• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Revel M105 Copy DIY build

LTig

Master Contributor
Forum Donor
Joined
Feb 27, 2019
Messages
5,760
Likes
9,442
Location
Europe
See section 1.3. ;) You want to alter the actual impedance not just the SPL of problem area.
I did read it. The claim is to lower the impedance of the crossover as seen by the driver to prevent higher distortion.

In configuration #2 the 5 kHz series notch filter increases the impedance. In configuration #3 the 10 kHz parallel filter does decrease impedance but it's still at least 300 mOhm (L3) and in series with the 5 kHz notch filter. In an active speaker the driver sees only the output impedance of the power amp which is easily 10 times lower than L3.
 

fluid

Addicted to Fun and Learning
Joined
Apr 19, 2021
Messages
691
Likes
1,196
I did read it. The claim is to lower the impedance of the crossover as seen by the driver to prevent higher distortion.
This is not how I read their application note.

Purifi Impedance.png

The parallel and series notches produce much the same effect on the main peak frequency, but the extra impedance of the series notch lowers the sub-harmonic distortion as well.

Series Parallel Notch.png
 

LTig

Master Contributor
Forum Donor
Joined
Feb 27, 2019
Messages
5,760
Likes
9,442
Location
Europe
This is not how I read their application note.

View attachment 213084
The parallel and series notches produce much the same effect on the main peak frequency, but the extra impedance of the series notch lowers the sub-harmonic distortion as well.

View attachment 213088
What you cite has the same effect as using an EQ in front of the power amp which lowers the level of the voltage source in the first place.

I had the understanding that the parallel notch filter has a damping effect.
 

Elfsberg

Member
Joined
Feb 14, 2022
Messages
50
Likes
26
Location
Sweden, Falun
Your DSP equalization just alters frequency response, right? This circuit alters driver motor behaviour.
 

LTig

Master Contributor
Forum Donor
Joined
Feb 27, 2019
Messages
5,760
Likes
9,442
Location
Europe
Your DSP equalization just alters frequency response, right? This circuit alters driver motor behaviour.
Yes. I don't understand how the circuit changes the driver motor behaviour, other than through damping which an active speaker would do better.
 

fluid

Addicted to Fun and Learning
Joined
Apr 19, 2021
Messages
691
Likes
1,196
What you cite has the same effect as using an EQ in front of the power amp which lowers the level of the voltage source in the first place.
No it doesn't and this is maybe why the whole thing does not make sense to you. The Parallel notch is much the same as using an active EQ filter.

A series notch made out of passive components does something that normally might not be wanted, create a high impedance at the filter frequency. But it is this high impedance at the peak frequency that supresses the SPL of the apparent voltage source in series and reduces the motor distortion lower in frequency.

The low and consistent impedance of an active crossover (which is normally a benefit) stops it from doing what the series notch is meant to do.
 

LTig

Master Contributor
Forum Donor
Joined
Feb 27, 2019
Messages
5,760
Likes
9,442
Location
Europe
No it doesn't and this is maybe why the whole thing does not make sense to you. The Parallel notch is much the same as using an active EQ filter.

A series notch made out of passive components does something that normally might not be wanted, create a high impedance at the filter frequency. But it is this high impedance at the peak frequency that supresses the SPL of the apparent voltage source in series and reduces the motor distortion lower in frequency.
So if I reduce the voltage using EQ exactly as the passive notch filter does (amplitude and phase), the driver will see the exact same voltage while being connected directly to the power amp. If it sees the exact same voltage it behaves exactly the same, except that any damping effects will be stronger due to the low output impedance of the power amp, which is a good thing.
 

fluid

Addicted to Fun and Learning
Joined
Apr 19, 2021
Messages
691
Likes
1,196
So if I reduce the voltage using EQ exactly as the passive notch filter does (amplitude and phase), the driver will see the exact same voltage while being connected directly to the power amp. If it sees the exact same voltage it behaves exactly the same, except that any damping effects will be stronger due to the low output impedance of the power amp, which is a good thing.
No it is not possible to make an active filter behave in the same way as a passive serial notch if the impedance is kept low. If the active section uses current drive then the impedance will be high and a similar effect to the serial notch will be seen. Bruno does this in the Kii 3, current drive amps for midrange distortion reduction.
 

