Skip to main content

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: Advice on small loudspeakers (Read 12708 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Advice on small loudspeakers

I am projecting a small pair of loudspeakers.
The use will be with a computer. Perhaps with an amp of 20W inbetween the two.

I have 2x100W 2" tweeters and 2x40W 4" car loudspeakers. Brands are good enough.
The enclosure is about 220mm (height), 142mm (width) and 143,8mm (depth).
It will be constructed with 9mm MDF.

Big question right now:

Do I go for the sealed design or do I go for the ported design?
I already know the pros and cons but I heard that some are particularly better with different kind of application.
Please answer ASAP, because I will be finishing the CAD models tonight and tomorrow I have to hand over the cabinetmaker.

Advice on small loudspeakers

Reply #1
Often different types of drivers are used in ported vs. sealed enclosures. Did you take it into account?

Are you aware of Thiele & Small parameters? Do you know their values for your drivers? Did you calculate frequency response for both ported and sealed enclosure? Do you know that the volume of the enclosure is a factor in the frequency response? Do you know that the tuning of the port (defined by the length and diameter of the tube) is also a factor?

I am sorry, but from your post it looks like you should learn a bit more about loudspeaker design before you finalize your CAD models.

If I misread you, apologies.
Ceterum censeo, there should be an "%is_stop_after_current%".

Advice on small loudspeakers

Reply #2
If you know the Thiele-Small paramaters you can use software to predict the bass response with a sealed or ported design.    But with car speakers you rarely get that information.  There is gizmo you can get for $100 USD that will measure the paramaters.    Although that's cheap compared to the equipment required to do it "the old-way", you'd probably be better-off spending the money on woofers that come with specs.

WinISD is a handy FREE speaker design application.

If you are "shooting in the dark", you'll usually get the best results with a sealed box that's as big as you can reasonably make it.    Ported design is tricky and you are more-likely to make a bad random ported design than a bad random sealed design.  And, a bad sealed design is likely to sound better than a bad ported design.  (A good ported design is likely to sound better than a good sealed design, but depending on the drivers a sealed design might be optimum.)

Also if your box size (volume) is locked-in, it's unlikely that you can optimize a ported design.   

Another option is to add a random port and just experiment with different port-lengths, and experiment with sealing-up the port.

Quote
I have 2x100W 2" tweeters and 2x40W 4" car loudspeakers.
Do you have a crossover network?    At a minimum, you need a capactor to block bass from your tweeters.

Advice on small loudspeakers

Reply #3
Ok, thanks for answers.

Well... I am not aware of those parameters really.

The box is really small. Is a bit larger than a general desktop loudspeakers. The speakers almost take the entire face of the box.
So it's no way like those home theater 5 feet towers where you can do a huge vent tube. It's fairly small.
I could replace these drivers on a later date.

See CAD  model:
http://www.hydrogenaudio.org/forums/index....ost&id=7337


Advice on small loudspeakers

Reply #4
The speakers almost take the entire face of the box.
This does not preclude using ported enclosure The port might be on the back of the speaker, or two smaller ports could be used like in Yamaha MSP3, or port that is not circular, like in KRK Rokit.

But if I were you, I would take a step back, and first learn a bit more about speaker construction. Right now you are like someone who knows that a computer has a mobo, a cpu, memory etc, and just buys some random components and tries to put them together, even though he does not know about voltages, sockets, bus frequencies, etc.

Of course, such computer simply will not work, and your speakers have a good chance to make some sound, but likely it won't be great sound, unless you are really lucky.


Several points that have not been made clear yet.

Not only you will need a crossover, but for sure an amplifier as well (not "perhaps").

Tweters must be closed-back so that pressure changes due to woofer operation do not affect them (normal tweeters usually are, but I know nothing about car speakers).

Usually some dampening material is used within the enclosure.


Replacing drivers later is probably not the best solution, because your box should be designed to work with particular drivers.

Once again, you choose your drivers and crossover design, then you enter the drivers' Thiele and Small parameters and the crossover design (basically, filter slopes) into a speaker design program, and then you can adjust box volume (and port parametert, if ported) until the frequency response looks good. That's how these programs usually work, I don't know the particulat program recommended by DVDdoug. And only after you have that, you design the dimensions of the enclosure.
Ceterum censeo, there should be an "%is_stop_after_current%".

