DIY subwoofer drawings and diagrams 75gdn. Calculation and design of the box

...or Building a subwoofer with your own hands.

Step One - Making a Decision.
I've always been attracted to high-quality sound.
Assuming that I was not alone in this, I rushed into the abyss of the Internet in search of thematic sites and forums where I could get answers to the questions that interested me. This lasted for some time, until quantity turned into quality. Until, as it seemed, quite suddenly, I realized that making a car subwoofer myself was quite possible and even interesting.

Step Two - Measuring parameters.

One summer day, a 75GDN-1-4 dynamic head was purchased.
Why 75-GDN. Here I must say that I have been a fan of S-90 speakers since 1989 (one of my pair of copies, which has served me faithfully for 18 years (!), can be seen in the photo below). And no matter how much they criticize everything Soviet now, I would still put these speakers in line with the well-known achievements of the 20th century engineers of the country in which I was born. My opinion. Of course, the opinion of respected people who built subs on the 75-GDN dynamics and achieved high results in car audio competitions played a big role. Speaking of opinions. In a nutshell, there are two points of view in RuNet. First - People! Don’t do nonsense, you can’t make candy out of crap, buy a normal (imported/expensive) head and get an amazing sound. The second, actually, is not even an opinion, but the given facts, like someone somewhere made a sub on a 75GDN and it worked out, and it wasn’t even bad. I think both are right. By the way, let’s discard for now the category of people who simply buy a subwoofer in a store or on the market, who may be confused by the question of what speaker is in their subwoofer, not to mention the question of what frequency the bass reflex is tuned to.
In short, the speaker was purchased, the first financial contribution was made and it was too late to retreat.
It is well known that these speakers suffer from a large (to put it mildly) spread of parameters. And, naturally, I wanted to get the most out of him,
Therefore, I decided to measure the Thiel-Small parameters myself. They say that on imported speakers, in the documentation, of course, these parameters are already written, and even the recommended volume and type of box is given. This is undoubtedly a plus for them. But I came across articles on the Internet about testing dynamic heads, from which I realized that manufacturers often indicate incorrect numbers. So, in order not to throw money away, it is still advisable to measure these parameters.
How to do this is written in many places, but I liked this resource -. By the way, if I don’t say something here, you will find the missing information there. If I have time, maybe I’ll write later, like I did. The only thing I can say is that there is no need to be afraid of this procedure, everything is real and quite simple. All I got by with was a voltmeter, a computer with a sound card, an amplifier and a program - a sound signal generator, the link to it is at the end of the article.
So my instance data is
Fs=32Hz Qts=0.595
Fs – resonance frequency in open space;
Qts – total quality factor of the speaker;
Knowing these parameters, you can begin to select the type of subwoofer.

Step Three - Calculations.
As you know, there are mainly three types of acoustic design -
1) Closed Box ( ZY)
2) Phase Inverter ( FI)
3)BandPass ( BP)
If your Qts is less than 0.8-1.0 (Fs/Qts=50), then it is recommended to choose ZY.
If Qts is less than 0.6 (Fs/Qts=85), then it is recommended to choose FI.
And if Fs/Qts=105, then you need to build a power supply.
My speaker turned out to be somewhere in the middle between ZY and FI.
Initially, I aimed to build a ZY - after all, I don’t need to bother with tunnel calculations, and it’s easier to make a box. The computer program gave me something like 200 liters of recommended box volume for my speaker. It is unrealistic to cram such a box into a passenger car, so a long struggle began for the volume and the best shape of the frequency response. In the end, I settled on a bass-reflex subwoofer with a volume of 63 liters and FI 25Hz. It probably took me about a week to make this decision; I kept sitting and looking at the frequency response curves, with different body volumes and FI parameters. Calculations were carried out in two programs: JBL SpeakerShop and WinISD beta.
The box was planned to be 65 liters, in the hope that the head itself and the internal fasteners would take up some space, and I would have some left over
just 63 liters. Carefully and not once, I ran to the car with a tape measure and sketched out a drawing of the box.

Step Four - Manufacturing.
As I already said, the FI frequency was planned to be 25Hz. External beauty for a subwoofer enclosure is not of the first importance, but... But.
When I looked at the plastic pipes I bought at the market, I saw that there was a way to cut them so that a small flange would remain, which would solve the decorative problems of FI.

Yes, plastic water pipes are ideal for bass reflex design. They are not expensive, there is a large selection of diameters, and they are easy to process. The only thing is that closer to the edge, the pipe has a slight expansion, which introduces an error into the bass reflex calculations. Everyone is used to the fact that the FI is usually round or rectangular, but on the Internet I came across calculations of very tricky sections and shapes. To be honest, I was too lazy to calculate with mathematical accuracy, I estimated it by eye. If I want to leave the selected frequency, then if the FI diameter increases, it will need to be lengthened. My FI was originally supposed to be 198mm. I decided to make 215mm. Even if I was wrong, I still think I didn’t go beyond 23-24Hz, which is quite acceptable for me. I decided to make the box from 16mm thick chipboard, solely for economic reasons - I had several sheets of this material, so there was no need to buy anything. Yes, yes, I’m ready for criticism, the wall is somewhat thin for such a volume. Solely, in order not to steal the internal volume, I decided to fasten the walls of the box using aluminum corners and M5 bolts. Again, the corners were found in the garage in the required quantity, which was also nice. I cut all the walls in one go; after all, a jigsaw is a thing. Only using it in an apartment turned out to be not entirely environmentally friendly. He also cut out a round hole for the speaker, a square hole for the contact pad and two holes for the bass reflexes. The second stage was to drill all the necessary holes for the speaker and wall ties. The latter were also drilled out for countersunk bolts. Then, four M6 nuts were attached to the inside of the front wall using the same epoxy to secure the speaker:

I planned to install the bass reflex pipes, as well as the contact pad, at the very end, already on top of the carpet, so that it would be beautiful and clear. The holes for the FI didn’t turn out very neatly; the pipes barely dangled in them without falling through along with the flanges, so I had to adjust them by filling the void with epoxy and sawdust. I did it this way. I lubricated the pipes with grease (I had it lying around in the trunk), wrapped it in paper, inserted it all into the holes in the front wall and filled the cracks (photo above). After hardening, thanks to the grease, the pipes were easily removed and set aside until the final phase of construction. This is how I got very tight and clear seats for the FI. To finish the topic with these pipes, I’ll say that after cutting them to length, I held the edges of each of the pipes over a gas burner, saving myself from working with a file. The result was very neat semicircular ends, and so what if they won’t be visible anyway. Thus, by this moment, the walls and corners with all the holes and a bunch of bolts and nuts lay in front of me. All that remained was to assemble all this so that the design would miraculously transform from a set of 2d planes into 3d and somehow become similar to the final product.
The box was assembled by laying a thin layer of “liquid nails” under the corners and at the junctions of the walls.

