I have an other measuring question for you all...
Now, this is not for a mabat design albeit quite close... I wonder how your 520G2 recreates a 1kHz sinc (the half positive part of a one sinus wave) measured at 30 cm 15 deg?
I created this stimuli in REW by Generator, Tones, Toneburst using half a cycle:
Run stimuli repetitive and at the same time run the REV Scope function (use the Trigger to make the wave form stay in the screen between pulses for easy viewing)
This is what I got when DUT was EQed flat 500-20k by the REV EQ function.
The filters used was: LR 8th 300Hz.
I had anticipated a more true replica of the pulse... the LF part in my speakers made more immediate copy with no strange reflections... (in this mesuement bass was off)
I have a little hard time to interpret the results. The first small positive "blipp" looks OK but if thats the stimuli and the rest is resonances then I'm surprised.. but perhaps it is what is happening!?
I hope this is still somewhat on topic on what is discussed here currently and is perhaps an alternative to the FFT analysis...
How yours is doing would be interesting to know.
//
Now, this is not for a mabat design albeit quite close... I wonder how your 520G2 recreates a 1kHz sinc (the half positive part of a one sinus wave) measured at 30 cm 15 deg?
I created this stimuli in REW by Generator, Tones, Toneburst using half a cycle:
Run stimuli repetitive and at the same time run the REV Scope function (use the Trigger to make the wave form stay in the screen between pulses for easy viewing)
This is what I got when DUT was EQed flat 500-20k by the REV EQ function.
The filters used was: LR 8th 300Hz.
I had anticipated a more true replica of the pulse... the LF part in my speakers made more immediate copy with no strange reflections... (in this mesuement bass was off)
I have a little hard time to interpret the results. The first small positive "blipp" looks OK but if thats the stimuli and the rest is resonances then I'm surprised.. but perhaps it is what is happening!?
I hope this is still somewhat on topic on what is discussed here currently and is perhaps an alternative to the FFT analysis...
How yours is doing would be interesting to know.
//
I Didn't see much on how to use the jig so i just ran a lag bolt into this 2x6. Worked perfectly. I printed this in elegoo rapid petg using 1.2mm nozzle, 0.6 layer height, 1.65mm line width. I love using these large nozzles for prints that dont require detail. Not only does it take less time but the prints are much stronger. Bondtech sells a 1.8mm nozzle that ill likely try. All of these petals are hollow with 4 layer walls. If they resonate i will just fill them with plaster of paris mixed with water and pva glue. They will basically be solid stone at that point.
Edit forgot to mention i used liquid nails fuze it to glue this project. Petg is very hard to find adhesive for. I have not tried this yet so hopefully it works. This glue is a flexible polyeurethane glue and it spreads well. Not at all like liquid nails construction glue. Clean up with mineral spirits. I also pre applied superglue on all the edges to be adheared just to give it something more to bond to.
Edit forgot to mention i used liquid nails fuze it to glue this project. Petg is very hard to find adhesive for. I have not tried this yet so hopefully it works. This glue is a flexible polyeurethane glue and it spreads well. Not at all like liquid nails construction glue. Clean up with mineral spirits. I also pre applied superglue on all the edges to be adheared just to give it something more to bond to.
Last edited:
No speaker can reproduce DC component and I guess it has also quite a wide bandwidth. It may seem surprising but this is what you get when you pass such signal throuh a system with that impulse response (or a filter, if you want). With the knowledge of the IR of the DUT, you coud even calculate this output as a convolution of the test signal with the IR - there's actually no need to measure it.
It's really not an alternative, but only a very small subset of what we get by measuring impulse response, which then describes the complete behaviour of the linear system, including an output to your test signal, or any other. You would need to move into a highly nonlinear region to divert from these principles.
Last edited:
The only thing I can say is thumbs up 👍🙂
Very clever overall, you should get flawless acoustic results, IMHO.
Last edited:
Hello everyone,
in the last days and weeks I have been experimenting a lot with ath and abec. My goal is to design a rectangular horn for a 1.4” driver, with a high focus on constant directivity. I have tried the Tritonia horn from @mabat and several versions with different parameters. However, the simulated directivity with abec does not come close to what I want to achieve. When I compare it with EXAR 400, for example, the results are worlds apart. Is it generally more complicated to design a rectangular horn with constant directivity, or are there any “golden rules” on how to design a rectangular horn with constant directivity?
I would be grateful for any help.
Kind regards
Klaus
in the last days and weeks I have been experimenting a lot with ath and abec. My goal is to design a rectangular horn for a 1.4” driver, with a high focus on constant directivity. I have tried the Tritonia horn from @mabat and several versions with different parameters. However, the simulated directivity with abec does not come close to what I want to achieve. When I compare it with EXAR 400, for example, the results are worlds apart. Is it generally more complicated to design a rectangular horn with constant directivity, or are there any “golden rules” on how to design a rectangular horn with constant directivity?
I would be grateful for any help.
Kind regards
Klaus
Thanks - I should probably do a non-EQd measurement as well. Which driver was that?
The pos or neg start... isn't it so that a band limited system in fact have a negative onset before the main signal comes along?
In that case your phase might be right and what is seen is then the DC "removed" stimuli and a few "ringings".
If it is, that looks better than mine - actually looks very good. But I will remeasure.
//
The pos or neg start... isn't it so that a band limited system in fact have a negative onset before the main signal comes along?
In that case your phase might be right and what is seen is then the DC "removed" stimuli and a few "ringings".
If it is, that looks better than mine - actually looks very good. But I will remeasure.
//
Thanks for your vote of confidence, and your contributions. can you explain, or direct me to, how the t520-25-ext throat goes together please.
what hardware do i need? do i use both flanges? is one glued to the throat tube?
An IR contains all information of a linear system. A frequency response also, as the Fourier transform can transform both in the other. The time or frequency response to any input can than be determined by convolution in the time domain or multiplication in the frequency domain (see convolution theorem).
Sometimes a new representation shows other things more clear (IR shows echos more intuitively e.g.).
Does this 'sinc' function (I know it as sin(x)/x) show any aspects more clear?
it does. I will sort out the hardware with a trip to Home Depot. I'm not even sure what threads are used for my compression drivers.
Ah, I never remember the bolt lengths. It's always good to have several lengths at hand for experiments - even different drivers have different lengths of the threads, so what fits for one driver can be too long for another. Typically these are all M6, only the smallest drivers use M5 sometimes.
Measured today on 460 horn:
Results:
A bit disappointed this evening as I had quite some hope for the DCX mod. Well...maybe I get another idea how to improve. Suggestings welcome 🙂
- modified DCX464
- unmodified DCX464 with original 1.4 adapter
- BMS 4554 "long adapter"
Results:
- The insert kills the LF gain. I assume it cannot "breathe" enough through the smaller channel. Unfortunately I have no idea how I could change this without affecting the HF part. Red trace is unmodified DCX the others the angle measurements of the modified DCX....
- BMS performs as promised. Great work Marcel!
A bit disappointed this evening as I had quite some hope for the DCX mod. Well...maybe I get another idea how to improve. Suggestings welcome 🙂
I would try to sand down and harden and seal the surfaces with epoxy. At least to be sure that it doesn't help 🙂
But the loss is huge, hard to anticipate.
But the loss is huge, hard to anticipate.