Advice on small loudspeakers

Reply #5
I get what you mean about not knowing RAM Bus, CAS, Frequencies vs. HT Frequency, CPU Clock and so on.
Yeah I don't know much about setting up a box. I made one once, it works fine.
I am not looking to build some kind of studio monitors or professional grade equipment. Just some decent loudspeakers.
I just wonder if all the meticulous details are necessary... and even with that I still need to be "lucky"? No way...

Ok here is the add-on info:

A crossover would be something interesting. If I had a scheme, perhaps I'd build it. There's so many schemes... Where to start?
An amplifier will be made after this (I am not sure if I need it right now...) Need amp specs and all of that? (One thing for sure is that it outputs 4 Ohms).
Tweeters are closed back.
The dampening material hasn't been forgotten.
The tweeter capacitor is not forgotten.

It is not clear to me if you're saying that the test must be peformed with the already constructed box with speakers or merely theoretically with software or both.

I also am not sure if "drivers" are a synonym for speakers. Or is it a different kind of speaker?
Anyway, the car speaker and tweeter don't have much of details...
So how do I go on if I don't have more details on this?

Car speaker:
Imp: 4 Ohms
Power (RMS): 50 W
SPL: 86 (dB/2.83 V/m)
Freq. Resp. 100 up to 8.000 Hz

Tweeter:
Imp: 4 Ohms
Sensitivity (2,83V a 1M): 108dB max. (0dB = 20µPa);
Max. Voltage (Pink Noise): 13V rms;
Power (RMS): 40 to 100W
Freq. Resp. 5KHz to 20 KHz;
Sound Dispersion: 90°.

Advice on small loudspeakers

Reply #6
Ok, I realized I just overlooked some of the things said. Apologies.
Well, this is too overwhelming right now. I don't have the time to seat and think about this, spend money on software, or dig details for cheapo drivers.
Anyway thanks for clarifications, it is understood enough what it takes to get a proper sounding loudspeaker.

Advice on small loudspeakers

Reply #7
One more observation (not that there aren't more observations to make or answers to give, but looks like you are giving up anyway)

Car speaker:
SPL: 86 (dB/2.83 V/m)

Tweeter:
Sensitivity (2,83V a 1M): 108dB max. (0dB = 20µPa);


It appears that when driven by a signal of the same amplitude, your tweeter will be 22 dB louder than the woofer. Which is a lot. This has to be compensated in the circuit. The minimal solution is placing an appropriate resistor in series with the tweeter, but it has also influence on how the crossover will work.
Ceterum censeo, there should be an "%is_stop_after_current%".

Advice on small loudspeakers

Reply #8
Well, about sealing and ported, I gave up. The wood worker already cut them and assembled them.
I would dedicate all my time to "build" one appropriate. It is just that I don't have the time right now.
I need to concentrate a lot of time on my academic thesis right now and just the thought of giving further detailed attention to this would take away the focus.

I have seen a video in YouTube that pretty much made me decide to go with sealed.
http://www.youtube.com/watch?v=v0OQ_y0F9Qk
I am not sure how expert the guy that made the video is. But after seeing the ported results, I won't take any chances.

But I am willing to go through a cross-over and circuits.
Have got any ideas of how to minimize the tweeter problem? What resistor would be appropriate? What about the capacitor? Does it go with resistor?

Any ideas for a cross-over?

Any advice is really welcome

I know how to build small amps with TDA I.C.s and am familiar with circuitry symbols.

Thanks a lot for the info. Please

Advice on small loudspeakers

Reply #9
The capacitor will serve as a simple high-pass filter for your tweeter. A minimal crossover. For a more proper crossover, you should also low-pass filter the woofer. The crossover will do two things: 1. protect your tweeter from damaging currents intended for the woofer, 2. separate the low and high frequencies for flat frequency response. The frequency ranges of your drivers (this is just a name for speakers, I mean component speakers like a woofer or a tweeter) overlap in the 5kHz-8kHz range. So even if your tweeter was balanced with (had the same sensitivity as) the woofer, in the overlap range they would work together and the overlap range would be too loud.

