Ive been getting some good results designing ported guitar cabs, for 1x12 or 2x12. This is my first post, and Id like to seek opinions as to whether my approach makes sense, or just happens to work.
This is for those cases where having built the cab, a bit more bass is wanted - many cabs don't need it. but, if they do...
Bass for a guitar, only goes down to 81hz, or a bit lower if you drop-tune
Im using WinISD, to pick a port length, diameter and number of, to set a bump in the plotted response at 50hz. This is below the frequencies used, but the upper tail of this bump raises the usable low end by a couple of db, which is what is wanted. It works on ISD and it works by testing in practice.
If you then look to see what the calculated box tuning frequency is, it is not 50hz but about 60hz, because this does not account for the speaker compliance.
But, and here is the first question, if I add internal cab volume Vb, to the speakers equivalent volume Vas, and use that total to volume the port by a simple equation, then I get exactly the same port dimensions as with WinISD, with my intended actual resonance at 50hz. Is there any correct basis for adding Vas to Vb? because it seems to lead to the correct result.
My other question is, I have been adjusting the port diameter to make the port length = wood thickness, so it is physically just a hole. This also seems to work in practice. Any down sides to this? (Im aware of possible chuffing noises with ports too small, but have not found this)
Thanks for looking - hoping for any input
This is for those cases where having built the cab, a bit more bass is wanted - many cabs don't need it. but, if they do...
Bass for a guitar, only goes down to 81hz, or a bit lower if you drop-tune
Im using WinISD, to pick a port length, diameter and number of, to set a bump in the plotted response at 50hz. This is below the frequencies used, but the upper tail of this bump raises the usable low end by a couple of db, which is what is wanted. It works on ISD and it works by testing in practice.
If you then look to see what the calculated box tuning frequency is, it is not 50hz but about 60hz, because this does not account for the speaker compliance.
But, and here is the first question, if I add internal cab volume Vb, to the speakers equivalent volume Vas, and use that total to volume the port by a simple equation, then I get exactly the same port dimensions as with WinISD, with my intended actual resonance at 50hz. Is there any correct basis for adding Vas to Vb? because it seems to lead to the correct result.
My other question is, I have been adjusting the port diameter to make the port length = wood thickness, so it is physically just a hole. This also seems to work in practice. Any down sides to this? (Im aware of possible chuffing noises with ports too small, but have not found this)
Thanks for looking - hoping for any input
Ok, thanks for responding. Ill try to be clearer:
Here are some plots for a WGS Green Beret, in a cab of 2 cuft which is 56.6l internal volume. The lower one is a closed back cab, while the upper one has ports, chosen in the way that I have described above, to give a bump in the response at about 50 hz. You can see from the data below the trace, that to do that, I have picked a resonant frequency of the box in WinISD of 60.78hz, even though the actual result is around 50hz.
When I plug numbers into a port formula, such as this one from DIYsubwoofers..
The Subwoofer DIY Page - Port Calculations
and if I just use the box volume, I get the same ports as a result, if I use the WinISD box frequency of 60.78hz. But, if I add the box volume Vb of 56.6l, to speaker effective Vas 33.5l, and put that (90.1l) into the formula, then i can use the frequency that I want, which matches the WinISD plot, ie 50hz.
So, the question is, is there a sound basis for adding Vb to Vas like this, for use in the formuls for predicting actual resonant frequency? it seem to me like there should be but Im interested in any comments. If its valid, then I can quickly work out port dimensions, based on box volume, adjusted for the speaker Vas.
Here are some plots for a WGS Green Beret, in a cab of 2 cuft which is 56.6l internal volume. The lower one is a closed back cab, while the upper one has ports, chosen in the way that I have described above, to give a bump in the response at about 50 hz. You can see from the data below the trace, that to do that, I have picked a resonant frequency of the box in WinISD of 60.78hz, even though the actual result is around 50hz.
When I plug numbers into a port formula, such as this one from DIYsubwoofers..
