Carmody S2000 DIY Speaker Kit Review

[amirm]

@amirm could not easily remove the crossover, but using meter, confirmed the 5 ohm resistor is inline with tweeter.

Thanks. So that should be tunable then. Has there been discussions about this elsewhere on this design?

 

[amirm]

Two things:
An impedance sweep is easy. Especially of a speaker. There shouldn’t be a noticeable difference between setups.

There can actually. My measurements are using Kelvin setup (4-wire) to remove the effect of the speaker cable for example. We are measuring pretty small values and even connection robustness can make a difference.

 

[andreasmaaan]

Regardless of all this, a difference of <0.4 Ohm wouldn't account for the measured difference in frequency response, anyway.

 

[hardisj]

Regardless of all this, a difference of <0.4 Ohm wouldn't account for the measured difference in frequency response, anyway.

bingo

My reply earlier was “is there a difference? I can’t tell, but it doesn’t look like it. If there is, that might explain things, though.”

that’s all I was trying to find out.

 

[Ericglo]

Are these things safe to use indoors as far as carbon monoxide and such?

I used mine indoors all the time. My only issue was lighting it would create an initial odor. After that, it was great.

I gave it to my dad after I moved. I used it in his garage and it worked great in there.

One other thing vs electric. If you ever have a power outage then this is a godsend. More than once I was left without power in an ice storm.

 

[amirm]

One other thing vs electric. If you ever have a power outage then this is a godsend. More than once I was left without power in an ice storm.

I can appreciate that. We have a generator for that It is a beast and sits outside, powered by a GM industrial engine!

 

[Rick Sykora]

Have not seen any...

A few thoughts in that respect though:

1. Of all Carmody’s designs (they are numerous) this one does not have much documentation. Notably had no measurements to validate build against.
2. TBH, at this price point, see most folks just enjoying the results of their work. Unless they are more serious hobbyists, they often lack the equipment to measure them anyway.
3. Between my measurements and listening, thought they were a pretty decent value. While it is easy to nitpick at some of the measurements, they do a number of things well too.

 

[andreasmaaan]

bingo

My reply earlier was “is there a difference? I can’t tell, but it doesn’t look like it. If there is, that might explain things, though.”

that’s all I was trying to find out.

Yeh, a <5% difference in impedance in that range would cause a <1dB difference in output. It looks like the difference between Rick and Amir's FRs is a >3 dB in this case.

 

[Maiky76]

. You did not post the radar plot for F328Be. Can we see them side by side?

Here are some thoughts on the EQ and a comparison with the F328Be

Spinorama with noEQ
Score: 4.91 , with sub: 7.32

Directivity:
Not bad on the horizontal plan.
Better stay at tweeter height or slightly above

Consistent LW except when listened under the tweeter height.

EQ design:
I was worried about the HF but it came out OK in the end.
I did not try to boost LF as the driver is already strained.
A lot of efforts required...

Score with EQ: 6.07, with Sub: 8.49
APO config file attached.

Carmody S2000 APO EQ 96000Hz
November132020-111537

Preamp: -4 dB

Filter 1: ON PK Fc 206 Hz Gain -2.22 dB Q 0.9
Filter 2: ON PK Fc 1140 Hz Gain -1.67 dB Q 2
Filter 3: ON PK Fc 1597 Hz Gain -1.67 dB Q 13.4
Filter 4: ON PK Fc 2595 Hz Gain -1.86 dB Q 6.54
Filter 5: ON PK Fc 3356 Hz Gain -2.23 dB Q 3.18
Filter 6: ON PK Fc 5128 Hz Gain 1.86 dB Q 2
Filter 7: ON PK Fc 9382 Hz Gain 3.3 dB Q 2.83
Filter 8: ON PK Fc 17454 Hz Gain 3.9 dB Q 4.47

Zoom PIR-LW-ON

Regression - Tonal:

Spinorama with EQ:

Noticeable improvements

Comparison with the Revel M105

Out of the box: no contest, M105 better

With EQ on Both speakers:
Very close. Not sure wether the Carmody will play as loud/cleanly though.
The M105 looks smoother and seems to have a better LW...

Finally the Comparison with the F328Be
The Carmony tracks the Revel pretty well up to the tweeter range

And the radar for @amirm

The rest of the data is attached.

