Mid-bass box & response dip

[yggdrasil]

A part of a new project I'm working on has a mid-bass box, closed, net volume 16.5 litres. Internal dimensions are 44cm x 31cm x 15cm. 5cm insulation on sides and back.

First box is put together. The impedance plot is very close to predicted.

When looking at the first response measurements it seems I have some issues ~700Hz, which matches the length of the box. Measurements are blue - nearfield, and green at 30cm in a hostile environment.

This issue must be addressed before I move.

Anyone with a good advice on this issue?

My first ideas are to:
a) add insulation at these sides
b) twist one (or both) of the side walls a few degrees

This driver will be used from 150-200Hz to 800-1000Hz.

Johnny

[Ian Mackenzie]

It would be useful to see the physical enclosure and the intended driver.

My take is its a diffraction issue caused by the combination of the height and width.

I doubt its not an internal thing. Test this by completely stuffing the box with insulation.

Difraction can be easily modelled.

You could measure the driver on an IEC baffle with the same size bax and it would probably look quite different.

Try adding some make shift wings and see what happens.

In any case I would not stress over it.



Bevelling the edges can make a difference too.

[yggdrasil]

Here's a picture of the box with one of the alternative drivers mounted.

I have both JBL 2202A or TAD TM1201.

The boxes will be used as rear chambers for mid-bass horns, and I want it to be as good as I can get it before introducing new issues caused by the horn.

I have calculated a 130Hz profile, and converted it to bi-radial format.

Johnny

[Ian Mackenzie]

Nice pic.

Like I said as soon as you mount a Driver on a baffle you will see ripples like all fuck has broken loose.

If you google you can find some free ware and model different baffle shapes.

What you can do is move the driver around the baffle and find a location to minimise the effect.

If the width is equal or greater than highest intended crossover point it will help.

Just add the horn loading and see what happens.

Its often not easy to predict effects of rear chambers.

[yggdrasil]

Thanks Ian.

Maybe I'll just try it with the horn. Will start fabrication next week.

Johnny

[4313B]

Do a couple ground plane measurements at the same 30cm distance and post the results. One measurement with the long side of the baffle vertical and one with the long side of the baffle horizontal.

[jerv]

I have both JBL 2202A or TAD TM1201.


Are the measured 600-800 Hz ripples present with both the 2202A and the TM1201?

Espen

[yggdrasil]

Will do the new measurements tomorrow.

So far I have only done measurements with the TAD driver. The JBL's are in daily use :)

[Loren42]

I would suggest flush mounting the driver to improve some of the diffraction.

Diffraction is the likely culprit with the upper end ripple. You can round over the edges of the cabinet to relieve that, but you really need a 2" radius to be effective, which is usually out of the question for most builders.

Box vibration may be another source to look at. Given how low you want to operate adding more damping will improve the situation. Stiffening the cabinet heavily helps, but you still need to absorb and transfer the acoustic energy from the rear of the cone into heat. Any energy that is not transferred to heat either vibrates the cabinet or worse, reflects back and vibrates the cone of the driver, yielding ripple and distortion.

[yggdrasil]

Finally got the time to do the measurements.

Notation: vertical = long side vertical.

First image with no gating. The box is placed on a tile floor angled down towards the microfon.

Second image with 9-10ms time window. Box is horisontal on the same floor, but with a thick wool carpet.

Third image is with 9-10ms time window. Box is horisontal on the tile floor.

Fourth image is with 9-10ms time window. Box is vertical on the tile floor.

The exact same time window is used on all graphs with time window.

It seems I jumped to the wrong conclusion with the nearfield measurement.

And thank you 4313B for guiding me in a meaningful direction for measuring this driver.

[pos]

Very interesting, please tell us more about your project!
How do you like the sound of the 1201 as a direct radiator so far?
Have you considered the 2020H for your horn setup?

[yggdrasil]

I actually looked for a pair of 2020H when these TAD's turned up...

I have sent music through this 1201, but not in a setup where it is possible to say much meaningful about it. Sounds nice is all I can say.