When the box was assembled and the bolts were tightened and pulled in a circle, the internal seams and walls were glued with fiberglass soaked in epoxy glue. Of course, this had to be done in stages - filling one wall with epoxy and laying fiberglass, we had to wait until everything hardened, and then move on to the next plane. However, this benefited the calluses - 80 hand-tightened bolts showed themselves very quickly. All the cracks on the outside were also filled with a mixture of epoxy and sawdust.
By the way, the epoxy just finished me off; it had the unpleasant property of running out very quickly. Perhaps only “liquid nails” were not inferior to it. Again about epoxy, I can give advice. Typically, epoxy does not come with a container in which it needs to be diluted. An excellent solution was found naturally. To mix the epoxy glue, where I added sawdust left over from cutting chipboard sheets, empty beer cans with the top cut off helped me a lot. I don’t know how it happened, but I didn’t experience a shortage of empty beer cans during the construction of the building. The next step was to putty the outside of the body. This was necessary because the bolt holes
almost everywhere they were drilled too heartily. After which I painted the structure with the paint that came to hand, thereby ticking the
column "moisture protection".

Why and for whom am I writing all this? Mainly for those who are just planning to build their first sub. Or for those who consider it too uninteresting or difficult, with the goal, of course, to convince them otherwise. So, when you go to buy carpet, stock up on paper and pencil. I probably presented an amusing spectacle to the store clerks. Well, imagine a man with bulging eyes standing in front of a roll of carpet. Doesn't move. Five, ten, fifteen minutes. But in fact, at this time my brain is feverishly
modeled the material pattern for the subwoofer housing. As a result, the carpet was purchased. Gray, matching the sewer pipes turned into bass reflexes. There I also purchased a material, the name of which I now can’t remember, something like batting, only three times thicker, let’s call it megavatin. I planned to use it to line the inside of the box. They say this virtually increases the volume of the box, I don’t know how it happens, it’s nothing other than witchcraft. Carpet was glued with “liquid nails”. Squeezed it out
surface and spread it over the entire surface with a rubber spatula.

Then he applied the carpet and rolled an empty glass beer bottle on top. You can probably do it full, I'll try it next time. I inserted the bass reflexes into the holes and poured epoxy through the hole intended for the speaker from the inside, laying small pieces of fiberglass here and there. Using the carpet gluing technology, I glued mega-wadding inside. The result was a very cozy, insulated birdhouse with a volume of 64 liters.

All that remained was to install the contact pad and screw on the speaker, which was urgently done. All!

The last line on the expense sheet - the decorative grille for the speaker - was left blank. Due to its high cost, I decided to purchase it after a test drive. It took me about a month and a half to do everything. And what you wanted, you had to go to work. I carry the baby into the car. 20-25 kilograms. Anticipating the unknown, I twist the wires...

Step Five - Result.
The very first listening in the courtyard of a residential building aroused the expectation that objects from residents who did not visit sites dedicated to car audio would fly at me from the windows. The result exceeded all my expectations. What else can I say, it’s just a bomb! Describing sound is a useless task, at least I am not able to scrape together enough writing qualities in myself to do this. All I can say is, I'm pleased.
Today the circuit is as follows: Head JVC-807, rear: JBL-936, sub: 75-GDN, amplifier LADA 100.4, two channels for the rear, two in bridge mode for the sub. There are plans to get rid of the rear, although I really like the JBL-936, and install component acoustics on the front. To the podiums, which, of course, I will do myself, and I will tell you...

P.S. - Expenses.
75-GDN 700rub
Fiberglass fabric 2m - 100r
Plastic pipe 2 pcs - 60r
Bolts and nuts 80pcs - 92r
Contact area - 65r
Cooking spatula (for applying epoxy) - 8r
ES 100g 40r x 3 = 120r
ES 150g 49r x 2 = 98r
ES 280g 72r x 3 = 216 rub.
"Liquid nails" - 52r x 2 = 104r
"Liquid nails" - 70 rubles x 4 = 280rub
Carpet - 220r
Textile - 180rub
Set of wires - 400rub
Amplifier terminals - 46r
Aluminum corners, chipboard and paint were already in stock
Total 2689 rub.
I admit that I might have forgotten to record a couple of trips to the stall for the next epoxy and “liquid nails”, and even if I include the beer I drank during construction, I still don’t think I went beyond 3 thousand rubles. But the main thing, of course, is not the money. The main thing is the pleasure you get from the fact that you made something with your own hands, and it also sounds great!
I lied about beer...

P.P.S - Useful links.
Subwoofer on 75GDN-1-4
Bandpass on 75GDN
JBL SpeakerShop
WinISD beta, subwoofer calculation program
Marchand Function Generator, an audio signal generator for a sound card
, also a sound signal generator, I wrote for this purpose myself. See section of the site Programs
75 GDN - club, I recommend visiting if you are planning to build a sub for 75 GDN

P.P.P.S.. Many people ask about FI sizes.
Here they are: length 215mm, diameter 55

How I made a shelf for acoustics -

It may seem strange, but the speaker is mainly characterized by three parameters proposed by Till and Small:

Fs- this is the resonance frequency of the speaker without any acoustic design. This is how it is measured - the speaker is suspended in the air at the greatest possible distance from surrounding objects, so that now its resonance will depend only on its own characteristics - the mass of the moving system and the stiffness of the suspension.
Qts- the ratio of the transfer function of the speaker at frequency Fs to the transfer function at frequencies where the amplitude-frequency response (AFC) of the speaker is horizontal, i.e. at frequencies above Fs. In other words, Qts characterizes the efficiency of the speaker at the resonant frequency.
Vas- the volume of air that has flexibility (the inverse value of elasticity) is the same as the movable speaker system.

