OB H frame Emin Alpha 15A & CRH-70 - Help needed

andb

Member
2008-11-28 7:09 pm
Prague
I've read over MJK's Jordan + Goldwood and many threads here and now am interested in building an open baffle speaker. I'm asking for some help from more experienced members to make sure I get this right. I'm hoping to build using an Eminence Alpha 15A with a Mark Audio CHR-70. I have the MA's, which is the main reason for this project.

I'd like to use passive crossovers and push these speakers with an audiosector LM3875 chipamp and a soon to be built Triode electronics st-70 clone. I have a small listening room, about 4mx5m, with the listening area being half of the shorter axis. Living in a flat in the city prevents me from listening as loud as I might want to, so I don't need to go for really high volume levels.

I see these issues that I need to resolve:
1. Will this driver combination work well?
Frequency response seems to be well matched, but the sensitivity of the 15A seems to be 10db higher. What is the best way to deal with this? Can these two drivers play nicely together with just a passive crossover?

If the CHR-70 isn't a good driver for the project, Ill can build a lotus box for them, give them to a friend, and choose another driver.

2. Proper dimensions for the H frame bass baffle (Alpha 15A)
In MLJ's research, he used a internal cross section area of 16” wide and 16” tall. The depth of the cavity was defined as 7.5”. Based on pictures, is it correct that the cavity depths in front and behind the driver are symmetrical?

MJK wrote:
The efficiency of a U or H frame could be increased by using a larger cross-sectional area. The depth of U frame or either side of H frame enclosures should be set to place the quarter wavelength resonance above the selected low pass crossover frequency.

Based on this, is it correct that the cross-sectional area should be tuned based on the efficiency of the other drivers in the system, in this case the CHR-70? Is the 16" square appropriate in relation to this FR driver? It seems that it will be ok since the sub is already much more efficient then the full range driver.

3. Proper dimensions for the fullrange baffle (CHR-70)
Im pretty new to this so tips on baffle size and driver location would really be appreciated. I'm also very interested how to derive these values.

4. Crossover wiring
Will this combination be work with just a passive crossover? I want to understand whats going on so I've ordered Weems' "Designing, Building, and Testing Your Own Speaker System with Projects". Hopefully I'll get a better understanding of passive crossovers from this. It would be great to be told what crossover components to start with for these 2 drivers as a starting point. Fortunately this is something Ill be able to play with over time, but input is very welcome. Suggestions about good sources for audio quality electrical parts in the EU are welcome.


I'm hoping I'll be able to build something that makes real music with this, so thanks in advance for help!
Andrew

p.s. Im also preparing a Fostex FE126eN A126 build, Dave is working on definitive plans, then I'll get cutting! I expect it will be interesting to compare these two speakers when they are done.
 
I ran the simulation assuming an active filter (I know you specified passive) and it looks very good. The H-frame for the Alpha 15A is 15" long and internal dimensions are 17" x 17". I tried a baffle with the CHR-70 mounted on top of the H-Frame that is 18.5" wide by 19" tall with the CHR-70 mounted 7" in from the edge and 14" up. The active x-over was set to be a 110 Hz low pass and 400 Hz high pass with 2nd-order Butterworth filters.
 
My idea was two eminence bass speakers each channel, but how to fill that in, do I have now change the H frame to 17 x 34 inch? or still use single H frame for each speaker .

MJK talks about two drivers in one baffle in this small PDF, which is a bit hidden on his web page. Basicly you use the double Sd, but put it in a frame with the dimensions for two stacked drivers (17 x 34 inch). It looks funny in the graphics, but should work alright.

Rudolf
 
Thanks

with playing it go now a little,my is 17 x 36, with two eminence beta 15a I get to 25 hz 85dB but have a bump of 95 dB at 40 hz, simulate stuffing on the walls do make a very nice graph without the bump so maybe there are some refections, I think the hframe needs to not see each other but that cannot be simulatied.

See that I have to put in some changes so I can always fill in single driver but mathcad see two drivers in the to driver sheet I make.

Oke, but I have to learn more first so I put in duoble vas etc.
 
Last edited:
I have did this (see pdf) after first learning some, I am still green behind the ears with Martins sheets, but get slow the ping feeling.

Anyway, I have use two visatons wsp26 s speakers in a single H frame and a fostex full bandwith speaker.

like to get some ideas reactions, because I have a H frame without a baffle between the upper and lower woofer, normal stacking has there a baffle but the sheet don,t work with that anyway.
 