For help with designing your circuits and calculating the component values check out this page.

You probably will need the "2-way crossover designer" and the "Driver attenuation circuit". But I would also recommend reading the relevant tutorials.






Ceterum censeo, there should be an "%is_stop_after_current%".

Advice on small loudspeakers

Reply #10
Ok, so I have here the 2-way crossover calculator.
http://www.diyaudioandvideo.com/Calculator/XOver/

Which value should I supply the frequency in hertz?

Another thing is that there are dozens of crossovers types... which one would be the probable best bet?

The driver attenuation circuit goes only for tweeter? Would it be placed before or after, or in series with drivers?

Power per channel is the max power written in the driver specs? Not the amp?

Advice on small loudspeakers

Reply #11
Ok, so I have here the 2-way crossover calculator.
http://www.diyaudioandvideo.com/Calculator/XOver/

Which value should I supply the frequency in hertz?
The value at which you want the crossover to split the signal into the low/woofer and high/tweeter part. Obviously it has to be in the drivers' overlap range, so somewhere between 5 and 8 kHz.

Another thing is that there are dozens of crossovers types... which one would be the probable best bet?
1st order filters are the simplest to build and have the gentlest filter slopes, which make the transition between woofer and tweeter gradual and the two ranges may blend better. But it requires the drivers to be reasonably flat far beyond the crossover point, i.e., beyond their intended range. With higher orders, filter slopes become steeper so you don't need the drivers to be flat over large frequency ranges. But the design gets more complex and blending of the ranges may be more difficult. The choice of crossover point and order should be also informed by the natural rollover of the drivers, which is apparently unknown in your case. I mean, you have numerical ranges of frequency response, which even don't specify whether the values are at -3 dB or at -10dB or what. Normally you should work with frequency response plots.

As for different types of crossovers within the same order, I cannot give advice. Let me just say, that 4th order Linkwitz-Riley seem to be pretty popular in active speakers, while passive speaker seem to often use 2nd order, sometimes 1st order (but I've seen 4th order Linkwitz-Riley as well).

In your case I would probably shoot for 2nd (maybe 1st) order, given how crude your design is anyway, it may matter very little. Note that for 2nd order (and, I think, 6th order as well) it is often recommended to invert the polarity of one speaker.

The driver attenuation circuit goes only for tweeter?
Of course. You want to match the sensitivity (i.e., sound intensity produced from the same-intensity electrical signal) of your drivers. Your tweeter is too loud, so you have to attenuate it. No reason to attenuate your woofer.

Would it be placed before or after, or in series with drivers?
Didn't you try out the calculator? It draws the circuit. One resistor is in series, one in parallel with the driver. No matter whether the series resistor goes "before" or "after".

Power per channel is the max power written in the driver specs? Not the amp?
It's the max power you want your tweeter to handle before the resistors catch fire ;-). The attenuation circuit will basically take away some of the power sent to the tweeter and dissipate it as heat through the resistors. The only reason you have to enter the power is that the calculator will give you minimum power ratings of the resistors. Actual power ratings may (and should) be higher than that.
Ceterum censeo, there should be an "%is_stop_after_current%".

Advice on small loudspeakers

Reply #12
Quote
Obviously it has to be in the drivers' overlap range, so somewhere between 5 and 8 kHz.


Is 6500 Hz, the middle, a good bet?

I read on that site that uneven orders are the recommended (so it would be either 1st or 3rd), that would avoid the polarity inversion trick.

Quote
Didn't you try out the calculator? It draws the circuit.

Yeah, I kinda got two circuits, the crossover have inductors and capacitors, and the attenuator have got resistors.

Quote
The attenuation circuit will basically take away some of the power sent to the tweeter and dissipate it as heat through the resistors.


It comes to my mind now that the box is sealed. And I just read the words heat & dissipation... It is not that I need ventilation, right?

Quote
The only reason you have to enter the power is that the calculator will give you minimum power ratings of the resistors. Actual power ratings may (and should) be higher than that.
Ok, so we have these 1W or 2W resistors which are huge... we're talking about those, right?
I entered 20W per channel, based on the amplifier power... I am not sure if I have done this right. However, I see weird results like one resistor being 18W. A 20W resistor would be the huge squared porcelain resistor. Is this more or less the deal?