When a speaker is placed in a closed box (CC), the flexibility of the air inside the box adds to the flexibility of the speaker's moving system and its resonant frequency changes. There is the following pattern: when a speaker is placed in a box of volume Vas, its resonant frequency Fs and quality factor Qts increase by 1.4 times. Measuring these parameters at first glance at the design is quite a pain, but having done this once, all doubts disappear - everything turns out to be quite simple .

First you need to prepare:

ZY I immediately dismissed it - why do I need a column with low efficiency? The speaker is not the most powerful anyway - nominal 50W, maximum 75W. In addition, considerable pressure is created inside the column, which requires special tightness. And the Fs/Qts setting is not suitable for my speaker. True, the ZY is also the smallest box of all - which can sometimes turn out to be important.
FI I didn’t want to do this for three reasons: 1) I need to make a super even hole and then cover it with something, otherwise my child will immediately condemn the speaker 2) I need to install a filter to cut off frequencies above 200Hz, otherwise the 75GDN then sounds not cool at all. 3) a normal low-end roll-off was obtained with a wild box size of 120-150 liters (another closet in the apartment), my wife would have kicked me out along with this subwoofer :) BUT! Fs/Qts=74, i.e. the speaker is most suitable for FI, and gives great lows, only it is SIZE 8(. Here we must take into account that almost the same speakers are suitable for single bandpasses as for bass reflexes.
Bandpass I liked it and liked it the most. There is no need to make a filter - the housing itself filters. The speaker is hidden inside so you can't puncture it. And calculations in the programs showed the best results with the appropriate size...

Calculation and design of the box.

Calculations showed that the Bandpass had relatively good dimensions and a good drop-off at the bottom, however, the drop-off still depended heavily on the volume and we had to make a compromise by slightly reducing the box to 65 liters. I carried out the calculations in three programs at once in order to check the accuracy of what I had built. The results were almost identical. I used WinISD 0.44, WinISD Pro Aplha and JBL SpeakerShop or BassBox (called find 10 differences). I liked the first program the most, the second was terribly buggy (that’s why it’s Alpha), but in some respects useful, the third simply confirmed my calculations (it has a very inconvenient interface - it’s bad to change parameters on the fly, selecting values for the size of cameras and bass reflexes, and after each boot you need to switch to the metric system). So what happened - look at the graphs (you can take the project files later).

It was produced in 1987, the suspension on it had already been replaced and was quite high quality. Let's take a look at the old AC-90 speaker housing.

Old speaker and my box

According to my calculations, 60 liters of pure volume and a tiny port gave a tuning of around 20 Hz, in my opinion, this is a gross mistake by the designers; even lovers of the worst blacks do not tune their boxes that low. There is absolutely no benefit from such a low setting, although if you look at the documentation of the 75GDN you can see Fs = 25 Hz (±5 Hz), but this is doubtful, since Fs is too low for a light paper 10″ suspension.

In general, what is there to talk about, we measure the real parameters of the speaker: Fs = 36.7 Hz Qts = 0.42, Vas = 44 liters. We think Bassbox 6 Pro has the optimal design. The subwoofer is planned to be installed in a VAZ 2112, we add the file of the transfer function of this car, we get the following graph:

Bass reflex 45 liters 38Hz. White graph of the frequency response of the subwoofer in a free field, red - taking into account the PF of the VAZ 2112.

The frequency response can be slightly equalized by lowering the port setting, but this is far from a pump and 38Hz will be enough. Such a box should play from 33 Hz to 57 Hz (at a level of -3 dB), frequencies above 60 Hz, the interior no longer amplifies and even dampens, so the midbass is simply obliged to take over from the subwoofer at 60 Hz. In the same Bassbox program, we set the power parameter to 100 W and look at the graph of the dependence of the linear displacement of the cone on the frequency, below 33 Hz the stroke begins to exceed x-max, accordingly the subsonic must be tuned exactly to this frequency, the 2nd order will be enough. The speaker’s fairly high sensitivity (more than 90 dB) allows you to save on the amplifier; 2*80W is enough if you connect it with a bridge to an 8-ohm head, and it will be easier for the amplifier to operate.

Assembly

There is nothing special here, the material is Soviet chipboard.

The speaker itself, box and pipe

By the way, its density is better than many modern materials, approximately like MDF. We saw it, assemble it using glue and screws, the port is a regular 110th pipe.

The speaker is quite old, I assembled the box with nostalgia))

Holes cut with a jigsaw

After assembly, I covered it with black leatherette and placed a layer of 5 mm foam rubber underneath.

The wire was brought out through a sealed hole, no unnecessary connections:

Photo to tease audiophiles. Regular PVA 2x2.5 mm²

The grille and port were painted with automotive paint and varnished

First, we stitch the sides using a sewing machine, then manually tighten them with threads.

View from the other side

All in black style

Unfortunately, I haven’t installed it in the car yet (in search of a simple two-channel), but home listening from the TDA was surprising, it sounds excellent, I would never trade it for stupid “pumps”, which are a dime a dozen right now. All the best and have a great weekend.

A subwoofer is a low-frequency loudspeaker that reproduces the lowest bass frequencies that are not included in the frequency range of full-range speaker systems working in conjunction with it.

The reason for writing this article was the rather small number of detailed articles on the Internet on ready-made structures. So I decided to make my contribution. For some reason, most of the articles concerned car subs - it’s simply amazing how many people we have who are ready to give up half the trunk in order to drive with a roar and jump rhythmically. In addition, noise at speed will still not give pleasure from the sound. My opinion is that you just need to buy good acoustics for your car, that’s usually enough. But at home you can listen to music in a quiet environment and watch a movie with surround sound. The article was intended as a brief summary of the theory with its reflection in practice, so I tried to summarize the theory from other articles - not to invent it myself.