LTig

Master Contributor
Forum Donor
Joined
Feb 27, 2019
Messages
5,760
Likes
9,442
Location
Europe
No it is not possible to make an active filter behave in the same way as a passive serial notch if the impedance is kept low. If the active section uses current drive then the impedance will be high and a similar effect to the serial notch will be seen. Bruno does this in the Kii 3, current drive amps for midrange distortion reduction.
With respect to this paper current drive prevents distortion due to voice coil heating. The notchfilter will behave somewhat (!) similar insofar as it increases the impedance of the feeding circuit so that voice coil heating will in fact lead to a somewhat higher voltage, thereby partially compensating the loss in SPL (partially compared to a real current drive amp). However this is at one specific frequency only. For all other frequencies the increase in impedance is negligible - in my view.
 

fluid

Addicted to Fun and Learning
Joined
Apr 19, 2021
Messages
691
Likes
1,196
With respect to this paper current drive prevents distortion due to voice coil heating.
That is part of it but the paper's conclusion says
"The principal advantages of current drive are seen to be an elimination of performance dependence on voicecoil resistance (which is thermally modulated) and also coil-inductive effects, which give rise to high-frequency distortion, along with nonlinear electromagnetic damping due to BI variations".

The notchfilter will behave somewhat (!) similar insofar as it increases the impedance of the feeding circuit so that voice coil heating will in fact lead to a somewhat higher voltage, thereby partially compensating the loss in SPL (partially compared to a real current drive amp). However this is at one specific frequency only. For all other frequencies the increase in impedance is negligible - in my view.
Indeed but because the high impedance is targeted at the main peak it's effective at reducing the sub harmonics. The current drive amplifier could provide the high impedance over a wide bandwidth and do away with the need for the passive series components.

The passive series notch is just a simple way to get most of the benefits without changing anything else in the system.
 

DanielT

Major Contributor
Joined
Oct 10, 2020
Messages
4,751
Likes
4,633
Location
Sweden - Слава Україні
One thing I was wondering why did you , McFly, remove the dispersion lens on the SB26ADC-C000-4?

You didn't think about adding a notch filter to push down the FR on your midrange after the crossover point? Maybe it is addressed in the thread? Apologies if I missed it in that case.:)

 
OP
McFly

McFly

Addicted to Fun and Learning
Joined
Mar 12, 2019
Messages
903
Likes
1,861
Location
NZ
@DanielT The factory phase plug/dispersion lens on the SB26ADC is removed to adapt to augerpros waveguide, and the waveguide with phase shield from augerpro has some physical printing issues, its too small to print consistently. The function of the factory (and augerpro's waveguide) phase shield acoustically, IMO, just pushes the aluminum domes resonant peak at ~24khz from ON axis to OFF axis, it doesn't actually get rid of it entirely.

And also I don't care, because I cant hear past 16.5kHz, and the 24kHz bell mode doesn't fold down into the audible band via IMD, like it does with the SEAS DXT tweeter. At least in my measurements.
 

DanielT

Major Contributor
Joined
Oct 10, 2020
Messages
4,751
Likes
4,633
Location
Sweden - Слава Україні
Ok why I asked the question was this, but please note it is for the tweeter by itself and not mounted in a wavegudie. Response from among a commercial speaker manufacturer.

Question:

RolffRojs:
Is there a better tweeter than the SB Acoustics SB26ADC-C000-4 between SEK 500-700
(about $60) ? Would cutting the center plug on this be a good idea as it seems to cause a choppy frequency response near 20kHz? This has previously been done on a SEAS H534 which got better spread after the mod

I-or:
Not if you ask me. You can be absolutely sure that the frequency response/dispersion will be much more uneven without the phase plug.

Petersteindel:
I agree with I-or. Forget about pinching the middle part. Without the center phase plug, you can throw the element away.


