dbx 223xs Crossover Review

[amirm]

This is a review and detailed measurements of the dbx 223xs active analog balanced crossover. It is on kind loan from a member and costs US $245.

I must say this is a nice and sturdy package for such a bargain price. Controls feel nice as well. Functionality though is confusing given the capability for mono vs stereo. Former provides 3-way crossover whereas the latter 2-way. There is some attempt at clarifying this with duplicate writings but it was hard. Same is reflected in the back:

If you have more patience than I , I am sure you can figure it out though.

Note: Harman owns dbx. Our company, Madrona Digital, buys some of their product lines but not these. Still, feel free to read any bias in my subjective remarks.

dbx 223xs Measurements
My focus here is level of transparency rather than quantifying filters. So I chose to use high-pass output for testing and set the filter very low so that it wouldn't impact most of the response. Here is our dashboard with input gains adjusted to get unity (which oddly was below 0):

This is very competent. Distortion at nearly -108 dB is very close to transparency. Our SINAD which is sum of noise+distortion drops down a bit due to those power supply spikes:

For 16 bit playback, you still have 6 dB of margin which is good.

Here is the frequency response as set up:

You can see the effect of the 40 Hz high pass. Response is extended past 20 kHz which is good.

Channel separation is well above average:

Baseline noise is rather high as displayed in IMD vs level measurement:

The high-pass filter is impacting this a bit.

We see a typical rise in distortion with frequency although it is mild:

Conclusions
Somehow I expected to see high levels of noise and distortion but did not find much. No, this is not state of the art but clears the bar of 16 bit performance well. If you are using ultra sensitive tweeters and such, you may hear some added hiss but otherwise it should be fine.

While a digital crossover is more perfect, you don't get the instant variability to adjust things as you have here.

I am going to recommend the dbx 223xs. It is a bargain given its build and performance.

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

Any donations are much appreciated using: https://www.audiosciencereview.com/forum/index.php?threads/how-to-support-audio-science-review.8150/

 

[DanielT]

Thanks Amir for the review and measurements.

That was a good result. It costs $170 in EU (in Sweden anyway). Perfect to plug in if you have subwoofers but lack the HP-LP fiter.

 

[amirm]

FYI there is a forum bug where I can't create a poll in this new forum.

 

[Lambda]

For the price and the class of product this is better then expected.

 

[Haskil]

Merci Amir !

 

[DanielT]

Out of pure curiosity, how are active analog crossovers constructed? Is the crossover part done via OP amps? Or maybe there are different technical, electronic solutions?

Edit:
I checked Wikipedia:

"An active crossover contains active components in its filters, such as transistors and operational amplifiers.[1][2][7] In recent years, the most commonly used active device is an operational amplifier."

Audio crossover - Wikipedia

en.wikipedia.org

 

[NYfan2]

We use quite some DBX products (also crossovers) because DBX has the name that it produces products that have a good price/quality value.
Good to see that measurements confirm this.

Thanks for testing Amir!

 

[Sokel]

Out of pure curiosity, how are active analog crossovers constructed? Is the crossover part done via OP amps? Or maybe there are different technical, electronic solutions?

Some DSP some only analog.I have used 4or 5 of them the DSP ones are easy to work with the analog ones are even easier but it's a little tricky if you want to time align.
The analog ones use opamps but I have seen discrete and also tube ones!
Most of the op amps have an application like this in their PDF.

That's the LME49720.

 

[Lambda]

Most of the op amps have an application like this in their PDF.

This is nothing specific to this model of the opamp.
It’t like the maul from a blender containing Smoothie recipes.

https://www.electronics-tutorials.ws/filter/second-order-filters.html

The more or less interesting question is how is it made frequency adjustable and.
With Slope / quality factor being more or less constant
Left and right channel benign in sync
How good is low cut and high cut x over frequency matching.

Probably everything is made with multi gang potentiometers and "gyrators"
so i would expect a lot of variation and and channel mismatch over the adjustment range.

State Variable Electronic Crossover

ESP Audio Projects - Continuously variable electronic crossover using state variable filters.

sound-au.com

The better way to do it would be with some sort of VCAs and not with multi gang potentiometers.

 

[LTig]

Out of pure curiosity, how are active analog crossovers constructed? Is the crossover part done via OP amps? Or maybe there are different technical, electronic solutions?

Have a look at the Rane legacy homepage, there are schematics for analogue crossovers (right at the top: AC22).

 

[DanielT]

This is nothing specific to this model of the opamp.
It’t like the maul from a blender containing Smoothie recipes.

https://www.electronics-tutorials.ws/filter/second-order-filters.html

The more or less interesting question is how is it made frequency adjustable and.
With Slope / quality factor being more or less constant
Left and right channel benign in sync
How good is low cut and high cut x over frequency matching.

