a small committee would get to a better answer than I could alone.

TKS is right. I first came across this from Walter Harley (www.cafewalter.com). He did a spectral analysis of bass signals and came to the same conclusion as TKS. The fundamental is important but the harmonics give us more info.

I would also say that your observations just prove to me that a good HPF in front of the power amp is worth its weight in gold.

[stevie]

[quote name='tks.se' timestamp='1412160565' post='2566219']
When playing a low B, most of the tone consists of harmonics to the fundamental. Here's the result of playing a low B and running the signal through a spectrum analyser:


[/quote]

What exactly was being measured here? Was the loudspeaker eq'd to be flat to 30Hz?

[6v6]

Interesting update!

FWIW in my build I made the fixed shelf tuned to ~50Hz, and later I added a removable extension which moves it down to ~42Hz tuning.

It's fair to say the difference is subtle, but audible, and since I fitted the shelf extension I'm getting occasional ugly sounds on hitting the low notes (only a 4 string but I'm running a 300w LM3 into one speaker) - it's interesting that you note port noise, I've not checked my plots but perhaps that explains it.

[size=4]I'll be going back to the original 50Hz tuning when I can be bothered to take the box apart to remove the shelf extension. [/size]

[Chienmortbb]

[quote name='stevie' timestamp='1412240912' post='2566993']


What exactly was being measured here? Was the loudspeaker eq'd to be flat to 30Hz?
[/quote] I think it is straight from the bass. There may be a buffer just to make sure the pups ate mot loaded by the spectrum analyser.

The point is that most of the imgirmstion is contained in the harmonics and the brain is very good at filling in the gaps.

[stevie]

I think most people will assume that it's straight from the bass because that's the only data that makes any sense, but the measurements shown are of SPL. So we need a bit more information on what was measured.

[Phil Starr]

I'd take the measurements as indicative only. All basses have a different timbre, or mix of harmonics. The pickup positioning is going to be critical. The further up the string the more fundamental and the less harmonics you'll pick up. Put it on the 12th fret and you'll get maximum pickup of the fundamental and the second harmonic will be absent as the node will be above the pup. The Pickup itself will have its own frequency response and where you pick the string will also affect the way it vibrates.

However even if this is one bass, through one amp and speaker it ties in with what you'd pretty much expect. I hope we do get some more information, my Swedish is non-existent so I'll have to wait.

Could you do your own measurements Stevie?

[stevie]

If I weren't expecting the builders to come in and demolish our internal chimneys tomorrow, I would have a go, but I'm going to be a bit tied up for the next few days. Apart from that, I don't have a 5-string.

I don't think it would be that complicated to do. You'd have to play pink noise through a hi-fi system or a bass amp and use a RATA to equalise it until it's flat between 30Hz and 200Hz. Then play your B and measure what you get on your RATA. That should represent the input the speaker receives for each frequency band. Perhaps somebody on here could give it a go......

It's well known that the energy of the bass guitar is concentrated in the first two harmonics rather than in the fundamental, and that applies to plenty of other instruments too. The second graph tks.se posted shows that the fundamental accounts for just 2% of the total energy of the note, which strikes me as rather low, unless an unequalised speaker system was used for the measurements.

Edited October 2, 2014 by stevie

[Phil Starr]

Tks talks about a cheap 10" speaker, though it isn't clear if all the tests took place with the same speaker. The implications for our design and any other designer are that we can take a few more liberties, which we kind of knew. I like to explore the extremes though so we can predict behaviour.

I hope that what people are getting from this is that though the maths gives you the predictions there is lots of room for setting the compromises in different places.

For those who don't know 6V6 built an earlier design based on the SM212 which is very similar to our design so his comments on his speaker are absolutely based on experience and relevant to our design

[stevie]

[quote name='6v6' timestamp='1412242726' post='2567012']
Interesting update!

FWIW in my build I made the fixed shelf tuned to ~50Hz, and later I added a removable extension which moves it down to ~42Hz tuning.

It's fair to say the difference is subtle, but audible, and since I fitted the shelf extension I'm getting occasional ugly sounds on hitting the low notes (only a 4 string but I'm running a 300w LM3 into one speaker) - it's interesting that you note port noise, I've not checked my plots but perhaps that explains it.

[/quote]

My guess is that your'e hitting xmax at 70/80Hz. If you look at Phil's excursion diagrams above you'll see that dropping the tuning to 40Hz reduces the power handling of the system quite a lot between 50 and 80Hz. If the problem didn't occur when your system was tuned to 50Hz, it's not likely to be noise, because lowering the tuning to 40Hz will move the turbulent noise from the ports down out of harm's way.

