For this post, let's spend some time looking and listening to the above device. It's the Rotel RSS-900 "speaker selector", MSRP around $200. This has been on the market for awhile now, I think released back around 2004. Here's the manual for it. I was graciously lent this box by a friend to check out as this is not a kind of product I've published on in this blog.
As you know, these days, loudspeakers tends to have quite low impedances with 4Ω ratings common and often when we look at impedance curves, it's not unusual to see values down to even below 2Ω for some frequencies (EPDR's typically even further down). Coupled with the fact that if one were to run multiple speakers to the same amplifier output, as per Ohm's law of parallel circuits, we risk extremely low impedance/resistance to the point of basically shorting out the amplifier. We obviously would not want to do this!
So it is then that we should protect the amplifier with devices like this Rotel switch if we need to run multiple speakers or for A/B listening with instantaneous switching between outputs.
We can see from the manual that this device is passive, and provides switching for one "Direct" speaker output and five switched, what I'll call "Indirect" speaker outputs. It is these 5 Indirect outputs that have "impedance compensation" and based on the manual, excess capacitance should also not be an issue.
Here's a look at the box from behind:
![]() |
| This specific unit seems to have been procured around 2013 based on one of the rear stickers. Banana plugs on the far right for input from amplifier. |
As you can see, we have two rows of speaker binding posts - left and right channels each row. It's rated for up to 200W and 4Ω speaker loads. The box is not large, measuring about 17.5" x 3" x 7" and 4.5lbs; sheet metal enclosure with reasonable non-gold plated binding posts that accept banana plugs which is what I used for testing. With bare wire, one probably can achieve better contact.
On the front, we have switches to turn on/off each speaker as arranged, correlating with the Direct and Indirect outputs we saw behind.
Okay, so what does this thing do to the electrical signal? Let's run some tests. First, to get a baseline, since this is about whether the passive device intrudes upon the electrical connection between amplifier and speaker, let's get the impedance, phase, and EPDR curve from the Sony SS-CS5M2 discussed last week with the "clone" Nordost-like 2.5m speaker cable examined last year:
Let's start with the "Direct" path. In principle, as per the name, this should be the cleanest output. Does connecting this Sony speaker + Nordost-like cable to the switch box with a short 3' 12AWG cable (normally leading to an amplifier) make a difference to the impedance/phase/EPDR graph?

The "Direct" pathway from an electrical perspective indeed measures almost in a transparent fashion. The only difference I see is the small 0.46Ω increase in the minimum impedance due to the extra length of wire, connector, internal conductor and switch added to the signal path to 6.3Ω (compared to 5.85Ω without the switch box).
So what about that multi-switched "Indirect" pathway? Let's plug the same speaker+cable into the first Indirect output and measure with all the other switches turned off:
Now we see a more significant difference. Notice that the minimum impedance now measures 8.46Ω. That's a 2.6Ω increase which was added by the switch over the minimal impedance of the speaker+cable and just over 2Ω above the speaker+cable+Rotel's Direct output. This suggests that the speaker switch provides a floor for how low the impedance/resistance can go when connecting multiple speakers to the switch. We can confirm this by shorting Indirect output 2 while the Sony+Nordost stays on Indirect output 1:
It's obviously never good to short the output for an amp! With this speaker box, we can see it kept the minimum impedance in this set-up at a protective ~3.6Ω. Note the flat phase response; the passive box did not shift the time domain in any meaningful way out to 45kHz as plotted.
Finally, just to double check to make sure it doesn't get nasty if we connect the Sony+Nordost to the Direct output but shorted one of the Indirect outputs:
Obviously a single amplifier connected to two outputs (one of which shorted in this case) will always result in lowered total impedance. The Rotel did a good job in keeping the minimum impedance at 5.85Ω.
Although I did not measure the impedance graph, I can confirm that the Direct output is unprotected and if you short that, it'll be essentially 0Ω using an LCR meter.
So how does this sound?
Unfortunately, I didn't take a picture of the listening session when I brought the Sony SS-CS5M2 to my living room and connected the switcher to both the Sony and Tannoy mX2 speakers to A/B listen.
![]() |
| The Sony SS-CS5M2 and Tannoy mX2. |
With the left Sony speaker connected straight to my old Denon 3802 receiver and the right connected to the Rotel switcher's Direct output, it was not easy to hear a difference; maybe just a slight reduction in output level through the Rotel shifting mono music to the left (non-Rotel) side, harder to tell with stereo music. As expected with the higher resistance/impedance through the Rotel Indirect output, the level difference was more noticeable but tonality otherwise sounded fine. I listened to a few audiophile test tracks (the usual stuff like Pink Floyd, some Dire Straits, a little Firebird Suite, etc.). As you might expect, it got boring due to lack of anything interesting/concerning quickly.
With the Sony and Tannoy connected to the switcher, it's easy to do A/B listening and note the difference in tonality between the two bookshelf speakers with the Tannoy sounding less treble extended than the Sony with its super-tweeter but still very pleasant and with its slightly larger woofer (5.7" compared to the Sony's 5.12"), the old Tannoy has a slightly better bass extension and overall more midrange lift. They were both about the same sensitivity +/-1dB with the Tannoy maybe a tad louder.
Obviously it's always cleaner to directly connect your speakers to the amplifier. I'm sure if I ran a 1kHz signal through this box there would be increased distortions. In casual listening though, I didn't notice any distortion or noise whether through the Direct or Indirect outputs during my testing. Something like this used in an audio showroom to allow quick speaker switching would be just fine to show off tonality, relative sensitivity of speakers, and for subjective evaluation.
In summary... The Rotel RSS-900 is a passive speaker switch, it works, and sounds fine as expected.🙂
I don't know if one can still buy this new but I've seen it available used on places like eBay, Audiogon, and the like typically for around USD$100 or so in very good condition. As usual, when buying older products, it would be good to do a little maintenance like cleaning the binding posts to minimize signal loss. I checked my friend's unit and all the Indirect outputs performed about equivalently.
Since I only borrowed this speaker switch, I did not open it up to have a look. However, there is a picture online and here's the Rotel RSS-900 Technical Manual where we can see the basic schematic:
![]() |
| Picture from Reclaimed Audio. Note the two green power resistors. |
There are of course many other speaker switches to consider if you need one but many I see these days on Amazon look pretty crappy with poor binding posts. For cheap, if you only need to select between 2 speakers in a low power system, something like the XtremPro 2-Way (US$36, 60W, 4-8Ω) with impedance matching and volume control probably would be fine. Going higher, the remote-controlled Nobsound Audio Splitter (US$70, rated to 300W, 4-8Ω) seems like a good deal. Even more fancy, the Douk Audio VU3 PRO (US$160, 350W) features VU meters, dual amp and speaker selector (2-in-2-out), plus remote control.
Hope you're enjoying the music as we head towards the Christmas season, audiophiles!
Let's end with some new music:
"Lovebomb" from The Neighbourhood's latest album (((((ultraSOUND))))) (2025, DR11 multichannel, and a horrible DR4 stereo version!, no surprise):











No comments:
Post a Comment