I had a pair of S-30 speakers connected to the computer through a homemade computer amplifier, so I used them for the fronts. There was nothing for the rear, so I bought 15AC-315 (small plastic speakers, but they sound very poor in the mid and high range) and the Vega-120 amplifier for a nominal fee. I decided to connect the central channel to the TV, since it can be separately adjusted from the remote control, and the sound there is sufficient to reproduce voices. And if you have a 29" TV, you can generally do without a central speaker, because there, as a rule, the sound is very decent. Then, of course, I began testing all the channels in films, since I only had a stereo amplifier. And then it turned out that that in other channels there is a lot of sound that is not in the front speakers :), and in the sub there are sounds that are not in any of the channels. Having read on the Internet about the recommendations of the best sub-cavods, as well as recommendations from friends who own a house. cinemas, I was obviously fired up.

And then a friend advised me to put this together myself, slipped me a couple of articles on 75GDN and recommended that I go to the market and look at the speaker. After much torment and thought, I bought a 75GDN-1-4 on the market for $23, as the cheapest and best described on the Internet, although I looked at different ones, including JBL from $80 to $250 :). I must say, the first thing I did before purchasing was to study the prices. The most affordable price for purchase turned out to be Sven-Audio, but it’s still a bit expensive and the sound quality is questionable. So, the simplest sub would cost $180, and a normal sub would cost about $300-350. All expenses for the production of the subwoofer were recorded, so I have calculated the cost of the design :), but first things first...

Theory and measurements of speaker parameters

To calculate the speaker, you need to know at least the basic parameters of the speakers. These parameters are in abundance on the Internet, but the trouble is that all as one, give implausible parameters for 75GDN-1-4 (I don’t know for others). On one of the sites I found a scanned passport for this speaker, and I trust these parameters more. There are probably reasons for this, one of them is that such heads were produced by a whole bunch of factories and for quite a long time by Soviet industry, so the parameters may have changed over time. But the fact is that they have changed, as it turned out, 2 times! And with the parameters from the reference books, when calculated in the program, the volume of the box turned out to be 5 liters, which alerted me. By the way, in almost all the articles I read there were recommendations to measure YOUR speaker parameters. The decision to take the parameters of my speaker was made after a week of fiddling with programs for calculating the subwoofer (despite the terrible laziness of doing this), in which it turned out that these parameters greatly affect the size of the box and the frequency response, respectively. In order to understand what parameters are needed and what they are used with, as well as how to get them without having any sensible measuring instruments at hand, this chapter will be discussed.

It may seem strange, but the speaker is mainly characterized by three parameters proposed by Till and Small:

Qts- the ratio of the transfer function of the speaker at frequency Fs to the transfer function at frequencies where the amplitude-frequency response (AFC) of the speaker is horizontal, i.e. at frequencies above Fs. In other words, Qts characterizes the efficiency of the speaker at the resonant frequency.

Vas– the volume of air that has flexibility (the inverse value of elasticity) is the same as the movable speaker system. When a speaker is placed in a closed box (CC), the flexibility of the air inside the box adds to the flexibility of the speaker's moving system and its resonant frequency changes. There is the following pattern: when a speaker is placed in a box of volume Vas, its resonant frequency Fs and quality factor Qts increase by 1.4 times.

Measuring these parameters at the first glance at the design is quite a hassle, but having done it once, all doubts disappear - everything turns out to be quite simple. First you need to prepare:

Download signal generator program for sound card
- you can also merge - the oscilloscope to the sound card. By connecting the output to the input, you can see what the generator is doing
- download my
- find a 1 kOhm resistor
- take a stereo, at least a power amplifier, because you need to amplify first the signal itself, and then the measured signal
- It is advisable to take a digital voltmeter, so as not to recalculate or change the ranges. I took digital and pointer, and compared the results to check.

Take a pen and paper

We launch the generator program, set the volume on the computer to the middle (otherwise the sine wave will be cut off), and the rest will be corrected with an amplifier.

We connect a voltmeter to points A and C (i.e. to the output of the amplifier), and set the voltage to 10-20 V at a frequency of 500-1000 Hz, adjusting the volume on the amplifier.

We connect the voltmeter to points B and C (i.e. to the speaker).

We set the generator to ~5Hz and REMOVE the SPEAKER away from all objects and walls (you can hang it up if possible). Practice has shown that, far from objects and on the floor, a speaker lying on the floor still gives different readings, but they are insignificant, but will affect the purity of the experiment.

When changing the frequency of the generator, we look at the voltmeter readings - we are interested in the maximum and minimum voltage. Approximately near the resonant frequency, the voltage increases sharply, and then drops sharply. At maximum voltage, we look at the frequency - this is Fs.

Thus we already have Fs. By changing the frequency upward relative to Fs, we find frequencies at which the voltmeter readings are constant and significantly less than Us (with a further increase in frequency, the voltage will begin to increase again). Let's write down this value, Um. It is better to repeat the procedure with a decrease in frequency.
From these records we get something like this graph:

Where, Fs is the resonant frequency, and Us is the corresponding voltage. Um is the minimum voltage, U12 is calculated in Excel after entering the data into the cells. Again we turn the frequency again and look for when the voltmeter readings coincide with the value of U12, remember the frequency. There should be two such values, as can be seen from the graph. These will be F1 and F2. We enter them into Excel. That's it - look at the value of Qts. I made two calculation methods to check the correctness of what I calculated :) in the end, the readings converged, and the small difference is explained by the calculation error. You don’t have to count Vas, but take it from the reference data, it is similar to the truth, and this parameter does not greatly affect the calculation of the box. In addition, to calculate it, you will have to build a plywood box that is sufficiently rigid and airtight, and then repeat the measurements. If you still want to calculate Vas, read the source text on how to do it. I set myself a goal to create something like a summary of the materials that I read, in order to discard the unnecessary, and for an in-depth study of the issue, I provide all the links at the end of the material. So we removed the speaker parameters. For 75GDN-1-4 this is what happened:

The most important parameter is Qts. How do you feel about the difference? TWICE! I must say that I downloaded about a dozen speaker reference books, other parameters in them fluctuated, but this one was surprisingly stable. So trust people after this...