The Subwoofer DIY Page - Port Calculations
and if I just use the box volume, I get the same ports as a result, if I use the WinISD box frequency of 60.78hz. But, if I add the box volume Vb of 56.6l, to speaker effective Vas 33.5l, and put that (90.1l) into the formula, then i can use the frequency that I want, which matches the WinISD plot, ie 50hz.
So, the question is, is there a sound basis for adding Vb to Vas like this, for use in the formuls for predicting actual resonant frequency? it seem to me like there should be but Im interested in any comments. If its valid, then I can quickly work out port dimensions, based on box volume, adjusted for the speaker Vas.
Hi,
Apparent success is nothing like doing the job properly.
Your guessing and you are mostly wrong, in all
respects in regards to good guitar cabinets.
Take the WGS Green Beret.
Its a complete nonsense to put such a driver
in a cabinet twice the Vas, in all practical
cases twice the Vas is enough for for a
relatively outstanding sealed 4x12.
rgds, sreten.
Apparent success is nothing like doing the job properly.
Your guessing and you are mostly wrong, in all
respects in regards to good guitar cabinets.
Take the WGS Green Beret.
Its a complete nonsense to put such a driver
in a cabinet twice the Vas, in all practical
cases twice the Vas is enough for for a
relatively outstanding sealed 4x12.
rgds, sreten.
Hi sreten. I see you are an eminent and long standing member here. But I do not agree with your statements.
I am not just 'guessing' nor, given that my cabs have been built, tested, measured, and found to work am I 'mostly wrong'. Nor is a 2cuft cab for a 1x12 'complete nonsense'. It is actually quite a common volume for this type of cab. 4x12 guitar cabs are usually between 5 and 5.5 cuft. I bit more volume per driver is helpful in a 1x12 cab.
However, if you wish to provide any helpful comment I will be genuinely grateful. I am seeking to understand the relationships that affect these cabs, which I think have different design priorities than hifi cabs.
I am not just 'guessing' nor, given that my cabs have been built, tested, measured, and found to work am I 'mostly wrong'. Nor is a 2cuft cab for a 1x12 'complete nonsense'. It is actually quite a common volume for this type of cab. 4x12 guitar cabs are usually between 5 and 5.5 cuft. I bit more volume per driver is helpful in a 1x12 cab.
However, if you wish to provide any helpful comment I will be genuinely grateful. I am seeking to understand the relationships that affect these cabs, which I think have different design priorities than hifi cabs.
Hi johndh .
I see you gained some 1.5dB around 100Hz, a frequency that matters in the Guitar world.
Cool.
That said, curve is already dropping at that frequency, even with that boost it's 1.5dB down from average speaker output and 5 or 6dB below that nice bump you got at around 150Hz.
If you have the time, I'd ask you to recalculate rising the nominal box tuning frequency in, say, 10 Hz steps, and see what happens.
I'm particularly interested in finding 2 values:
a) what would the bump be if the box (or system) were tuned at 100hZ .
I guess it would be chest punching bass from a quite small cabinet.
We're not striving for Hi Fi or flat response here , bigger than life is often good (just ask Kiss make up artist 😉 )
of course, it might be a little too much, so I'm also interested in:
b) what box tuning frequency to choose (guess around 80Hz but you tell me) to move that peak down in frequency to 100Hz , which will also tame it a little.
Thanks for experimenting and posting 🙂
I see you gained some 1.5dB around 100Hz, a frequency that matters in the Guitar world.
Cool.
That said, curve is already dropping at that frequency, even with that boost it's 1.5dB down from average speaker output and 5 or 6dB below that nice bump you got at around 150Hz.
If you have the time, I'd ask you to recalculate rising the nominal box tuning frequency in, say, 10 Hz steps, and see what happens.
I'm particularly interested in finding 2 values:
a) what would the bump be if the box (or system) were tuned at 100hZ .
I guess it would be chest punching bass from a quite small cabinet.
We're not striving for Hi Fi or flat response here , bigger than life is often good (just ask Kiss make up artist 😉 )
of course, it might be a little too much, so I'm also interested in:
b) what box tuning frequency to choose (guess around 80Hz but you tell me) to move that peak down in frequency to 100Hz , which will also tame it a little.