Peter's speaker:


 
OP
McFly

McFly

Addicted to Fun and Learning
Joined
Mar 12, 2019
Messages
903
Likes
1,861
Location
NZ
If I were using the SB26ADC without a waveguide, I'd leave the phase plug there. Also - good dome protection! I've already smooshed one of my SB26ADC domes after moving it into a waveguide :facepalm:
 

DanielT

Major Contributor
Joined
Oct 10, 2020
Messages
4,751
Likes
4,633
Location
Sweden - Слава Україні
Sounds reasonable.:) Sad with your smooshing. Shit happens, but it's just part of fiddling with things.:)
You get to say like Lemmy in the song Ace of Spades:

You win some, lose some, all the same to me....
The pleasure is to play, makes no difference what you say


Hm, I know the hell how to do it, because I have the same drivers as you plus bass drivers for a three-way construction. If I'm going to construct without a waveguide but with a wide baffle and round baffle corners. Something like this Heco:
4406765.jpg


Or with Waveguide, which my friend can print out.

I have to take a deeper look at how I will place the speakers in the room, what might be the appropriate design principle with the tweeter then.:)

Edit:
By the way, very impressive your DIY I must say!:)
 
Last edited:

Trdat

Addicted to Fun and Learning
Forum Donor
Joined
Sep 6, 2019
Messages
967
Likes
396
Location
Yerevan "Sydney Born"
While patiently waiting for a calibrated mic to turn up, here's a teaser on another speaker I'm working on - a scaled up version of this concept with an augerpro 8" waveguide. Output much higher than the 105 copy.

Hard to get flatten the power response warble around the XO, being a two way. True sensitivity is 88db (not shown). Keens might be able to pick the woofer by the breakup 'hint'

Another bonus is this is a dead plain box speaker, no roundovers.
You didn't really specify which woofer you used here?

Also, if you have finished a project similar to the Revel M105 with a decent but affordable woofer with the Augerpro 8'' it would be great if you can post the simulation including the crosssover components or the scaled up version option you mention above. It seems the orignal you posted is great, happy with it just that a 5 inch woofer seems too small that's all.
 

Paweł L

Member
Joined
Aug 18, 2021
Messages
94
Likes
45
The midwoofer used is the Satori Textreme cone. I think it could be the MW19TX, it's 4ohm version (ah, that 4-5kHz break up). The MW16TX looks quite good in Vance's measurements - no dip past cone edge resonance.

The CRC cone from the more affordable SB series is supposedly the best sounding version - possibly no ALU glare McFly described, no paper mushiness.
 
Last edited:
OP
McFly

McFly

Addicted to Fun and Learning
Joined
Mar 12, 2019
Messages
903
Likes
1,861
Location
NZ
Finally found some time to make some tweaks. I don't think I can get this 2-way sounding any better - I could go Be tweeter and purifi woofer, but I really think it'd be diminishing returns. I don't know if there's a better all rounder woofer than the scan speak.

For those that have just jumped in - 2 way, 12L bass reflex box, 15w8530-k00 woofer and Sb26adc tweeter in augerpros (somasonus - linked here) 5 inch waveguide with no phase shield.

I have the 4ohm poly version of the SB woofer here, so will try a more budget oriented build soon. Augerpro's waveguides are what makes these builds work so well.

REV 6 Six-pack.png

REV 6 Six-pack Z.png

REV 6 XO-schema-1.png
 
OP
McFly

McFly

Addicted to Fun and Learning
Joined
Mar 12, 2019
Messages
903
Likes
1,861
Location
NZ
Also, as promised, for those that want to play with the project, attached the response files and the VCAD project file.
 

Attachments

  • 2WAY SB26 15W WAVEGUIDED VCAD.zip
    329.2 KB · Views: 300
OP
McFly

McFly

Addicted to Fun and Learning
Joined
Mar 12, 2019
Messages
903
Likes
1,861
Location
NZ
And here is the as-built for comparison to the simulation above. At this stage I only had time for the 1 measurement, will do full spins next weekend.

15w8530 sb26adc as built 1m hor 0 tweet axis.jpg


+/- 1dB within the limitations of the 4ms window. Unbelievably flat for a passive. Both speakers tested red and black, they only deviate by 0.25db max.
 
Top Bottom