There are only three most common types of acoustic design:

Drawer type

Closed Box
(ZY)
Closed

PhaseInverter
(FI)
Vented

Bandpass (BP)
4th and 6th orders
Bandpass

Selection criterion

Qts< 0.8-1.0 , оптимально 0,7
Fs/Qts=50

Qts<0.6, оптимум - 0,39
Fs/Qts=85

Fs/Qts=105

Distinctive characteristics

This is the easiest type of speaker system to manufacture.

Despite the simplicity of the design, it has many advantages, but the C.P.P. is the smallest compared to any other type of acoustic design - as a consequence, the need for considerable power and possible failure of the speaker (from excessive efforts:)

To calculate the characteristics, there is only one parameter - the volume of the box.

In its operating range, the bass reflex creates completely greenhouse conditions for the speaker, and exactly at the tuning frequency the oscillation amplitude is minimal, and most of the sound is emitted by the tunnel. The permissible input power is maximum here, and the distortion introduced by the speaker is, on the contrary, minimal.

The bass reflex is much more capricious in the selection of parameters and settings, since three parameters are subject to selection for a specific speaker: box volume, cross-section and tunnel length.

Efficiency champion.

By selecting the appropriate volumes and frequency tuning of the front chamber, it is possible to build a subwoofer with a wide bandwidth but limited output, that is, the bell will be low and wide, or one with a narrow bandwidth and very high efficiency. in this strip.

The bandpass is a capricious thing to calculate and the most labor-intensive to manufacture. But the speaker is buried inside - there is less risk of damaging the speaker and there is practically no need for a bandpass filter (although in practice it turned out that it is still desirable)

FI I didn’t want to do this for three reasons: 1) I need to make a super even hole and then cover it with something, otherwise my child will immediately condemn the speaker 2) I need to install a filter to cut off frequencies above 200Hz, otherwise the 75GDN then sounds not cool at all. 3) a normal low-end roll-off was obtained with a wild box size of 120-150 liters (another closet in the apartment), my wife would have kicked me out along with this subwoofer :) BUT! Fs/Qts=74, i.e. the speaker is most suitable for FI, and gives great lows, only it is SIZE 8(. Here we must take into account that almost the same speakers are suitable for single bandpasses as for bass reflexes.
Bandpass I liked it and liked it the most. There is no need to make a filter - the housing itself filters. The speaker is hidden inside so you can't puncture it. And calculations in the programs showed the best results with the appropriate size...

Calculation and design of the box

Calculations showed that the Bandpass had relatively good dimensions and a good drop-off at the bottom, however, the drop-off still depended heavily on the volume and we had to make a compromise by slightly reducing the box to 65 liters. I carried out the calculations in three programs at once in order to check the accuracy of what I had built. The results were almost identical. I used , WinISD Pro Aplha and or (find 10 differences called). I liked the first program the most, the second was terribly buggy (that’s why it’s Alpha), but in some respects useful, the third simply confirmed my calculations (it has a very inconvenient interface - it’s bad to change parameters on the fly, selecting values for the size of cameras and bass reflexes, and after each boot you need to switch to the metric system). So what happened - look at the graphs (you can take the project files later):

This can be seen in the comparison between BP and FI. With equal box volumes, the FI is significantly inferior to the Bandpass. And for large ones it’s the other way around. So if you want to have a new closet at home, then you should take a closer look at FI. About the hump in the middle, see below. And almost the same in JBL SS:

Here the volume of the FI is slightly smaller, but it is still 2 times larger than the given BP. For PSU, the volume is also critical; you can make it larger, thereby reducing the dip in the middle and improving the roll-off at the bottom.

The calculation comes down to selecting the displacement of the cameras and the frequencies to which they are tuned, and checking the frequency response. I think the other three graphs will not bother you :).

The hole in the middle could not be removed at all - such is the speaker :). I didn’t want to increase the volume, I had to fit into the room, by the way, the box is quite big anyway. But I think such a failure can be neglected - after all, a 3dB dip is very small (the graph is simply stretched in height), and if you take into account the uneven frequency response of the speaker itself at 10dB, then you can simply forget about it. In addition, this is still an idealized frequency response, in life everything is much more complicated and confusing :) The dip can be made smaller by narrowing the frequency, but I wanted to extend the frequency response to 200Hz, which was not entirely possible, but it certainly got to 150Hz :). I’ll note right away that the middle is still audible in the subwoofer through a thick hole, so an active filter won’t hurt, which I then did in the amplifier.

The calculation of bass reflexes comes down to setting the internal diameter of the pipe in meters and checking the “Vent mach” value to green, when it turns red it’s bad – there’s too much air flow, i.e. it’s no longer a subwoofer, but a musical instrument of the “trumpet” type. Here you need to look at the length of the bass reflex so that it fits into the box, and preferably up to the middle. For a long time I could not fit the size, because by increasing the diameter, for normal air flow, the length of the phase immediately became immeasurable. The length depends on the diameter and frequency to which the phase is tuned - so you can also play with the frequency. In this case, the frequency response will change, keep this in mind.

At first, a sub was made with identical pipes, fortunately in the upper one, the program shows less air resistance, but after assembling the 100W amplifier and listening, it turned out that after ~50W the effect of air exhaust (popping) began just in the upper compartment (the smallest one, however) ). I had to disassemble everything and cut a large hole for a thick 105mm pipe inside, so I almost fit right into the height of the camera - 2cm remained. Considering that you also need to put a sound absorber there, this is a very small margin. I used plastic sewer pipes. Here I note that there are 70mm pipes, but they are not as big as 50mm and 105mm. As the upper phase increased, the lower phase immediately began to work better. I really don’t recommend using 50mm - for such a fool it’s very small. The result is this: for a sub, the dimensions of the elements are the most important.

Below in the windows are all the parameters of the speaker, but you can take all this in project files for both programs: for WinISD and for JBL SpeakerShop.