I am attatching some results I got, entering the values:

http://www.hydrogenaudio.org/forums/index....showtopic=99400

Advice on small loudspeakers

Reply #13
I read on that site that uneven orders are the recommended (so it would be either 1st or 3rd), that would avoid the polarity inversion trick.
Yes, that's what the site says, however 2nd order (and 4th order, but these do not require polarity inversion AFAIR) are used pretty often anyway.

It comes to my mind now that the box is sealed. And I just read the words heat & dissipation... It is not that I need ventilation, right?
Then your box wouldn't be sealed anymore. You theoretically could use a radiator, this is used for amplifiers of more powerful powered speakers. In your case it should not be necessary though. First, the actual power delivered to the tweeter will only rarely be full 20W, unless you are actually trying to fry the tweeters (or resistors). Second, if there was no attneuator network, all 20W would go into the tweeter, which would also convert much of it into heat. Speakers are not very efficient and at best a few % of electrical power is converted to acoustic power, the resto goes into heat. So I don't think this is going to be a serious issue, although I don't guarantee this. A tweeter is exposed to the outside, so it can loose some heat more easily.


Ok, so we have these 1W or 2W resistors which are huge... we're talking about those, right?
I entered 20W per channel, based on the amplifier power... I am not sure if I have done this right. However, I see weird results like one resistor being 18W. A 20W resistor would be the huge squared porcelain resistor. Is this more or less the deal?
Your problems results from trying to match two drivers of very different sensitivity. I think that the proper solution is to find better-matched drivers. Or go with a large resistor.

Now, I had my try into bulding (or rather rebuilding speakers) years ago, I learned quite a lot then, and I passed most of I learned to you. I think it would better for you if you found a dedicated speaker/DYI forum and ask your specific questions there, like whether it makes any sense to try to match these drivers, or specific questions about circuits. People who actively build speakers at this time will answers such questions better than I ever could.
Ceterum censeo, there should be an "%is_stop_after_current%".

Advice on small loudspeakers

Reply #14
pawelq, thanks for all the info.
Yes, it can be trick to start from the beginning when constructing loudspeakers.
This is what I am going to do

Since I don't have much time to spend on this now, I will make an effort to not make any gross mistakes.
I am in the need of those speakers right now.

However, in the future, I will seat down and meticulously perform a construction from ground zero.

I went to the DIY forums to see if someone helps me out with the values for the circuitry.

Thanks again.

Advice on small loudspeakers

Reply #15
pawelq,

Please take a look at values for my midrange:
I just found it over the internet.

Name: 4HTS400

Fs = 115,1 Hz

Qms = 3,652

Vas = 0,99 liters

Cms = 0,221 mm/N

Mms = 8,66 g

Rms = 1,715 kg/s

Xmax = 2, mm

Xmech = 3, mm

P-Dia = 85, mm

Sd = 56,19 sq.cm

P-Vd = 0,0112 liters

Qes = 0,745

Re = 3,51 ohms

Z = 4,461 ohms

BL = 5,43 Tm

Pe = 50, watts

Qts = 0,618

no = 0,195 %

1-W SPL = 84,9 dB

2.83-V SPL = 88,68 dB

Advice on small loudspeakers

Reply #16
I am playing around with WinISD, and soon I will post here the results.

Advice on small loudspeakers

Reply #17
I am projecting a small pair of loudspeakers.
The use will be with a computer. Perhaps with an amp of 20W inbetween the two.

I have 2x100W 2" tweeters and 2x40W 4" car loudspeakers. Brands are good enough.
The enclosure is about 220mm (height), 142mm (width) and 143,8mm (depth).
It will be constructed with 9mm MDF.

Big question right now:

Do I go for the sealed design or do I go for the ported design?
I already know the pros and cons but I heard that some are particularly better with different kind of application.
Please answer ASAP, because I will be finishing the CAD models tonight and tomorrow I have to hand over the cabinetmaker.


Nothing wrong with the advice you've been getting, but just sayin'

If you want to talk with more people with a wider breadth of opinion and experience, try one of the do-it-yourself forums.