The project is going to replace my old 4-way's. It all started out as a crossover upgrade. But I soon realised that the complete speaker in one box is not a DIY friendly concept when there are so many drivers and horns to try out, so it has sort of evolved into a set of single drivers/horns project.

I will build new boxes for the 2245's. Relatively low and wide ~10 cubic feet net volume. The dimensions are selected from the dimensions of the horn for the 12" mid-bass. I haven't decided on tuning or eq.

I have modelled a 130Hz hyperbolic exponential horn. Dimensions were chosen so that I can make it without any bends. After converting it to bi-radial format, the mouth will be approximately 3 feet wide and 1 foot tall. The horn will be ~2 feet deep plus the rearchamber. This combo will hopefully play usable in the 200-1000Hz range. I have started cutting pieces for it, and I'm a little overwhelmed by the dimensions. Didn't look like that on the CAD program...

For the HF I started out getting a pair of those cheep 2435's on eBay a few years ago, then a pair of TAD 4001's found it's way to the workshop before finally this winter, thanks to 4313B, a pair of 476Be came home to papa :D. It started out as a 4-way, but with the 476Be it looks like I'll be building a 3-way. I dont' have any UHF driver in house that will improve on the 476.

I have also made a hyperbolic exponential horn for the 1.5" drivers. Cutoff 400Hz. The first measurements looks ok, but again new measurements will have to be taken in a friendly environment, not the workshop.

I have an active crossover for this project on the drawing bench. Most of the internals will be using the Nelson Pass B1 as a building block. In addition there will be a balanced to single ended buffer on the input and the option of single ended to balanced buffer on the output. I plan on making the filter modules on small replacable modules. There will also be a slot for an eq module on each band. The output buffer will most likely aslo be on a replacable module.

To power this I have 4 different Nelson Pass designed DIY amps. One relatively big (2x F4 + frontend) 160wpc that will be driving the woofers and 3 smaller one's (Zenlite balanced, F5 and Zen9).

[pos]

Where did you get your TM1201 from?
I have a pair of 2020H sitting on my shelf that I can lend you if you still want to experiment with these

[yggdrasil]

I actually got the 1201's off eBay. New :D.

I might take you up on that offer, after everything is up and running. Main plan is to have it all playing by the end of this year.

[pos]

Hi Johnny

Any update on that project?

[yggdrasil]

Not much progress to report.. Got a little sidetracked - again.

I have done some work on the active filter. Most of the PCB's have been etched, and I have started soldering on the power supply.

I had to build a new Pre-amp in between. Similar modular approach as the active filter, with the same connector for the amplifier modules as the active filter has on the output modules. Actually the amplifier modules used in the pre-amp is borrowed from the active filter project.

[pos]

So you didn't had time to build and measure the midhorn yet?

[yggdrasil]

Nope. Not yet.

[pos]

Hi Johnny
So what's up now? :D

[yggdrasil]

Slowly getting there(too much work and not enough fun this year). Top and bottom halves of first horn are ready. The size and weight are.... lets just say - they will be noticed. Probably january.

I did get some work done on the new active crossover for this project. Actually have been burning it in since april/may.

[pos]

I bet it will be impressive when done, both visually and sonically :D
Did you use the correct mmd for the TM1201 when modeling the horn in hornresp? (it is around 52g, not the 5g quoted in the official T/S)

My system is similar to yours in concept (TL1801+TM1201+2450SL/Be on H9800, active filtering and hypex amp modules), but I use the TM1201 as a direct radiator only (and it is still a work in progress, like yours, sloooowly getting there...). I was thinking about adding a small waveguide around the TM1201 just to smooth the directivity crossover with the H9800, but this is already quite complicated to do (for me at least).

I am very curious to see how your mid horn will turn out. Remember I still have some 2020H you can borrow if you want to try them in your horn :)

[yggdrasil]

Finally got one horn finished. A bit larger than what I visualized from the drawing. They are barely movable for one man, without the drivers.