Well, the volumes have been determined, it’s time to calculate the box and structure. Since I do 3D modeling as part of my job, I did just that - I took the SolidWorks program and created a 3D model there. If you noticed, 3D graphics are also involved in the site design :) The program itself calculated the volume for me. It’s difficult to do this yourself accurately, since all the connecting bars in the structure eat up a decent amount of space, and the design had to be invented and changed on the fly. Another problem was the material and its thickness, and the program immediately made it possible to see the dimensions of each element, taking into account, precisely, the thickness of the sheet and the connection with each other, i.e. I automatically got the dimensions of each element.

Separately about the material. Well, I didn’t even think about MDF, although of course the best option. The task was to find 20-22 mm chipboard, but it turned out to be practically impossible. The most common chipboard is 16mm or 18mm laminated Polish. 16mm is not enough, and laminated on both sides is probably bad for sound, and it also turns out to be expensive. Then I realized how right I was by not buying laminated chipboard. And it took about 1.5 weeks to find a regular 18mm one. Those that were in stores are impossible to take home, because the sheet is very large. I have already called all the offices and visited all the markets. My nerves were starting to give in - I almost regretted that I contacted the sub, because everything except the box material had already been purchased. I started thinking about how to glue 16mm chipboard and 4mm plywood, but I stubbornly didn’t want to glue it - this requires special glue and a strong press. And then a friend called and asked for help to bring cement to his home. So, while we were running around looking for cement and its owners throughout the industrial zone, we accidentally came across an office selling and sawing that very notorious laminated Polish chipboard. They used 22mm chipboard sheets as spacers in the racks. But the owner was not there and we had to wait... After looking again for the owners of the cement and not finding them :) We returned to the chipboard again. The owner did not give me the 22mm, citing the fact that it was difficult to pull out, and they seemed to have already sagged (in short, I was too lazy to pull out) and offered me 28mm. This is cool, I thought, and refused, looking at these thick sheets. And then he offered me 18mm, simple sanded chipboard - it turns out that it is used as transport sheets for laminated (top and bottom). So, all the good stuff, including sawing on an imported machine, cost me $5 (cutting $0.3/m). You can’t cut it like this in life - evenly, exactly to size. Draw your own conclusions...

Yes, what am I talking about? Oh yes - about the box. Let's see what happened:

As you guessed, this is a 3D model. The length of the bass reflexes is visible in the screenshots of the program above and is 19 cm at the top and 25 cm at the bottom, the internal diameters are 105 mm and 70 mm, respectively. There is a hole in the back wall for the connector socket. The legs were made of steel spikes at the factory and hardened. I chose the size according to my taste. In the articles I found, people made the spikes 2 times larger, but I didn’t want to put the sub too high so that the spikes wouldn’t be visible, because the design is not low anyway. You need 4 spikes, I tried it on 3 stands - a terribly unstable structure. In the columns you can get by with three, since they are not deep and their center of gravity is in the front. I leveled the height of the spikes with a washer; only one was needed, then I checked it on a obviously flat surface (the back wall).

Assembling the box

Stage 1

Assembling the middle and lower shelves. First, I rounded the upper inner edges of the beam. The 20x30mm beam is fastened with 45mm long screws. I pre-drilled holes for the screws everywhere. The entire assembly of wooden units was placed on EXTRA PVA - this is a thicker PVA for stronger gluing. All corners were coated with silicone sealant for wood - it is harder than usual for glass, etc. But I think anyone can do it.

Stage 2

We screw the side walls to the bars from the inside with 35mm screws. The middle shelf must be measured fairly evenly. Don't forget the glue. We attach timber to the side walls at the front. It turns out to be a beam, in front, around the entire perimeter of the large chamber.

Stage 3

We glue the bass reflexes to the front wall. I glued them with liquid metal - such a solid multi-component garbage, it smells like epoxy. She kneads it with her hands, it turns out like plasticine - soft and terribly sticky. It hardens almost instantly as soon as it cools down, and if it doesn’t cool down, it still hardens after 5-10 minutes, so I then removed it from my hands with pumice stone along with the skin :) I made a thickening along the entire radius at the base, for reliable fastening - they still stick out far. In addition, I personally was not able to perfectly accurately cut such a hole with a jigsaw, and these unevenness were miraculously sealed with liquid metal.

I wrapped the bass reflexes with window paper tape in 5-7 layers, smoothing everything very tightly. Then I covered them with linoleum with insulation. The thickness of the phases turned out to be about 7 mm.

Stage 4

We screw the front wall to the bars from the inside with 35mm screws.

In the upper part we screw it onto the corners for greater rigidity of the structure. I used plastic furniture ones with a lid.

Stage 5

We install the speaker on the sealant, fasten it with bolts (I don’t remember the size). We coat it well in a circle - you can see the white sealant in the photo. Solder the wires. I attached a block with connectors to the back wall.

Stage 6

Now all this needs to be covered with a sound absorber. There are many material options. I used foamed linoleum with insulation, and then, in the end, thick padding polyester ~2.5-4 cm, which, in addition to pasting, was also fluffed in a large chamber. The linoleum also needs to be glued to the top and back walls. Can be pasted over in several layers. Thick batting is a good thing, but I couldn’t get it.

Stage 7

We put the top cover and the back cover on the sealant, without glue - then you may have to remove them. I fastened them with furniture ties under an internal hexagon - an excellent thing, it fastens very firmly, the main thing is not to tear it off if you tighten it with a drill, like I do. I attached it at the end, here you have to be careful with the chipboard - it can delaminate. For ties, first drill one long hole, and then enlarge the screwed hole with a larger drill. Two ties are at the top in the front wall, the rest are on top of the lid, and the back one is just around the entire perimeter and in the center. You can see how it is attracted by the protruding sealant, but it is attracted well.