Thanks for experimenting and posting 🙂
Hi JMFahey - thanks for your comments.
That stepped plot is quite easy to make, here it is, tuning the system in approximate 10hz steps from 50 to 100hz:
You can see that the calculated bump in response keeps rising. In going above where I sit it at 50hz, I would start to get worried about excessive prominence of the low mids - you can see it rising through the plots. Also, by putting it safely well below actual guitar frequencies, my belief is that the results are not too critical and the desired small extra rise in low bass will not come with too many side effects, even if box and speaker parameters vary a bit.
I first did this approach on my own cab, with a Celestion V30 in it. As a closed box, it was good but just wanted a bit more lows, and this method worked very well. I could hear it clearly, testing by blocking and unblocking the ports, and the results could also be seen when recording the response with white noise and plotting it.
That stepped plot is quite easy to make, here it is, tuning the system in approximate 10hz steps from 50 to 100hz:
You can see that the calculated bump in response keeps rising. In going above where I sit it at 50hz, I would start to get worried about excessive prominence of the low mids - you can see it rising through the plots. Also, by putting it safely well below actual guitar frequencies, my belief is that the results are not too critical and the desired small extra rise in low bass will not come with too many side effects, even if box and speaker parameters vary a bit.
I first did this approach on my own cab, with a Celestion V30 in it. As a closed box, it was good but just wanted a bit more lows, and this method worked very well. I could hear it clearly, testing by blocking and unblocking the ports, and the results could also be seen when recording the response with white noise and plotting it.
I use a Marshall DSL and sometimes a Crate PowerBlock. Its not specific though, any kind of guitar amp. Others who have built cabs with this spproach use whatever they have.
Also, Im not saying that this necessarily makes a cab better. Only that if a given cab is judged to be wanting of a bit more bass, then this can be a rational and effective way to do it.
Also, Im not saying that this necessarily makes a cab better. Only that if a given cab is judged to be wanting of a bit more bass, then this can be a rational and effective way to do it.
Thanks for doing the simulations and posting. 🙂
Not the results I expected 🙁 but , oh well, reality rules.
I expected the peak to go down somewhat in frequency but I guess that the high Q rigid suspension speaker kept it firmly on its free air resonance.
I guess you measured with a standard high damping SS amp.
Thanks again. 🙂
Not the results I expected 🙁 but , oh well, reality rules.
I expected the peak to go down somewhat in frequency but I guess that the high Q rigid suspension speaker kept it firmly on its free air resonance.
I guess you measured with a standard high damping SS amp.
Thanks again. 🙂
Hi,
The priorities are indeed different, generally as loud as
possible, as small as possible and as light as possible.
Drivers are never put in cabinets twice their Vas,
an absolute sensible maximum is the same as
Vas for vented and about half Vas for sealed.
The maximum output of an open Bass E string is at
80Hz, the 2nd harmonic, and many bass cabinets
are designed to peak around this area, rather
than being concerned with maximum 40Hz level.
Tuning vented bass cabinets is moreorless choosing
E, or drop D, or low B (5 string) for the tuning.
For guitars simply double the frequencies. 50Hz
tuning in a cabinet twice Vas volume is way off.
Guitar cabinets also come in the open backed
variety, for higher Q guitar speakers which
have no bass but a large hump in the mid
around 150Hz to 250Hz depending on size,
centred on the baffle "wrap" frequency.
There are also for PA so called "power handling"
alignments, where the port barely changes the
response, but allows higher SPL at the tuning,
where the driver may be running out of excursion,
this also needs highpassing just below the vent
frequency, to prevent driver overexcursion.
Drop D tuning is probably the best for a vented guitar cab.
Note that for peaking responses and peaking
open backed cabinets a simple default 1st order
high pass in the amplifier can flatten any low
mid excess, and leave a default with all controls
set flat of a bass rolloff below about say 150Hz.
You then design the amps EQ around that response.
