• 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!

Carmody S2000 DIY Speaker Kit Review

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,390
Location
Seattle Area
This is a review and detailed measurements of the Carmody S2000 DIY speaker as built by our own member, @Rick Sykora. He has nicely documented his build of S2000 speaker from the kit. According to him the cost for a pair of these with everything included, sans labor to assemble is about US $244.

I like that the kit comes with Baltic Birch plywood which gives it a nice half-way finished look:

DIY Carmody S2000 Review Speaker DIY Kit.jpg


You could literally use it like this or finish it with a stain and top coat. I would personally use it as is.

And yes, these are very small speakers.

The back is not exciting other than showing the port:

DIY Carmody S2000 Review Speaker back panel port DIY Kit.jpg


The binding terminals Rick used are easy to turn since they are not inset in a cup like commercial speakers are.

Measurements that you are about to see were performed using the Klippel Near-field Scanner (NFS). This is a robotic measurement system that analyzes the speaker all around and is able (using advanced mathematics and dual scan) to subtract room reflections (so where I measure it doesn't matter). It also measures the speaker at close distance ("near-field") which sharply reduces the impact of room noise. Both of these factors enable testing in ordinary rooms yet results that can be more accurate than an anechoic chamber. In a nutshell, the measurements show the actual sound coming out of the speaker independent of the room.

I performed over 800 measurement which resulted in error rate of less than 1% up to about 3 kHz but then climbed up to 2% or so.

Temperature was 58 degrees F. Measurement location is at sea level so you compute the pressure.

Measurements are compliant with latest speaker research into what can predict the speaker preference and is standardized in CEA/CTA-2034 ANSI specifications. Likewise listening tests are performed per research that shows mono listening is much more revealing of differences between speakers than stereo or multichannel.

Reference axis was the tweeter center.

Carmody S2000 Measurements
Acoustic measurements can be grouped in a way that can be perceptually analyzed to determine how good a speaker is and how it can be used in a room. This so called spinorama shows us just about everything we need to know about the speaker with respect to tonality and some flaws:

DIY Carmody S2000 Speaker spinorama CEA2034 frequency Response Measurements.png


At first the measurements look very flawed but if we focus in, the response is actually quite good between 100 Hz to about 3 kHz. Then we lose energy before some kind of resonance around 11 kHz gives us a bit back. Directivity is also good until the graphs bunch up around 4 kHz to 5 kHz region (crossover is supposed to be 3 kHz). Smaller woofers tend to make directivity errors smaller because they are closer in size to the small tweeters.

Efficiency is around 84 dB which is fairly low as expected.

Early window reflections which are the strongest sum up to a smooth response:

DIY Carmody S2000 Speaker spinorama CEA2034 early window frequency Response Measurements.png


So room dependency should be low.

Putting the two together we get a surprisingly good predicted-in-room response:
DIY Carmody S2000 Speaker spinorama CEA2034 Predicted In-room frequency Response Measurements.png


I noted that the regression line slopes down to the tune of 10 dB. This will be a good exercise to see if this if the results are too dull or right.

Drive responses shows some mismatch leading up to the crossover region:

DIY Carmody S2000 Speaker driver frequency response measurements.png


Doesn't seem to be that the Crossover is really at 3 kHz. Note that this is the acoustic crossover, not electronic.

Impedance dips down to below 4 ohm as is typical in most speaker we measure:

DIY Carmody S2000 Speaker impedance and phase measurements.png


Beamwidth is pretty flat and reasonably wide:
DIY Carmody S2000 Speaker Horizontal Beamwidth measurements.png


Color coding the amplitude in contour map says the same:

DIY Carmody S2000 Speaker Horizontal directivity measurements.png


Vertical directivity is typical so try to stay from slightly below tweeter to above it:
DIY Carmody S2000 Speaker Vertical directivity measurements.png


Otherwise you fall in that response ditch and loose energy in the important upper mid-range area.