Attachments

  • visatonws26s-fostex.xmcd.zip
    211.1 KB · Views: 54
Last edited:
I have use two visatons wsp26 s speakers in a single H frame and a fostex full bandwith speaker.
Try to work with better precision. You got the divided Re of the Visatons wrong.
Why didn’t you enter the proposed “theoretical values” as Actual Component Values in the “Crossover Definition”?
You might lower the low pass filter frequency or raise the high pass filter frequency to avoid the response bump at 250 Hz.

like to get some ideas reactions, because I have a H frame without a baffle between the upper and lower woofer, normal stacking has there a baffle ...
If people put a brace between the upper and lower driver, it is only for mechanical reasons (and recommended). There is no acoustical necessity for it. And leaving it off doesn't change anything acoustically either.

Rudolf
 
Thank you very much for your advise, what this for my new technologie from Martin concerns, I have to practice a lot more to get it the good way, but I am have found the good way.

I a a amp designer, so, speakerboxes is for me new, what calculations concerns.

The bump I presume is because of fase reactions? there is a need of some distance between the two, (low and mid). Martin say I have to divide the RE when parallel speakers, I have the thiele parameters of visaton paralleled,
it is 2.85 ohm or specifications are not right, normal with one speaker it is 5.7 ohm I now that filtering is normally calculated with 8 ohm, Martin uses dc resistance, I have tryed to calculate filter components with a other software, using 8 and 4 ohm, but what do the sheet with the theoretical values” as Actual Component Values? .. I have already ask this to Martin but yet answerm so I wait, I have to learn this way of calculations, I am used on normal software for this, I already see that the Martin sheets are much more usefull, but I have to learn more.

But first I have to solve the winzip crash in windows 7 after a last update the winrar program crash with opening winrar or winzip, it annoying me very much, windows 7 gets more unstable with any update.. tomorrow I go further with testing and set here the result.

I will try your advise, thanks again.

kees
 
Last edited:
The bump I presume is because of fase reactions? there is a need of some distance between the two, (low and mid).
No phase reaction - the low and high pass filter frequencies need to be farther apart.
Martin say I have to divide the RE when parallel speakers, I have the thiele parameters of visaton paralleled, it is 2.85 ohm
Yes it is 2.85 Ohm, but you entered 2.35 Ohm in the sheet ;)
I have tryed to calculate filter components with a other software, using 8 and 4 ohm, but what do the sheet with the theoretical values” as Actual Component Values? ..
You have defined the crossover filter frequencies in the "Crossover Definition". From them the worksheet calculates the "Theoretical Values". You need to enter those values (rounded to the nearest "market" values) into the "Actual Component Values" column. See attached picture. The "Actual Component Values" are not calculated automatically!
There are detailed "Instructions" on how to do this below the "H frame Enclosure Geometry Definition". Did you read them? :rolleyes:
 

Attachments

  • MJK kees52.gif
    MJK kees52.gif
    19 KB · Views: 423
Ahh oke, a little mistake, I correct it, thanks.

Yes I did read them but boxes are not my thing, amp designing is more my thing
but I want to learn how to play with boxes because it is fun and give some feeling
of that everything is from my hands, and it is much more cheaper, I have also play
with the tl, and now after reading I have a excel sheet for calculate tl tube length
to put in the mathcad.

But after some studie I am surpised that I am already so far that I can simulate,
it is more difficuld then a box calc program, but it is more precise.

thanks for your help.
 
Last edited:
I have done the corrections, I now now how it works, thanks to you all.

I have now everything within plus minus 3 dB what is best I get until now.

I have used fostex speakers but also a visaton ws13ng basmid what I can us with a extra tweeter, this I can not simulate but is not a disaster.

now this all on my designed amp.

real diy testing, real fun. - YouTube

audio is recorder through a mobile phone so it is not big sound, but it is open and easy.

Just one thing left, what I have to do with this in the sheet, is this the speaker offset in h-frame?

(Fraction of the Total Length of the Front Cavity : 0.001 < x < 0.999)

thanks again.
 

Attachments

  • visatonws26s-fostex.xmcd.zip
    207.1 KB · Views: 12
Last edited:
Well it sing, and it go low, however, it is sensitive for where I am in my room, a meter for the box it has a beautifull bass, but 3 meters before the box the bas dropps for shure with 3 db minimal, I presume the h frame has to stand on the floor, and in a corner, what not so is with the testing.

the 0.5 is that I have to move the speaker baffle to correct for the baffle itselfs? can you explane some more for me please, thank you.

here is some music from youtube, the box sound this way, 30 hz, it go there easely but you can,t hear that with the camera recording, and is mono (use just one channel so you miss a lot).

