Acoustics TQWP – Voight pipe on 4A-28


Created on 03/29/2006 12:31 PM.

Updated 03.22.

Author: Petrukhin Ilya.

Music has always accompanied me. At first it was “Spring-202”, later “Olimp-004”, “Amfiton-002”, “S-90”. Then there was a long pause and “Panasonic RX-DT 30”. And now – a new wave of “longing for nostalgia.”

On industrial affairs he took up automotive acoustics. I had to surf the Internet a lot to fill the information gaps. Everywhere there were links to sites describing, in the rosiest colors, the virtues of Transmission Line loudspeakers. And, without saying a word, the English and the French, and the Danes, and the Australians and the Japanese played the same tune.

As a result, the virus of music lovingly penetrated my soul and firmly grabbed the claws of doubt: – “Why are we worse?” Having searched the Russian-speaking internet and found nothing, I had to test everything myself.

The choice fell on a column of the type “Reverse Horn”, in other words “Trumpet Voyta” or for foreigners “TQWP”. The fact is that I had already listened to the “S-90”, and I did not like this sound at all. The bottoms are very intrusive. I connected this drawback with the large delay of the phase inverter, but now I think differently. So I chose a column without a phasic.

Just as opportune, a friend presented a pair of 4A-28 from long-term storage. Having found a description of improving the characteristics of these speakers, he executed both of them according to Gorynych’s recipe. The result is the following Thiel-Small parameters:

Re 12.48 Ohm
Qms 1.501
Qes 0.802
Qts 0.52
Vas 79 l
Le 0.202 mH
L1 1.057 mH
R1 9.41 ohm
Fs 53.9 Hz
dF 50-17000 Hz
Cms 0.422 mm / H
Mms 14,1 g
Rms 4.65 kg / sec
Bl 10.4 N / A
Rice. 1. Frequency characteristics of loudspeaker 4A-28 in open space.

Having pulled out files with a mathematical model of horn columns from MJ King, and having bought a disk with MatCAD2000, he began a thorny path of trial and error in selecting the parameters of the Voight tube.

After three days (two Saturdays and one Sunday) of experiments using the “scientific poke” method, I picked up something similar to the characteristics of the original source. The resulting column data puzzled me STRONGLY. Nothing like a nifty primary source column! My monster looked depressing!


  • horn height 161cm,
  • cross-sectional area of ​​the closed end of the horn 782 cm2,
  • the cross-sectional area of ​​the open end of the horn is 1364 cm2,
  • the speaker is installed 32 cm from the closed end of the horn (1/5 of the length of the horn),
  • filling the horn (IMPORTANT!) 10g / liter of volume with “Dacron” fiber (this is probably according to our synthetic winterizer).

Nothing like a wedge with a speaker in the middle! However, the frequency response characteristics look very encouraging:

Here and in the figure below, the characteristics of the Voight pipe are drawn in red, and the infinite screen in blue.

For reference, I also give the Z-characteristic of the column:

When I calmed down, I picked up AutoCAD, and using these parameters, I drew a vertical column with a horn turned forward. Please do not judge harshly, do not have a know-how in the design of speakers.

I ordered details from laminated chipboard at the company. Everything had to fit, even the slanting ends.

I came up with the design according to the principle: not a single screw on the outside. Fasteners from the inside through oblique triangular rails 40 * 40mm with screws and Moment-Montage glue. Here is a general view in the process:

The place for the speaker was reinforced with an additional sheet of 10mm plywood:

The curved horn output was made from a sheet of 4mm fiberboard:

I thought to fill the gaps between a thin sheet of fiberboard (marked with red arrows in the figure) and the body with polyurethane foam, but for the rigidity and stability of the column I filled it with concrete. It took one and a half buckets. Here he stepped on the first rake: chipboard swells from moisture (:-(). Won.

A comrade advised to put several struts for rigidity. I put two of 10mm plywood between the front and back walls of the speaker (they are not in the photos yet). After assembling the column, the inner surface was treated with automotive anti-noise mastic. A jar of mastic was used for each column.

Here is the second rake: for each bell you need 1.7 kg of padding polyester.

I thought it was a little. It turned out to be about 20 running meters!

Calculate the price! After some internal struggle, I managed to get by with mineral wool. It took 3 sq.m. on both speakers. This stuff dried up for about a week. I couldn’t stand it any longer: I put the speakers on and turned it on to try.

Look at the happy electronics engineer!

The first impression is of course something! In general, I got a lot of pleasant emotions while still making the speakers, but for the first time to hear what you did with your hands! … I played some dynamics in TQWP. By the way, the first speaker played three days earlier, and after turning on the second speaker, the difference in sound and sound pressure is enormous! Then the truth is, everything returned to normal.

When I brought the speakers home, I measured the frequency response in the near field of the speaker and the output of the horn. Vector has not yet learned how to add to obtain a full frequency response, tk. I will measure the phase response in about a month, after the microphone amplifier is manufactured:

And the measured Z-characteristic showed that there is very little damping material. It is necessary to add mineral wool, or still apply a synthetic winterizer (:-().

I’m not a very good listener, but the sound quality is not yet what I want. The panorama is somewhat blurred.

One friend said about the frequency range – “Good middle. From middle bottom to middle top. “

I disagree with him, there are enough tops. The very bottom is not enough.

So, the next step is to make an audio section for these speakers. I have ORBIT (Russian Kvod). I want to build a 25W / 16 ohm class A amplifier. I came up with the scheme myself. The output current source in the lower leg will be positive feedback, so that the transistors in the X.H. will heat up at 12 watts. The radiator thought it was enough. Let’s listen !?

Read also