Probably everything is made with multi gang potentiometers and "gyrators"
so i would expect a lot of variation and and channel mismatch over the adjustment range.

State Variable Electronic Crossover

ESP Audio Projects - Continuously variable electronic crossover using state variable filters.

sound-au.com

The better way to do it would be with some sort of VCAs and not with multi gang potentiometers.

Good questions, clearly relevant!

If I were to use it as an HP-LP filter between subwoofer and speakers, I would in any case want to have at least a 24dB slope on the filter.

 

[Tatteredmidnight]

I’ve been using a dbx 223xl to crossover my monitors and sub in my living room and it has been a great addition for the price. I was able to pick it up used from GC for $80. Very nice to see the xs model perform so well, gives me more confidence in my choice (my understanding is that the units have very similar performance).

Edit: just compared the manufacturer specs and they are actually slightly better for the older xl model. Not sure if those would result in it measuring slightly better, but an interesting choice on dbx’s part.

 

[Sokel]

Probably everything is made with multi gang potentiometers and "gyrators"
so i would expect a lot of variation and and channel mismatch over the adjustment range.

State Variable Electronic Crossover

ESP Audio Projects - Continuously variable electronic crossover using state variable filters.

sound-au.com

The better way to do it would be with some sort of VCAs and not with multi gang potentiometers

Even if Amir's sample was ok you're right,speccially the pro ones (that's the sub-forum we are after all).
That's why I use multiturn trimmers (usually 22 turns),you can adjust it down to last millivolt until you're ok so you take the measurement of the pot get rid of those too and you put a simple resistor.

 

[Koeitje]

Nice to see a review of a DBX crossover, I've looked into using one of these back when my system was a bit different but didn't want to take the risk. Turns out it would have been fine . Curious to see how it stacks up versus a digital crossover, because that needs AD/DA conversion unless you are feeding it a digital signal. What are the odds that such a thing will be better than this in terms of distortion?

 

[anmpr1]

Anecdote: my unit arrived with defective switches (low cut was permanently on). At the price point you don't expect perfection, however the overall build quality seemed a cut above what you would expect for the dollar. I had occasion to contact dbx technical support via email, and received a reply within 24 hours, answering my question.

Dealing with Harman return (by way of email) was not satisfactory. But after three weeks of back and forth I finally received an email shipping label for return.

For these sorts of items, my advice is to purchase from your local guitar store, and not direct from dbx (or any manufacturer). Guitar Center has many local stores, making return easy. Sweetwater (and the others) offer no hassle free returns. I suspect that retailers are more set up for returns/exchanges than manufacturers.

FWIW, the past quite a few years I've used a dbx 215 EQ. It has similar build quality to the 223 crossover, and I've experienced no problems with it. If there is a downside, it is that the 1U form factor doesn't give you much 'wiggle room' for the sliders, making fine adjustments (especially at +/- 12dB) less exacting than you might want it. Their newer 1231 dual band 1/3 octave EQ avoids this limitation altogether, and would be a better choice (but is more money--still not 'high end' expensive). Of course rotary knobs on this crossover avoid this problem. It was easy and intuitive to dial in crossover settings.

Lastly, both EQ and crossover feature switchable 40Hz low cut. I wonder about that choice, and am not sure why dbx chose that value. With an analog system you can use it switched in, as a rumble filter for your record player.

 

[DonH56]

Wow, a review of a component I have used for decades, and it actually has decent performance! Thanks @amirm! This has been my "go-to" analog crossover for many, many years.

The crossover circuit is a standard Linkwitz-Riley (L-R), don't remember what opamps are inside or anything else about the circuit. The L-R topology is a cascade of two second-order Butterworth stages to provide phase and amplitude matching at the crossover frequency with 24 dB/octave (fourth-order) roll offs.

For many years "pro" subwoofers were crossed at 40 Hz, not sure if that is more than an industry standard, but it has been that way for decades (at least IME).

 

[mdsimon2]

There are a few comments in the review suggesting that the two -120 dB power supply spikes are the cause of the poor SNR, SINAD and IMD measurements but I am not seeing how two -120 dB spikes (only present in one channel) result in 102 dB SNR in both channels. Rather it seems that this thing has a relatively high noise floor across the entire audio band.

Michael

 

[DWPress]

Surprisingly good indeed. I'd love to see how it affects the signal full spectrum though. Since your testing in mono anyway is it possible to set up a 3-way XO on the unit and see how it does with multiple slopes? Sounds like the other major limitation is being stuck with LR4 as the only option.

Just curious, I do it digitally but once upon a time I looked seriously at these units.

 

[DonR]

In the absence of a poll... Happy Panther... it looks to be a fine piece of kit for the mid-range performance.

 

[Ra1zel]

My conclusion is:

Thank God for digital XO

Edit: don't get me wrong I would rate this as great, but if I already have sota amps and dac I wouldn't want to spoil my perfectionism fun.