EDIT: forgot to post the distortion graphs originally:

DIY Carmody S2000 Speaker THD Relative Distortion measurements.png


DIY Carmody S2000 Speaker THD Distortion measurements.png


Carmody S2000 Speaker Listening Tests
Cursory look at the frequency response measurements would say this speaker should not sound good. But good sound it produced! Not only initially but through a full suite of tracks covering every genre and tonality. Most impressive were male vocals which had a wonderful warmth to them courtesy of the slight bass boost (and possibly reduction of high frequencies). Female vocals were nice too with no lispiness. Track after track was reproduced well and a nice diffused aura around the speaker.

A very nice characteristic was power handling. I could turn it up and up and it simply did not degrade like a lot of budget speakers do. When presented with deep bass, it simply did not play them. I have learned to appreciate this over a speaker that tries to play them but then distorts, bottoms out, etc. Don't do what you can't do I say!

There is one flaw there though: if the content of the music track is mostly low frequency, the level drops way down almost making you think someone lowered the volume way down. This made a few tracks not listenable. Actually I could listen to them by turning up the volume but then was afraid of the next loud note blowing them up! :)

Conclusions
I will confess that I hate doing reviews like this. Objective scores are likely to be low creating a lot of cries of "how could you like it so much." Early on in my speaker testing I realized that what is important to my ears is good response from 200 to 2000 Hz. Dips in that region can be really problematic as there is so much "life" in music in that region. Testing this speaker made me realize perhaps I don't like as much high frequency content and want a more sloped down response.

Something that doesn't jump out of the measurements and is a key reason I listen is power handling. Yes, I show you the distortion graphs but those are two points and you can't hear what I am hearing. In this case as I noted, the speaker incrementally gets louder and louder without falling apart like just about every budget speaker. I want to experience music, not the limitations of the speaker and the S2000 does a very good job here.

I imagine nothing is more complicated than getting a cheap, small speaker to sound good. The components are simplicity in themselves. But balancing the compromises must be a herculean affair. Whatever that formula is, it worked for my ears.

I suggest giving this build a try. If you don't want to build them yourself, ask Rick to build them for you. You will wind up with a unique speaker that sounds really good, and fits everywhere.

I am happy to give a strong recommendation to Carmody S2000 DIY Speaker.

------------
As always, questions, comments, recommendations, etc. are welcome.

Appreciate any donations using: https://www.audiosciencereview.com/forum/index.php?threads/how-to-support-audio-science-review.8150/
 

Attachments

  • Carmody S2000 DIY Speaker Spinorama.zip
    87.1 KB · Views: 289
Last edited:

MZKM

Major Contributor
Forum Donor
Joined
Dec 1, 2018
Messages
4,240
Likes
11,463
Location
Land O’ Lakes, FL
Preference Rating
SCORE: 4.9
SCORE w/sub: 7.3

Sensitivity: 83.9dB (spec: 87dB)
Frequency response: +/-5.9dB 50Hz-20kHz ; +/-4.1dB 80Hz-20kHz


Spinorama (12).png

Horizontal Directivity (3).png
Horizontal Directivity Normalized (7).png
Vertical Directivity (3).png
Vertical Directivity Normalized (7).png
chart - 2020-11-12T191518.341.png
All graphs/data

__________
Measurement mic positioning accuracy

Slightly to the right of center horizontally:
chart - 2020-11-12T192231.029.png


Slightly below center vertically:
chart - 2020-11-12T192205.526.png
 
Last edited:

Xyrium

Addicted to Fun and Learning
Forum Donor
Joined
Aug 3, 2018
Messages
574
Likes
493
Distortion plots?

What's up with the gap around the port?

And I'm with you Amir, I like a downward sloping response as well. I'd almost wager that if a product had a bump at 20kHz, I wouldn't consciously hear it, but it would irritate me.
 

Rick Sykora

Major Contributor
Forum Donor
Joined
Jan 14, 2020
Messages
3,517
Likes
7,028
Location
Stow, Ohio USA
Distortion plots?

What's up with the gap around the port?

No gap. Used butyl rubber to seal and some appears to have oozed out while awaiting review. :oops:

It can be trimmed away.
 

Xyrium

Addicted to Fun and Learning
Forum Donor
Joined
Aug 3, 2018
Messages
574
Likes
493
No gap. Used butyl rubber to seal and some appears to have oozed out while awaiting review. :oops:

It can be trimmed away.