[yggdrasil]

These are taken straight in front at different distances:

30cm:
57993

60cm:
57994

1m:
57995

Impulse response:
57996

Distortion:
57997

[yggdrasil]

Offset measurements:

30 degrees horisontal offset at 30cm:
57998

~15 degrees vertical offset, at 1m:
57999

[yggdrasil]

Next step is finishing the second horn before final sanding and oiling. Hopefully it will not take as much time as the first one....

[ivica]

Finally got one horn finished. A bit larger than what I visualized from the drawing. They are barely movable for one man, without the drivers.

Hello yggdrasil,

I am amazed when I have realized that You are talking here about the horn applied to the 12-inch driver such as 2202. I am not in the position to judge whether 2202 is appropriated driver for such loading, but anyhow there are lot of questions about sensitivity 'pairing" of the drivers, but I am sure that a skilful person who can make such horn, will appropriately tweak network too.

I have one question, may out of the thread topic, but attached figure in #22 post here, has presented the box which I am dreaming:

-2245
-2202
-2397 horn + say 2441
-2405

do You have any more words about your "old" box, any measurements, any listening 'testing', any ...., any ......, any .....

Regards
Ivica

[yggdrasil]

The old system has 2245's, 2202A, LE85 on small Smith's. The 077's are currently not in use. Rolloff starts at ~14KHz, and becomes significant ~17KHz, which is fine with me.

Last iteration on them was re-voicing the LE85/Smith combo, lowering the lowest frequency output. After that iteration there was no longer any need for the 077. The process ended with charge coupling the network.

Lately I have also moved them away from the back wall and raised them up from the floor, giving a much better bass frequency linearity.

IMO the sound is very good and IME you chicken out before they do. Playing on them is simply fun and addictive.

I have only 2 concerns with them. The first being that all drivers are in one box. The second concern is impulse response - which is not spot on through the frequency spectrum. I.e. the bass comes little later than the rest.

The boxes have a footprint of 55x55cm, are 125cm high and the top is 55x50 cm. The dog box volume is ~30l, which in retrospect is unnecesary large.


The above measurements of the new horn is with TAD TM1201, not the 2202A. Judging by a short session running this driver/horn combo, the TAD driver is ridiculously good.

[pos]

:applaud:
V-E-R-Y impressive build Johnny!
Kudos!

It deserve a measurement outside in an as near as an anechoic situation as possible (but I can understand how difficult it can be to move this beast).
Your 30cm measurement is to close (especially for directivity measurements), and the other ones are blurred by the room.

It would be very interesting to compare the frequency response (and directivity) of the TM1201 with and without the horn!

With your current measurement we can already guess that you gain *at least* 10dB in the midbass!

[yggdrasil]

There are some measurements in post #10 flat on the floor.

They will not be moved outside for measurements. Outdoor climate north of the arctic circle in winter is far too hostile for measurements..... Even summer is too hostile most days.

[ivica]

The old system has 2245's, 2202A, LE85 on small Smith's. The 077's are currently not in use. Rolloff starts at ~14KHz, and becomes significant ~17KHz, which is fine with me.

Last iteration on them was re-voicing the LE85/Smith combo, lowering the lowest frequency output. After that iteration there was no longer any need for the 077. The process ended with charge coupling the network.

Lately I have also moved them away from the back wall and raised them up from the floor, giving a much better bass frequency linearity.

IMO the sound is very good and IME you chicken out before they do. Playing on them is simply fun and addictive.

I have only 2 concerns with them. The first being that all drivers are in one box. The second concern is impulse response - which is not spot on through the frequency spectrum. I.e. the bass comes little later than the rest.

The boxes have a footprint of 55x55cm, are 125cm high and the top is 55x50 cm. The dog box volume is ~30l, which in retrospect is unnecesary large.


The above measurements of the new horn is with TAD TM1201, not the 2202A. Judging by a short session running this driver/horn combo, the TAD driver is ridiculously good.

Hi yggdrasil,

Many thanks for short overview of your "old" system, especially about 'dog box' for 2202 size, for which I would expect over 20Lit would be OK.

Regards
Ivica