(And the typical very large presence peak that
sets guitar (+ bass) speakers apart from PA.)
I've no doubt that driver in 2cuft tuned to 50Hz
for an extended shelf bass alignment would
sound good, but its simply not road practical,
compared to say a much smaller open back.
Noting a 2x12 is 3dB more efficient, doubling
SPL, and a 4x12 6dB more efficient than 1x12.
Including power handling a 2x12 goes 6dB
and a 4x12 12dB louder than a 1x12.
rgds, sreten.
Last edited:
An assumption made in the T/S theory is that the driver is operating as a piston in a linear range below Xmax. For most guitar speakers Xmax is very small or zero. The voice coil is about the same length as the magnetic gap and cone stiffness is often compromised in favor of efficiency.
Another assumption is that the speaker is driven by a voltage source, meaning an amplifier with a high damping factor. Most guitar amplifiers don't have a very high damping factor. Many solid state amps deliberately reduce the damping factor with mixed voltage and current feedback. The software can adjust for this if you increase the value of Re, the voice coil resistance by 1 to 4 ohms.
Another assumption is that the speaker is driven by a voltage source, meaning an amplifier with a high damping factor. Most guitar amplifiers don't have a very high damping factor. Many solid state amps deliberately reduce the damping factor with mixed voltage and current feedback. The software can adjust for this if you increase the value of Re, the voice coil resistance by 1 to 4 ohms.
Hi,
True Xmax is generaly very small or zero, that gives a very
gradual overload characteristic, but no clean bass output.
Yes designing speakers for valve amplifiers with high
output impedances does complicate matters a lot.
mh-audio.nl - Home
The software will not adjust Qts if you change Re,
simpler to change Qts according to the above.
The 12" 100Hz Fs 1cuft Vas 0.94 Qts driver becomes
most suitable for open backed cabinets where high
amplifier output impedance drives up the Qts.
rgds, sreten.
True Xmax is generaly very small or zero, that gives a very
gradual overload characteristic, but no clean bass output.
Yes designing speakers for valve amplifiers with high
output impedances does complicate matters a lot.
mh-audio.nl - Home
The software will not adjust Qts if you change Re,
simpler to change Qts according to the above.
The 12" 100Hz Fs 1cuft Vas 0.94 Qts driver becomes
most suitable for open backed cabinets where high
amplifier output impedance drives up the Qts.
rgds, sreten.
Last edited:
Thankyou for the replies, which I have learnt much from. It seems that the T/S parameters start to become less valid for guitar speakers as volume increases. Also the points about amp damping are interesting. It would be good to try to make the corrections and see what the effect is, and more importantly whether it leads to different design decisions.
Sometimes an imperfect theory is still helpful if the limitations are understood. A 'flat earth' world model is perfectly fine for planning a route across town, but not for a round-the-world trip!
I want to establish that in discussing 1x12 cabs of about 2cu.ft, im not talking about anything unusual or impractical.
Avatar are a popular brand. Here is a closed-back 1x12 with 1.9cu.ft internal volume:
Contemporary 112
Here is a Marshall 1933 cab from the early 80's. It is about 1.7 to 1.8 cu.ft, and you can see a front port:
Marshall's current top 1x12 is a 1912 model. It is about 1.8cuft, ported:
1912 150W Cabinet | Products | Marshall Amps
This mesa 1x12 cab is about 2.1cu.ft:
MESA Engineering 1 x 12 Rectifier Guitar Cabinet
Here is a cab that I built:
guitarcab5.jpg Photo by JohnDHewitt | Photobucket
It is sized on the front face to match my combo amp, but deeper, more like the mesa above, with 2.2cu.ft. After a few variations, three rear ports tuned as I have described in this thread got it to sound how I wanted it.
This cab was built by a member on the Marshall forum, to my design:
IMG_9068.jpg Photo by OlleO | Photobucket
Its wider than mine, to match his DSL combo, and about the same volume. It has replaceable back panels to try ported, closed or semi-open.
Sometimes an imperfect theory is still helpful if the limitations are understood. A 'flat earth' world model is perfectly fine for planning a route across town, but not for a round-the-world trip!