[tks.se]

Phil is right,

The measurements is an indication only, the intention was to look at the distribution of harmonics, compared to the fundamental. The test was done with an electric bass (Ibanez SR900) plugged in to a spectrum analyser through a simple mixing board (set flat). I can't say what the reference level is, and it isn't really relevant, since the difference in dB between the harmonics was the interesting thing to look at. The test was done quite quickly and a long time ago, before my first DIY project, when I wanted an answer to the exact thing Phil was looking about - should you assume that everything goes out through the fundamental (31 Hz in this case)? I popped over to the electrical engineering department at the university, borrowed the equipment while it was available for a short while, plugged in the bass and did some testing. The results were consistent, regardless what note was played, plucking style, bass eq settings etc, there were always harmonics that measured above the fundamental, often 4-5 of them. Data was just recorded for one note, the open B-string, as pictured. Again, disregard the reference level. As mentioned before: other basses/playing styles etc would shift the distribution, but the thing to focus on is still that it's not a single 31 Hz sine wave that goes out when you play a low B, which is what many bass players seem to believe, I used to be one of them (before educating myself).

The test with the 10" speaker was separate from the measurement, but that test confirmed the same thing, it's easier to get a lot of excursion from the harmonics than the fundamental. Picture a chart plotted from excursion vs frequency. If you start at the higher frequencies, the excursion rises as the frequency gets lower, but is shifted downards again when you get close to the tuning frequency, and then sharply up when you've passed the tuning frequency. The "hill" you get above the tuning frequency (highest excursion per signal strength, disregarding frequencies below the tuning frequency) is what the test with the 10" speaker described before refers to. In that case, the excursion of the fundamental (34 Hz in that case) was calculated to be higher than that of the ~68 Hz harmonic. Let's assume that it was the fundamental that made the voice coil former hit the back plate. Playing even lower notes would mean even lower fundamental frequency, which means even higher excursion, as we move further below the tuning frequency. However, going lower made the problem go away: less excursion. This means that the fundamental couldn't have been responsible, since when playing lower notes, the fundamental requires more excursion, while the harmonic is shifted to a frequency requiring less excursion.

Wow, don't know if that made any sense, keep in mind that English is my third language, and trying to explain anything technical becomes a challenge, I don't know how well it translates.

This is relevant for the SM212 as well. We've had two prototypes based around the SM212, so we have real life experience of it too, hence the earlier "nice speaker"-comment.

Everything written in this thread was only meant as a friendly tip, not in any way telling you how to run your project. I'm sorry if any of you took it that way.

[Phil Starr]

Your English is fantastic, the only evidence of it being a second language, never mind a third is that you are slightly more polite than we are with each other Please don't worry about sharing ideas either, it is good to share and I certainly don't have all the answers. Any suggestions you make are welcome.


Thanks for clearing up how you generated your data. It is something I had wanted to do and it would be good to try a range of basses. It would be nice to see what the difference between a P and J bass actually was for instance. It is always good to have an input from people who are actually manufacturing speakers. I wish you every success. For those who haven't noticed TKS are already on their own thread and you can see some of their speakers here [url="https://www.facebook.com/pages/tks-Engineering/154937774584383"]https://www.facebook...154937774584383[/url] .

Edited October 3, 2014 by Phil Starr

[6v6]

[quote name='stevie' timestamp='1412270572' post='2567394']
My guess is that your'e hitting xmax at 70/80Hz. If you look at Phil's excursion diagrams above you'll see that dropping the tuning to 40Hz reduces the power handling of the system quite a lot between 50 and 80Hz. If the problem didn't occur when your system was tuned to 50Hz, it's not likely to be noise, because lowering the tuning to 40Hz will move the turbulent noise from the ports down out of harm's way.
[/quote]

Aha, I'll have to check my own plots to confirm, but given it's the same speaker in a similar sized box you may well be right.

I'll return the tuning to 50Hz before the next gig to hopefully avoid causing any driver damage [size=4] [/size]

[tauzero]

[quote name='tks.se' timestamp='1412160565' post='2566219']
When playing a low B, most of the tone consists of harmonics to the fundamental. Here's the result of playing a low B and running the signal through a spectrum analyser:


[/quote]

That's an interesting result, and one which I'm rather surprised by. I was expecting a succession of even harmonics and not much in the way of odd, but the odd harmonics are there in force. The reason I'm surprised is that with even harmonics there's a node at the 12th fret, while with odd harmonics there's an antinode there. Mind you, the fundamental also has an antinode there, so perhaps I shouldn't be so surprised.

[Chienmortbb]

I mentioned Walter Harley's article on Bass Signals. It can be found here

http://cafewalter.com/2000/04/20/the-nature-of-electric-bass-signals/

[tauzero]

Also very interesting. Presumably the high initial peak for a half cycle is because that's the half cycle where the string is struck, so it's starting with a larger displacement than it can ever get to when it's vibrating freely, and the amplitude of the signal is proportional to the string velocity, which is highest when the string is going past the midpoint of its travel immediately after the string has been plucked. Easy to work out when you see the answer in the back of the book...