In the photographs there is still the first option, with a small upper bass reflex and padding polyester not yet pasted over. Later I covered the bars with linoleum too. This was, so to speak, the first test assembly, in order to listen to whether it would sound or not. Sounded. But so far I only had a 25W amplifier, it was fine for now.... Then I gave up the whole thing and started assembling a 100W amplifier, I had to test the design to the maximum possible. And it was not for nothing that I was afraid of a more powerful amplifier - then it turned out that at maximum the upper phase squelched, but the case was sealed the first time - it was not for nothing that I thoroughly sealed all the seams and corners with sealant. I had to disassemble the entire structure, cut out a larger hole and install a thicker pipe. Here I have already completely “insulated” everything with padding polyester. Inside it became soft, white and warm like a snowdrift. What can I say, it sounded quite good - smooth, soft bass. At maximum volume of the amplifier, the speaker begins to jam, but we must take into account that its nominal value is 50W. It's enough for an apartment higher than the roof, especially since it shouldn't be heard in music, just when you cut it - as volume is added to the music and such a solid low sound. In short - good work...

I plan to cover the entire structure with wood-look film. But first, you need to putty and sand the edges - no matter how accurately I didn’t cut it on a machine, no matter how I tried to carefully assemble everything - there is still a total error of 0.5-1mm. That's why laminated chipboard is still bad, because you have to immediately glue the tape to the end - in short, it will be crooked.

And you can argue about the feasibility of building it yourself, looking at the costs...

Construction cost

speaker 75GDN-1-4	$23,32
Chipboard	$5,18
mounting socket (connectors)	$2,33
fasteners (screws)	$3,11
linoleum	$5,39
padding polyester	$3,11
timber	$1,04
the wire	$3,89
pipe 70mm	$2,59
pipe 105mm	$2,59
dye	$5,13
film	$7,77
putty for wood	$1,40
glue Moment	$1,24
PVA glue	$1,30
skin	$3,89
thorns	$2,59
liquid metal	$3,11
platbands	I'm still dreaming about them
Total	$78,96

It's certainly not $300. But I note - this is an inactive subwoofer. To compare with a purchased one, you need to add here a preamplifier with a tunable filter, a phase shifter and a power amplifier of about 100 W, and you shouldn’t forget about the power supply (possibly the most expensive one in it). Since I decided to build an amplifier for 6 channels at once, I did not bother with building the amplifier into the sub. Still, there is a lot of vibration inside the case, and it’s difficult to combine the tightness of the box with the ventilation of the amplifier. In general, an amplifier is a thankless job - an awful lot of hassle. I probably won’t show such heroism again. A separate article will be written about the results of designing an amplifier based on the TDA7294...

Subwoofer on 75GDN-1-4

Article taken from the site pavel.artmech.com Author of the article: Novik P.E.

Introduction.

A subwoofer is a low-frequency loudspeaker that reproduces the lowest bass frequencies that are not included in the frequency range of full-range speaker systems working in conjunction with it.

The reason for writing this article was the rather small number of detailed articles on the Internet on ready-made structures. So I decided to make my contribution. For some reason, most of the articles concerned car subs - it’s simply amazing how many people we have who are ready to give up half the trunk in order to drive with a roar and jump rhythmically :). In addition, noise at speed will still not give pleasure from the sound. My opinion is that you just need to buy good acoustics for your car, that’s usually enough. But at home you can listen to music in a quiet environment and watch a movie with surround sound. The article was intended as a brief summary of the theory with its reflection in practice, so I tried to summarize the theory from other articles - not to invent it myself.

What inspired me to sculpt such a structure? A whole series of events led to this. The DVD generation is coming, everyone around is preoccupied with buying players and discs (fortunately, they now cost $4-6). It was already difficult to stay away, and my long-standing dream of transferring the family archive from a video camera to digital did not give me rest. I sat down on the Internet, studying articles on DVDs, from equipment to creating my own discs, fortunately I already knew how to digitize and process videos. It was decided to wait until the summer, when DVD writers would become a little cheaper, and buy all this newfangled equipment. To begin with, it was decided to assemble everything ourselves in order to get by at the lowest possible cost. The first thing I bought was a DVD player, which was purchased with the intention of testing my DVD film library on a home player. Moreover, I couldn’t wait until the summer :) I chose it based on price and capabilities, and also because I don’t have a receiver and probably won’t have one in the near future, so it had to have a decoder and an output for 6 channels at once. As a result, I bought a BBK-919PS (at that time it still had a Panasonic drive). I had a pair of S-30 speakers connected to the computer through a homemade computer amplifier, so I used them for the fronts. There was nothing for the rear, so I bought 15AC-315 (small plastic speakers, but they sound very poor in the mid and high range) and the Vega-120 amplifier for a nominal fee. I decided to connect the central channel to the TV, since it can be separately adjusted from the remote control, and the sound there is sufficient to reproduce voices. And if you have a 29" TV, you can generally do without a central speaker, because there, as a rule, the sound is very decent. Then, of course, I began testing all the channels in films, since I only had a stereo amplifier. And then it turned out that that in other channels there is a lot of sound that is not in the front speakers :), and in the sub there are sounds that are not in any of the channels. Having read on the Internet about the recommendations of the best sub-cavods, as well as recommendations from friends who own a house. cinemas, I was obviously fired up. And then a friend advised me to put this together myself, slipped me a couple of articles on 75GDN and recommended that I go to the market and look at the speaker. After much torment and reflection, I bought 75GDN-1-4 at the market for $23, as cheap and best described on the Internet, although I looked at different ones, including JBL from $80 to $250 :).I must say, the first thing I did before purchasing was to study the prices. The most affordable price for purchase turned out to be Sven- Audio, but still a bit expensive and the sound quality is questionable. So, the simplest sub would cost $180, and a normal sub would cost about $300-350. All expenses for the production of the subwoofer were recorded, so I have calculated the cost of the design :), but first things first...

Theory and measurements of speaker parameters.

To calculate the speaker, you need to know at least the basic parameters of the speakers. These parameters are in abundance on the Internet, but the trouble is that all as one, give implausible parameters for 75GDN-1-4 (I don’t know for others). On one of the sites I found scanned passport on this speaker, I trust these parameters more. There are probably reasons for this, one of them is that such heads were produced by a whole bunch of factories and for quite a long time by Soviet industry, so the parameters may have changed over time. But the fact is that they have changed, as it turned out, 2 times! And with the parameters from the reference books, when calculated in the program, the volume of the box turned out to be 5 liters, which alerted me. By the way, in almost all the articles I read there were recommendations to measure YOUR speaker parameters. The decision to take the parameters of my speaker was made after a week of fiddling with programs for calculating the subwoofer (despite the terrible laziness of doing this), in which it turned out that these parameters greatly affect the size of the box and the frequency response, respectively. In order to understand what parameters are needed and what they are used with, as well as how to get them without having any sensible measuring instruments at hand, this chapter will be discussed. I will write briefly; anyone who wants can read in more detail in the attached literature from which the material was collected.