 

[Mudjock]

I am assuming the tweeter response is so low just due to the designer going a tiny bit overboard in taming it. As even though is does rise up, the PIR curve shows it is a smidge too low

I first met the designer about 13 years ago (his first DIY event). He does have a recording background and tends to voice his designs to not sound harsh even on poorly recorded material, so treble can get a little recessed, especially in the BBC dip area (3-5 kHz). With my designs, I have noticed that Amir's NFS rolls off the response somewhat above 15 kHz (due to fewer data points being taken up there IIRC), so the agreement between Rick's measurements and the NFS is a little closer than it first appears.

Looking at that crossover, I'm guessing the 5 ohm resistor is in series with the tweeter and the 15 ohm is in parallel.

I'm glad to see that these did well, albeit not without controversy. Thanks to Rick for the build, Amir for the measurements and review.

 

[amirm]

And the radar for @amirm

Good grief!!! Something is seriously wrong with the scoring system or my measurements of F328Be. Then again it demonstrates why I liked the sound of S2000.

 

[JohnBooty]

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.

When it comes to power handling, now imagine the MTM version!

I built the MTM version of the S2000 for my brother in law (lucky guy!) and was super impressed. I would agree with Paul Carmody's description of the MTMs: "I honestly don't know where the limit is on these, but they were playing at 95 dB at my house and didn't seem to be breaking a sweat."

Considering their inexpensive price and small cabinet size, the power handling and output of the S2000 MTM (and their siblings, the Overnight Sensation MTM) designs are fairly unreal. Both MTM designs are about 91dB efficient and will happily take 100W.

baltic birch

Always a treat IMO when a kit is baltic birth instead of MDF. Obviously baltic birch is impractical above a certain size, but I generally prefer a nice stained look to a painted look. And staining is fairly idiot-proof, relative to painting.

Birch isn't as dense or inert as MDF, but you can always add some Dynamat to the inside of a birch cabinet. Sure helps it pass the "knuckle rap test."

 

[Newman]

So, a speaker that measures significant flaws but sounds subjectively good in sighted listening to one person, gets a knocked-it-out-of-the-park panther.

Got it.

 

[JohnBooty]

So, a speaker that measures significant flaws but sounds subjectively good in sighted listening to one person, gets a knocked-it-out-of-the-park panther.
Got it.

The estimated in-room response is good and ultimately that's what matters more than on-axis anechoic measurements, unless of course you live in an anechoic chamber.

Also, I think folks greatly overestimate the impact of imperfections in the high treble range. The fundamental frequencies of most musical content is octaves below the "problem" areas of this speaker's direct frequency response.

I do agree with the fundamental premise of ASR -- we should always be striving for engineering excellence because correct objective performance will lead us to the best sound. But, it is also always interesting to correlate those measurements with actual human enjoyment because that is also the end goal.

 

[ferongr]

If I get super rich from this site one day , I can put a mini-split system in there.

USian mini-split prices are outrageous. I paid 350€ + 60€ installation labor for a 12000 BTU inverter mini-split here (Greece).

 

[amirm]

USian mini-split prices are outrageouts. I paid 350€ + 60€ installation labor for a 12000 BTU inverter mini-split here (Greece).

Yes, I don't understand why. My brother lived in UK for a while and told me to get one of them years ago. When I told him how expensive they were here, he could not believe it. I had one installed for my woodshop and it cost $3,000 I think! He too said it cost 350 pounds or so to get one.

Part of the problem is high cost of labor here. And the units require a certified HVAC technical to charge them or some such thing.

 

[Hemi-Demon]

So, a speaker that measures significant flaws but sounds subjectively good in sighted listening to one person, gets a knocked-it-out-of-the-park panther.

Got it.

Gotta agree with this take. The subjective rating seems to go against rather poor measurables. Is the value in it being so cheap?

 

[ferongr]

All mini-splits sold here come precharged with enough refrigerant for something like 10 meters of line. Installation usually takes less than an hour. The most labor intensive part of the job is preparing the lines and wrapping them in tape, and pulling a vaccum on the circuit (that goes on while electrics are connected on the internal unit)

Of course, usually qualified technicians ("psyktiki" - coolant system tecnicians, catch all trade for all things that deal with phase change temperature regulation systems) do the installation most of the time, but the ease of it means that many electricians do it too for residential units.

 

[Rja4000]

Something that doesn't jump out of the measurements and is a key reason I listen is power handling.

Hi Amir
Thanks for this review again.
About your comment above:
Is there a way to measure that somehow ?
This is so important for listening experience.

 

[pavuol]

So, as long as you don't plan to listen to it above the mid-80's it would be OK. But that's some very high distortion, all things considered.

ME: le me google:
"mid-80s music playlist recommendation, distortionless"