Do-it-yourself audio and audioholics come to mind.

Advice on small loudspeakers

Reply #18
Yes, I have been there already and am running a topic there.
I found here to be more friendly though.
It's good to know parallel opinions, and the opinions here have been equally good as there.

Advice on small loudspeakers

Reply #19
Sealed.

The fact that you have to ask the question tells me the right answer, for you. Why? Because a sealed, acoustic suspension design has a markedly larger margin of error in box design (internal volume) and is much simpler (it's just a box, no ports). You can goof up by as much as 5-10% in internal volume and still get great results, heck you can probably even goof up by 15-20% and at least have listenable results!

With a ported design, you don't have this luxury.

How might you goof up, you might ask? Say, as just one example, you follow all you instructions perfectly, but then notice a panel vibration at loud levels, you worry is adding a coloration to the sound. You can go back in and add reinforcements, such as corner buttresses or cross bars (to make your box less ringy) and not fear that these additions alter the effective, overall internal volume in an audibly negative way.

edit to add: Although not as efficient as ported designs, acoustic suspension designs usually have good bass extension, low distortion, and are more agreeable to adding EQ to eek out even greater bass extension Conversely, trying to EQ the bottom octave of a ported design to increase its low bass extension is often asking for trouble.

Advice on small loudspeakers

Reply #20
mzil,

I was playing with WinISD beta, and I entered the available values for my midrange to calculate a box.
The application calculated a 4,1 Litre box, measuring about 34cm (h) x 18cm (w) x 11,5 (d).

For my extreme luck, the box I have designed earlier hits on the nail as being a box that is 4,1 Litre as well. 22,4 (h) x 16 (w) x 16 (d).
I guess I haven't lost any of the already made cut of the woods.

However, the program is suggesting a ported design. Because the EBP is 154 Hz.  (EBP = fs/Qes).
Now you are coming and saying for me to go sealed, well, should I do this after getting the information from the application?

I would have to design a port on this box, either by adding a 1,4cm tube of 10cm diameter, by the back or in the front (the program recommends the front).
I could ask the wood maker to cut the circle and I would be able to insert a 4 mm hidraulic tube. I am not sure what the others will think of this. Perhaps they will suggest me to make the squared port in wood shape.

Anyways, another doubt I have is that theses calculations are based on the midrange.
I am not sure if the piezo tweeter comes into play on this. Obviously, in WinISD, if I specify 2 drivers, I would be talking about the same 4" midrange driver, then it would increase the box litre to about 8L. But I am not sure... I need advice on this.

Curious to hear what pawelq has to say about.


Advice on small loudspeakers

Reply #21
First of all, I would like to thank the folks who didn't give up on me but persevered to make me understand the whole thing. Specially Mr. pawelq.
I have gathered, by using WinISD, the values for my midrange driver. (I still don't know if I need to take the tweeter here into account).
Luckly it seems that I ended up with a box with the same required volume, however, I still have to make an intervention about the port.
There are also things to take into account...

a) The piezo tweeter tweak attenuator.
b) The crossover that will be 6000Hz/12dB cut.

I have contacted someone nearby that actually make inductors, so I should be getting the right components.
I just have to figure now if I go with 1st order crossover or not.
Please have a look at my results and carefully advice. I think it was worth the effort to dig this up.







Advice on small loudspeakers

Reply #22
I still don't know if I need to take the tweeter here into account


The only aspect of the tweeter that you need to take into account from the point of view of box dimensions is the space the tweeter takes - it reduces the internal space of the box, but it *should* be totally enclosed, so unaffected by the box. If not, you will probably have a problem with back pressure from the bass/mid affecting the tweeter membrane (in worst case blowing it out).

Advice on small loudspeakers

Reply #23
Car speakers are designed for cars.

Advice on small loudspeakers

Reply #24
Quote
f not, you will probably have a problem with back pressure from the bass/mid affecting the tweeter membrane


Tweeter is closed back. It is a piezo. Leson TLC1. Google for it and see how it looks.

Quote
Car speakers are designed for cars.


I agree. But aren't you being too extreme on this?
Someone gave me those drivers and I don't feel like trashing them away.
I also abolished music in car because it is a proven distracting factor for car accidents.