It may seem strange, but the speaker is mainly characterized by three parameters proposed by Till and Small:

Measuring these parameters at the first glance at the design is quite a hassle, but having done it once, all doubts disappear - everything turns out to be quite simple.

First you need to prepare:

download signal generator program for sound card download Marchand Function Generator

you can still drain download Oscilloscope 2.51- an oscilloscope for the sound card. By connecting the output to the input, you can see what the generator does :)

download mine calculation file in Excel

find a 1K resistor

take a stereo, at least a power amplifier, because you need to amplify first the signal itself, and then the measured signal

It is advisable to take a digital voltmeter, so as not to recalculate or change the ranges. I took digital and pointer, and compared the results to check.

Take a pen and paper

We launch the generator program, set the volume on the computer to the middle (otherwise the sine wave will be cut off), and the rest will be corrected with an amplifier.

We connect a voltmeter to points A and C (i.e. to the output of the amplifier), and set the voltage to 10-20 V at a frequency of 500-1000 Hz, adjusting the volume on the amplifier.

We connect the voltmeter to points B and C (i.e. to the speaker).

Thus we already have Fs. By changing the frequency upward relative to Fs, we find frequencies at which the voltmeter readings are constant and significantly less than Us (with a further increase in frequency, the voltage will begin to increase again). Let's write down this value, Um. It is better to repeat the procedure with a decrease in frequency. From these records we get something like this graph:

Where, Fs is the resonant frequency, and Us is the corresponding voltage. Um is the minimum voltage, U12 is calculated in Excel after entering the data into the cells. Again we turn the frequency again and look for when the voltmeter readings coincide with the value of U12, remember the frequency. There should be two such values, as can be seen from the graph. These will be F1 and F2. We enter them into Excel. That's it - look at the value of Qts. I made two calculation methods to check the correctness of what I calculated :) in the end, the readings converged, and the small difference is explained by the calculation error. You don’t have to count Vas, but take it from the reference data, it is similar to the truth, and this parameter does not greatly affect the calculation of the box. In addition, to calculate it, you will have to build a plywood box that is sufficiently rigid and airtight, and then repeat the measurements. If you still want to calculate Vas, read source text how to do it. I set myself a goal to create something like a summary of the materials that I read, in order to discard the unnecessary, and for an in-depth study of the issue, I provide all the links at the end of the material. So we removed the speaker parameters. For 75GDN-1-4 this is what happened:

There are only three most common types of acoustic design:

Drawer type

ClosedBox (ZY) Closed

Phase Inverter (FI) Vented

Bandpass (BP) 4th and 6th orders Bandpass

Selection criterion

Qts< 0.8-1.0 , оптимально 0,7 Fs/Qts=50

Qts<0.6, оптимум - 0,39 Fs/Qts=85

Distinctive characteristics

This is the easiest type of speaker system to manufacture.

To calculate the characteristics, there is only one parameter - the volume of the box.

Efficiency champion.

ZY I immediately dismissed it - why do I need a column with low efficiency? The speaker is not the most powerful anyway - nominal 50W, maximum 75W. In addition, considerable pressure is created inside the column, which requires special tightness. And the Fs/Qts setting is not suitable for my speaker. True, the ZY is also the smallest box of all - which can sometimes turn out to be important. FI I didn’t want to do this for three reasons: 1) I need to make a super even hole and then cover it with something, otherwise my child will immediately condemn the speaker 2) I need to install a filter to cut off frequencies above 200Hz, otherwise the 75GDN then sounds not cool at all. 3) a normal low-end roll-off was obtained with a wild box size of 120-150 liters (another closet in the apartment), my wife would have kicked me out along with this subwoofer :) BUT! Fs/Qts=74, i.e. the speaker is most suitable for FI, and gives great lows, only it is SIZE 8(. Here we must take into account that almost the same speakers are suitable for single bandpasses as for bass reflexes. Bandpass I liked it and liked it the most. There is no need to make a filter - the housing itself filters. The speaker is hidden inside so you can't puncture it. And calculations in the programs showed the best results with the appropriate size...

Calculation and design of the box.

Calculations showed that the Bandpass had relatively good dimensions and a good drop-off at the bottom, however, the drop-off still depended heavily on the volume and we had to make a compromise by slightly reducing the box to 65 liters. I carried out the calculations in three programs at once in order to check the accuracy of what I had built. The results were almost identical. I used WinISD 0.44, WinISD Pro Aplha And download JBL SpeakerShop or download BassBox(find 10 differences is called). I liked the first program the most, the second was terribly buggy (that’s why it’s Alpha), but in some respects useful, the third simply confirmed my calculations (it has a very inconvenient interface - it’s bad to change parameters on the fly, selecting values for the size of cameras and bass reflexes, and after each boot you need to switch to the metric system). So what happened - look at the graphs (you can take the project files later):

At first, a sub was made with identical pipes, fortunately in the upper one, the program shows less air resistance, but after assembling the 100W amplifier and listening, it turned out that after ~50W the effect of air exhaust (popping) began just in the upper compartment (the smallest one, however) ). I had to disassemble everything and cut a large hole for a thick 105mm pipe inside, so I almost fit right into the height of the camera - 2cm remained. Considering that you also need to put a sound absorber there, this is a very small margin. I used plastic sewer pipes. Here I note that there are 70mm pipes, but they are not as big as 50mm and 105mm. As the upper phase increased, the lower phase immediately began to work better. I really don’t recommend using 50mm - for such a fool it’s very small. The result is this - for a sub, the sizes of the elements are the most important :).

Below in the windows are all the parameters of the speaker, but you can take all this in project files for both programs: for WinISD and for JBL SpeakerShop.

Yes, what am I talking about? Oh yes - about the box. Let's see what happened: