My new OB Speaker

[Hari Iyer]

After using my current open baffle for almost 6 years, I am now planning a new OB project venture for my own use. The current baffle had some issues with size w.r.t bass performance and also visually blocks my TV for folks to watch due to their 14.5 x 50 inch baffle in my small living room. Folks at home have to move the baffle aside to watch TV which was not comfortable for me on a daily basis. This new design will need to address this issue if it has to have a WAF and DAF ( Daughter).
Some designs came up to my mind like having a dual push-pull type H- frame subwoofer that can be crossed below 150hz and the OB mids taking over the rest of the bandwidth. As the OB needs to handle only beyond 150Hz ( which is my planned -6dB point), they can roll-off from 300Hz. The baffle need not be designed for low end duty and can have a narrow profile. I shall contine using the same drivers as I don't think anything wrong with them. Listing them here once again,
1. Cleveland Audio 12" Alnico woofer , reconed with Jupiter paper and Aluminium Vc former
2. Ahuja 12" ferrite woofer, again reconed with Jupiter paper and Audiotone VC
3. Sachenwerk 8" field coil driver
4. JBL Selenium CD

Below is the final design idea.





BTW, this design would not have seen light of the day without the help of FM @sadik who was generous enough for doing all the Autocad diagrams and answering all my queries. for that. I have sourced Pinewood for this project and there are mixed opinions about this in the internet. I would like to read less about the messaging that are penned about this wood and would go ahead building with this wood and discover for myself abouts it's pros and cons. Generalization in audio have never worked for me as each project is different imo. The current status about this project is today the wood is been cut at a CNC routing workshop and delivery is expected around Tuesday next week. Shall update this thread with further details. Stay tuned.
Below is the simulated response for the narrow baffle with the field coil driver,


Thanks for looking.

 

[arj]

...his new design will need to address this issue if it has to have a WAF and DAF ( Daughter).

While WAF and DAF is the acceptance of an object by the husband , HAF is the acceptance ( or objection) of the Husband By the Wife/daughter..

BTW, this design would not have seen light of the day without the help of FM @sadik who was generous enough for doing all the Autocad diagrams and answering all my queries. for that. I have sourced Pinewood for this project and there are mixed opinions about this in the internet. I would like to read less about the messaging that are penned about this wood and would go ahead building with this wood and discover for myself abouts it's pros and cons. Generalization in audio have never worked for me as each project is different imo. The current status about this project is today the wood is been cut at a CNC routing workshop and delivery is expected around Tuesday next week. Shall update this thread with further details. Stay tuned.
Below is the simulated response for the narrow baffle with the field coil driver,


Thanks for looking.

@sadik is a Champ ! Really looking forward to your experiment on this one

 

[Hari Iyer]

Hari, H frames do dig low but essentially sound like subs. If you get your desired response in U frames, then think about them. U frames are musical and give you a freedom of crossing higher too. And why not use exactly same dual woofers ?

Even if you decide H frames, don’t do Push Pull. They sound insipid. I consider H frames to be essentially dipole subs which can be crossed from 140hz. (Higher order slopes)

I initially thought about Uframe, but the panel dimensions become unwieldy for my available living room space TBH. The main goal of this project is to provide some walking space to family members and not hijack more space . The hframe provides a better compromise on a space vs low end extension parameter. For my baffle height of 25" the F6 is around 270Hz. I shall cross them an octave less around 135Hz for my -6dB to be around 190Hz. Initially I shall fine tune with my 31+31 band dbx analgue equalizer as that's easy to do. Once finalized shall add a passive line level low pass filter to remove any coloration from active EQ.

 

[Hari Iyer]

Made some more simulations with the exact baffle size as beow,

 

[Hari Iyer]

Completed assembley of the OB and H- frames. Version 1 of voicing too completed. They sound awesome and gave me goosebumps - ironically with the same drivers but a different design approach. Even the crossover network is same.

 

[Yogesh]

Completed assembley of the OB and H- frames. Version 1 of voicing too completed. They sound awesome and gave me goosebumps - ironically with the same drivers but a different design approach. Even the crossover network is same.

Pictures or it never happened!

 

[Hari Iyer]

Pictures or it never happened!

Your belief is important. Nothing else matter. Come over and discover.

 

[Hari Iyer]

Could manage 2 coats of linseed oil last weekend.

 

[arj]

Hi hari, is this based on the Linkwitz design ?

 

[Hari Iyer]

Hi hari, is this based on the Linkwitz design ?

Yes and mo. The narrow open baffle mimicks the Linkwitz LX 521.4 shape. But the similarities end there. The LX is a 4-way active crossover design. Mine is a 3way hybrid passive design.

The tubes monoblocks are connected to the mids and highs. The mids are not having any crossover and they play flat from 300Hz rill 3khz. The roll-off is smooth hence don't need any filter. The highs use a -12dB/ octave passive filter.

The LX use a w-frame subwoofer box with a different kind of woofers. I am using what I already have and mine is a H-frame design. They are having a passive LPF @ 140Hz (-3dB) as a line level filter. This is fed to my 20 year old Denon 1604 AVR to drive them.

 

[Hari Iyer]

Earlier I used a dbx analogue GE for low frequency rolloff and bass boost from 20hz to 60hz. Though they sounded very nice and extended, I found them to be a little coloured what the recording engineer did not intend to have.. Hence now removed the GE and went the passive LPF route. With this the bass extension will be around 45Hz compared to the previous 30hz, but more balanced.

 

[Hari Iyer]

Finally today went for a major modification of my OB setup,
- I removed the Denon AVR and connected the h-frames directly to my tubes parallel to the OB. The bass was a bit low as they were getting masked by mids and highs from the OB - but was there.
- I added a fixed LP shelving # -3dB begin from 340hz and with -3dB point around 480hz. This bought the mids and highs at level with the lows
- the speaker impedance for the tubes dropped from 4 ohms to 1.7 ohms, but there was no fatigue or struggle on the amplifier to play this load.
- Next step is to try a single 1.25mH inductor on the H- frame to rolloff them at 340Hz at -6db/ octave and check if there is any further improvement in the bass.

FM @captrajesh is planning listen tomorrow evening. Hope to complete all mods by then for a critical review-

 

[Hari Iyer]

It took me two months to arrive at this final crossover version. I have been listening this for past one week and have no complaints. Current crossover uses all Keltron Electrolytic capacitors and inductors wound by me. The low frequency is biamped and crosses @ 150hz. I have ordered audio grade oil filled capacitors and better air core inductors for the final assembly. This hopefully shall be assembled this weekend. Thanks for looking.

PS: My laptop broke last month and in a way it was blessing in disguise. I could not measure the speakers due to this and had to rely on my ears for designing. What I have noticed is - you can only measure, measure and measure and arrive at your target response that is flat and you can also listen, listen and listen and arrive at the same target response. But when you listen to both system they are going to sound completely different. My experience says that, it really does not matter if you only measure and design but If you listen and design the outcome is much much better. In my case earlier Mt midrange was crossed at 3000hz by measuring and by listening I realized that the design frequency should be at 1500hz. Also by listening I realized that the tweeter needs to be design frequency should be at 3000hz. So when the two sum up at 2121hz they actually crossover with a flat summed response. I had measured this on a KRK audio spectrum on my Samsung TAB to crosscheck.

 

[Hari Iyer]

Some more modifications done on the crossover. Now it's 4th order LR for both drivers. Have completed assembly with a total of 13 components. 8 are filter components and rest for impedance and level matching. All capacitors are oil filled type and resistors are hand wound by me with magnanin wires. One Lpad is made out of nichrome wires. Inductors are all aircore commercial type. Bypass capacitor are orange drops. With this I shall be concluding this project as there doesn't seem to be much scope for further fine tuning. Now with is project my speaker building experience increase from 3% to 5% -
Thanks for looking.

 

[Hari Iyer]

As my laptop broke a month and a half ago, I could not use REW to design or measure the crossover and had to rely on my ears to design the crossover and my years of accumulated skills. I have used an online log frequency sweep generator and KRK audio spectrum analyzer just now for the first time after they are assembled 2 days ago. Have a look. Thanks.

 

[Hari Iyer]

Listened to my new 4th oder LR crossover for a week with audiophile components etc. I have to admit that for my setup and ears, I shall prefer a 1st order crossover anytime. The 4th order crossover is having a poor transient response and is very sluggish and sounds quite laid-back. What I find is the lack of energy and dynamism in the music which gives me a feeling something is not right. They do measure quite flat and do not have any major peaks or dips in their response, but they don't sound involving. You can always argue that listening is subjective and maybe my ears prefer the added distortion and coloration introduced by the cone breakups etc.

My new laptop arrives next week and itcwill be interesting to measure these crossover with the REW installed. Till that time I shall not work on anything new or modify the current setup. Shall post an update as appropriate.

 

[Hari Iyer]

Will be designing a first order crossover with just Resistor and Capacitor and omit the inductor totally. Inductor makes the crossover a bit sluggish. The design frequency will be an octave less for the woofer and an octave more for the tweeter as the rolloff is shallow.

 

[Hari Iyer]

A few images before and after fixing REW and adjusting levels in my 4th order LR crossover,

REW setup issue,


Before level adjust in crossover,

After Level adjust in crossover

crossover freq is 220Hz and 2121Hz

 

[Hari Iyer]

After living with the 4th order LR crossover for a month, i finally decided to move away from it and go back to the humble first order crossover. The reason for this movement was,
- Poor dynamic response of the speaker. Loud and soft passages in music was not playing well
- Poor trainsent response. Orchestra music was sounding horrible and so was Jazz music not playing well
- Laid back response of the speaker was making it to sound very boring.

Ironically the speakers did not measure bad as you can see below,


They are reasonably flat from 70Hz till 20KHz with no major peak or dip in FR. Also the minimum phase is almost at zero deg for the entire BW.

Below is the step response,



You can clearly see the 360 deg phase shift between the drivers and as per commercial standards this step response is quite good for a 4th order speaker and is generally acceptable shape where the tweeter rises and at the fall of the tweeter the mid-range picksup and later at the fall of the midrange the woofer picks up without any disconnect.

If you decided to buy this speaker only by looking at the measurements then you could be disappointed.

After listening to this speaker with various music for a month, i decided to switch back to my 1st order BW crossover. This crossover is a bit different than the usual stuff where you use only a single inductor and single capacitor for the network. Instead of using an inductor, i used a resistor to make an RC low-pass filter for the woofer and similar RC high-pass filter for the tweeter. Here are the measurements as below,



The FR is similar to the 4th order network with a very good minimum phase response.



The classic right triangle step can be seen with a very good dirver integration between highs, mids and lows without any discontinuity.

Subjectively both the 1st order and 4th order has no comparison / competition. You can listen to the 1st order for an entire day and still wanting more for the smoothness of the FR, the dynamic range in soft / loud passages and its excellent transient response. Vocals sound as if its happening in front of you and turning left / right to check in-between is a distraction i- i would say. The low-end response integration with the mids and highs is so good that it needs to be heard to be belived. All in all a well thought out modification and not relying only on the obejective performance and challenging it with subjective follow-up. For end-users it should always be the listening performance that should be the final verdict and not get lost in the objective world as it can very well fool you easily.

Thanks for looking.

 

[Hari Iyer]

Finally the cherry wood wave guide has seen light of the day. Still it's work in process. Hoping it to be delivered by this weekend.