[Bill Fitzmaurice]

You'll see a shift in the fundamental to harmonic ratio as you go up the scale. A vibrating source works best where its dimension is 1/4 wavelength. With an open string 36 inches long 1/4 wavelength is 94Hz. The fundamental of the low B would be stronger if your B string was 1/4 wavelength long, but at 113 inches that would be somewhat impractical.
This shift explains why the same notes played at the lower end of the fretboard and the upper end of the fretboard sound slightly different. The shorter string length will have less fundamental content. I avoid playing a note on a lower string fretted above the seventh fret in favor of the same note played on a higher string for this reason. For that matter I avoid going above the seventh fret at all unless I have to. I play bass, not guitar.

[quote]I was expecting a succession of even harmonics and not much in the way of odd, but the odd harmonics are there in force.[/quote]That's normal. There is a difference in the ratio of even versus odd harmonics when different classes of amps are pushed to clipping. This in part explains why tube and SS sound different.

Edited October 4, 2014 by Bill Fitzmaurice

[LukeFRC]

[quote name='Bill Fitzmaurice' timestamp='1412426978' post='2568847']
That's normal. There is a difference in the ratio of even versus odd harmonics when different classes of amps are pushed to clipping. This in part explains why tube and SS sound different.
[/quote] that's interesting (as far as I understand it)

[tauzero]

[quote name='Bill Fitzmaurice' timestamp='1412426978' post='2568847']
You'll see a shift in the fundamental to harmonic ratio as you go up the scale. A vibrating source works best where its dimension is 1/4 wavelength. With an open string 36 inches long 1/4 wavelength is 94Hz. The fundamental of the low B would be stronger if your B string was 1/4 wavelength long, but at 113 inches that would be somewhat impractical.
This shift explains why the same notes played at the lower end of the fretboard and the upper end of the fretboard sound slightly different. The shorter string length will have less fundamental content. I avoid playing a note on a lower string fretted above the seventh fret in favor of the same note played on a higher string for this reason. For that matter I avoid going above the seventh fret at all unless I have to. I play bass, not guitar.
[/quote]

Being experimentally minded and not having delved into the theory, if I want a note to sound deeper, I play it on a lower string further up the fretboard, and it works. Of course, there's not just the string length to consider, you're also moving the midpoint of the vibrating length of string closer to the pickups, and for that matter, if you don't change your right hand position, you're plucking closer to the midpoint of the vibrating length.

[Bill Fitzmaurice]

[quote name='tauzero' timestamp='1412434928' post='2568948']
Being experimentally minded and not having delved into the theory, if I want a note to sound deeper, I play it on a lower string further up the fretboard, and it works. Of course, there's not just the string length to consider, you're also moving the midpoint of the vibrating length of string closer to the pickups, and for that matter, if you don't change your right hand position, you're plucking closer to the midpoint of the vibrating length.
[/quote]There's truly deeper, and there's perceived as deeper. The shorter string length will accentuate the harmonics, and as they occur where both speakers and ears work better it will seem deeper. To really appreciate what happens when you play as low as possible on the neck you need speakers that work really well down to 40Hz or so, and most don't. For instance, as a young lad I (and everyone else for that matter) avoided going below A like the plague, as my speakers couldn't handle it and it sounded bad when I ventured down low. When I abandoned store bought speakers for my own creations that were happy playing an open E full tilt I changed the way I play. For instance, playing a typical blues riff in 'C' I don't go up to the third and fifth, I go down, playing them off the F and G on my E string.

[LukeFRC]

the three musketeers, bill and others will probably have this/not need it... but someone linked to it on TB and it may be interesting for some... https://docs.google.com/file/d/0B_l73GVBBlIUNTc2YTc2YTktNTNmYi00YWNmLTlmY2ItZTU0MTRhMzdkYTAy/edit?pli=1

[Phil Starr]

[quote name='LukeFRC' timestamp='1413217846' post='2575968']
the three musketeers, bill and others will probably have this/not need it... but someone linked to it on TB and it may be interesting for some... [url="https://docs.google.com/file/d/0B_l73GVBBlIUNTc2YTc2YTktNTNmYi00YWNmLTlmY2ItZTU0MTRhMzdkYTAy/edit?pli=1"]https://docs.google....YTAy/edit?pli=1[/url]
[/quote]

Yep it's a great book and a good thing to have around if you want to delve a little deeper and understand some of what the computer is up to when it calculates box sizes for you. The other book I'd recommend is http://eu.wiley.com/WileyCDA/WileyTitle/productCd-0470094303.html though it is more technical in places.

[Owen]

http://www.hypex.nl/product/2012-11-23-13-44-01/2012-11-23-13-46-04/psc2-700.html