The low pass for the bas speakers, not higher then 100 hz and minimum 18 db octave, the sound above 150 hrz is coulored like you wil hear in this video, here is is 6 db octave and 250 hz, for up, bessel fuction work best, 12 db or more, it give no fase problems..

I did hear from other members that the H frame coloured the sound, and so I have to say yes it does give hollow sound, so low pass not hihger then 100 and maybe even 80 hz.

First H frame test with visaton wsp26s - YouTube

regards
 
Last edited:
it is sensitive for where I am in my room, a meter for the box it has a beautifull bass, but 3 meters before the box the bas dropps for shure with 3 db minimal,
As with every loudspeaker, bass is sensitive to the room modes. This is nothing unique for a dipole.
I presume the h frame has to stand on the floor
Yes please, and without clothes next to it.
and in a corner
No, please!
the 0.5 is that I have to move the speaker baffle to correct for the baffle itselfs?
Sorry, I don't understand your question. You may need to visualize that.
here is some music from youtube, the box sound this way, 30 hz, it go there easely but you can,t hear that with the camera recording, and is mono (use just one channel so you miss a lot).
I don't judge sound quality from Youtube movies. No way.
The low pass for the bas speakers, not higher then 100 hz and minimum 18 db octave, the sound above 150 hrz is coulored like you wil hear in this video, here is is 6 db octave and 250 hz, for up, bessel fuction work best, 12 db or more, it give no fase problems..
Please follow MJKs method in how to equalise H frames. Start with a 6 dB low pass at 50 Hz or lower to get a linear response. Add a second low pass for the crossover.
I did hear from other members that the H frame coloured the sound, and so I have to say yes it does give hollow sound, so low pass not hihger then 100 and maybe even 80 hz.
Every change in baffle size changes the response of a dipole. I would not call it "colour". Please equalize the H frame in the right way and it will be useful up to 200 Hz.
 
Hi Rudolf

I have simulate with carpet in the room, I have change the low pass a little with what I have, so I must get the parts I need, but now it sound very nice, I have try with and without feedback on the hybrid amp, this amp is very good, the first in my live that make sound so easy.

I have test tones used to find the cut off, 27 hz and below bas is gone, 30 hz give 86 dB watt, so very close to martins sheet, impressive. holding a paper for the printer before the h frame a meter or so and the paper go as moving as the conus of the woofers itslefs.

And a other good thing, the bass don't boom through walls ang go to the nabures a normal box does.

I have still to use the bass, it is very different, colourless and as you now i have to use to it.

schematic of my amp design is here also.

thanks

What I did mean with the 0.5 the offset what is in the sheet, for me is 0.7 for a good respons together with the other fill ins, but how have I to use the speaker offset in the box I then make, shift the speaker 0.7 inch to the front? of back?

I mean this, how to implement.

ξ := 0.7 (Fraction of the Total Length of the Front Cavity : 0.001 < <
0.999)

I know one thing I get Goosebumps of the music even if I hear just one mono channel.

You say this

Every change in baffle size changes the response of a dipole. I would not call it "colour". Please equalize the H frame in the right way and it will be useful up to 200 Hz.

I do not equalize with a equalizer but with filtering, I go make a filter as used with baffle correction if I measure bumps and falls, but as I now hear it is quite liniair til the lowest region in music, with ease, I have only a hybrid amp and a volume pot that is al and the best.

Vocals are sooo beautifull from the amp and youtube if it sounds there nice it is alays much better in real time.

regards
 

Attachments

  • ScreenHunter_01 Apr. 21 19.08.gif
    ScreenHunter_01 Apr. 21 19.08.gif
    37.8 KB · Views: 355
  • Frame_Corner_Passive_visatonws26s-fostex.zip
    209.8 KB · Views: 17
Last edited:
What I did mean with the 0.5 the offset what is in the sheet, for me is 0.7 for a good respons together with the other fill ins, but how have I to use the speaker offset in the box I then make, shift the speaker 0.7 inch to the front? of back?
Just look at your attached worksheet :D
Page 12 of your PDF, bottom "Side view". There you see how the baffle is moved backward with ξ := 0.7 in the H frame.
BTW: I've never seen any advantage in building such an asymmetric H frame.
I know one thing I get Goosebumps of the music even if I hear just one mono channel.
Mission accomplished. What to expect more. :)

Probably you should change "radius := 1m" in the "Listening Position" at page 11 to a more realstic value. But that is marginal. ;)

Rudolf