Sweet. That port sure isn't going anywhere! ;)
 

MZKM

Major Contributor
Forum Donor
Joined
Dec 1, 2018
Messages
4,240
Likes
11,463
Location
Land O’ Lakes, FL
So what PIR slope generates maximum preference in the formula?
I believe the way it works is, the steeper the better; which is an issue with how the formula works (it was made from "standard" speakers, so they didn't have to worry about super wide dispersion or say omni). The calculation is simply the correlation coeffient formula you would learn in any stats class:
1*T52-LSuLQyq-6I2c1lkj-A.png

For anyone wondering:
1) You have a linear (or log) regression line, basically what line best fits the data.
2) Find how far each measured point is from the regression.
3) See if that variance can be correlated with frequency (If there is more/less variance in SPL as frequency increases/decreases, the score is better; if the variance is "random", it scores poorly; so, if the variance is the same but the SPL is sloping down more than usual in the treble, that is seen as a better correlation).
 
Last edited:

Rick Sykora

Major Contributor
Forum Donor
Joined
Jan 14, 2020
Messages
3,517
Likes
7,028
Location
Stow, Ohio USA
Sweet. That port sure isn't going anywhere! ;)

I can be a little obsessive and must have overdone. IME, air leaks around the port often are a reason why a build did not match the design. It does not take much of a leak to shift the system impedance.
 
Last edited:
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,390
Location
Seattle Area
I believe the way it works is, the steeper the better; which is an issue with how the formula works (it was made from "standard" speakers, so they didn't have to worry about super wide dispersion or say omni). The calculation is simply the correlation coeffient formula you would learn in any stats class:
1*T52-LSuLQyq-6I2c1lkj-A.png
Thanks. So beyond those factors, having a speaker with less high frequency energy to begin with, would also do better.
 

Helicopter

Major Contributor
Forum Donor
Joined
Aug 13, 2020
Messages
2,693
Likes
3,944
Location
Michigan
Very nice to see this. With such good predicted in room response and power handling, it makes sense this speaker sounded good. Thanks Rick and Amir for putting this together.
 

Xyrium

Addicted to Fun and Learning
Forum Donor
Joined
Aug 3, 2018
Messages
574
Likes
493
I can be a little obsessive and much have overdone. IME, air leaks around the port often are a reason why a build did not match the design. It does not take much of a leak to shift the system impedance.

Funny you say this. I assembled a sealed version of Krutke's ZRT, and I sealed the inside seams with some sort of rubberized caulk. Then I did the same to the speaker terminals on the inside of the cab. I definitely went a little overboard, and can appreciate your attention to those types of details!
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,390
Location
Seattle Area
Slightly to the right of center:
index.php
With light and small speakers it is harder to align them well. Touch them and they move too much which was the case here.
 

MZKM

Major Contributor
Forum Donor
Joined
Dec 1, 2018
Messages
4,240
Likes
11,463
Location
Land O’ Lakes, FL
Thanks. So beyond those factors, having a speaker with less high frequency energy to begin with, would also do better.
I believe so.

Olive doesn't address this specifically in the paper, but he does state the ideal slope may be speaker dependent (I guess meaning, as long as there isn't a directivity mismatch, it doesn't matter). So, I don't know why he didn't normalize the PIR to its own slope, and then run the AAD or NBD on those, that way you are only looking for linearity and not slope. One can easily do this, and then we do have Spins and listener preference scores for the original 13 bookshelf speakers used if one wants to try and alter the formula.
 

MZKM

Major Contributor
Forum Donor
Joined
Dec 1, 2018
Messages
4,240
Likes
11,463
Location
Land O’ Lakes, FL
Yeh, with such small woofer and high crossover frequency, reference axis shifts below tweeter.
No, I mean the mic was too low, it shows that +10 degrees vertically has higher SPL.

With light and small speakers it is harder to align them well. Touch them and they move too much which was the case here.
I mentioned this elsewhere, but for how expensive the machine is, you'd think it would have an auto-centering feature where it plays a 20kHz sine wave and simply moves along a plane perpendicular to the speaker until it registers the highest SPL.
 
Top Bottom