I want to establish that in discussing 1x12 cabs of about 2cu.ft, im not talking about anything unusual or impractical.
Avatar are a popular brand. Here is a closed-back 1x12 with 1.9cu.ft internal volume:
Contemporary 112
Here is a Marshall 1933 cab from the early 80's. It is about 1.7 to 1.8 cu.ft, and you can see a front port:
An externally hosted image should be here but it was not working when we last tested it.
Marshall's current top 1x12 is a 1912 model. It is about 1.8cuft, ported:
1912 150W Cabinet | Products | Marshall Amps
This mesa 1x12 cab is about 2.1cu.ft:
MESA Engineering 1 x 12 Rectifier Guitar Cabinet
Here is a cab that I built:
guitarcab5.jpg Photo by JohnDHewitt | Photobucket
It is sized on the front face to match my combo amp, but deeper, more like the mesa above, with 2.2cu.ft. After a few variations, three rear ports tuned as I have described in this thread got it to sound how I wanted it.
This cab was built by a member on the Marshall forum, to my design:
IMG_9068.jpg Photo by OlleO | Photobucket
Its wider than mine, to match his DSL combo, and about the same volume. It has replaceable back panels to try ported, closed or semi-open.
That looks more like an opening for an optional tweeter, which would probably be of more use than a port. The lack of wadding also leads me to believe it's not a port.
I'm rather bemused by what you're trying to do? - guitar cabinets are very well tried and tested, and almost all of them (giving the sound that guitarists want) aren't ported.
I once built a ported cabinet for an EVM12L reissue. My experience was that the driver had such low Q and Vas, that a traditional ported alignment would end up with a very small cabinet and hardly any output below 120Hz. So I tuned the cabinet a bit lower, similar to the original Mesa Thiele cabinet.
http://scopeboy.com/speaker/evm12L_cab_design.jpg
http://scopeboy.com/speaker/IMGP1405_small.jpg
I tried various cabinet sizes and tunings. I found that a too big cabinet gave a horrible quality of bass, thin, weak and boomy all at once.
Guitar speakers vary widely. I'm sure many of them are even worse suited to a vented alignment than the EVM12L. There is a Rivera speaker shootout on Youtube where for some reason they tested all of the speakers in the same vented cabinet. Some of them have truly horrid sounding bass.
http://scopeboy.com/speaker/evm12L_cab_design.jpg
http://scopeboy.com/speaker/IMGP1405_small.jpg
I tried various cabinet sizes and tunings. I found that a too big cabinet gave a horrible quality of bass, thin, weak and boomy all at once.
Guitar speakers vary widely. I'm sure many of them are even worse suited to a vented alignment than the EVM12L. There is a Rivera speaker shootout on Youtube where for some reason they tested all of the speakers in the same vented cabinet. Some of them have truly horrid sounding bass.
Last edited:
I don't get this. If a 12" guitar speaker is 5% efficient, then 2 x 12" will still be 5% efficient (conservation of energy)?
Nope. There are about 10 different explanations of this, but they all amount to the same thing: a speaker's efficiency increases when it is surrounded by other speakers driven with the same signal.
Tweeters in guitar cabs sound truly horrible with any kind of driven sound.
Why are you bemused by what I am doing? It seems simple enough to me, using ports to add a little bass, if wanted. 'If' is the key word. It works, and it is not at all uncommon in small cabs. Not saying its needed on every cab. Its a matter of taste. My approach is keep the tuning low, then its not too critical for exact speaker properties and it doesn't get boomy, it just adds a little bit, if the user wants this.
Last edited:
Yes, I didn't believe it either - conservation of energy etc. So I tried it, setting up two similar cabs side by side and spitting the same power between them. It did indeed get measurably louder from the front. I reckon the payback in energy terms may be some associated cancellation of volume if you are at a skew angle to the group.
And maybe it explains why a 4x12, can sound 'bigger' than a 1z12, even with similar drivers and similar cab volume per speaker
- Status
- Not open for further replies.