• 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!

Technical Review of the NJ&FX AUDIO PW-6 “HiFi Audio Power Switcher” (A/B switch)

KEFCarver

Active Member
Joined
Mar 29, 2021
Messages
114
Likes
132
Location
Tucson, AZ
This is a review of the NJ&FXAUDIO PW-6 “HiFi Audio Power Switcher”. I had been looking for an A/B speaker/amplifier switcher and really liked the Douk VU3 “Little Bear”, which I purchased. I was happy with until I tested it and discovered that it had a +1dB boost in the frequency response at 20hz, and while it was the same with any input, and in theory you would probably not hear a difference, it bothered me and I returned it. I posted my write-up here: https://www.audiosciencereview.com/forum/index.php?attachments/douk-audio-switcher-pdf.220808/.

I later found out that Amir had already done a review of the Douk VU3: https://www.audiosciencereview.com/forum/index.php?threads/douk-vu3-review-vu-meters.26827/.

I did not have the budget to spend $1400 on the Van Alstine a/b/x comparator, and eventually came across the PW-6:

1675806318368.png


1675806339390.png


I purchased it on Ebay for about $70 (at the time): https://www.ebay.com/itm/263331053496 .It appears to be part of the Douk Audio family. It comes with a small remote and a 12vdc wall wart. The PW-6 weighs a bit more than a pound and is constructed with what I am guessing is aluminum for the housing- it has a nice solid feel to it. It is approximately 7”x5”x1.5”. The connectors are standard banana jack/wire/spade lugs, with 0.75” spacings. There is not a lot of room between them so using really large spade lugs may be a problem; however, all my testing was done using banana jacks and I had no problems. I used the banana to rca adapters I had built for my testing of the Douk VU3:

1675806384834.png


I used a QuantAsylum QA402 Audio Analyzer for my measurements, and used a 2vrms (6dBV) signal level for my testing. The RCA test cables are 4’ long, and the frequency span for all measurements was 20hz-20khz. The first measurement was to check the frequency response. The Right channel (in red) has the QA402 input and output cables connected together with an rca barrel adapter, so it is bypassing the PW-6. There is very little difference between the two, though you can see just a small amount of 60hz on the Left channel (yellow) which is going from A Left to B Left:

1675806425523.png


For the next plot, the Left channel is bypassing the PW-6 and the Right channel is going through the PW-6 from A Right to B Right. It looks mostly the same except now the Right channel has a little 60hz glitch:


1675806499418.png


Things are looking pretty encouraging so far. The next thing to look at was the THD/SNR. Here the Left Channel is bypassing the PW-6 and the Right channel is going through the PW-6 from A Right to B Right.

1675806537567.png


While the Right Channel is noisier than the Left, it is still pretty insignificant- at least for what I would be A/B testing, which are typically older amplifiers and speakers.

I then decided to look at the Crosstalk (isolation) with the signal path A-B selected and leakage measured at path C:

1675806593716.png


Not very good isolation to say the least due to the 60hz leakage. Up to this point I had been using the 12Vdc wall wart that came with the unit. I decided to try to power the PW-6 via the microusb input using a powered usb hub that I have, and that cleaned things up considerably. Here is the Crosstalk (isolation) with the signal path A-B selected and leakage re-measured at path C:

1675806624377.png


The takeaway is that the supplied 12Vdc wall wart is noisy and should not be used.

I then re-measured the THD/SNR for the A-B path with the Right Channel bypassing the PW-6 and got these results:

1675806651056.png


The low level noise cleaned up considerably with the powered USB hub adapter, but there was not really any difference going through the PW-6. Pretty much the same for the path going A-C:

1675806679135.png


Next the frequency response from path A-C was measured and looks pretty much like the path A-B, but without the very small 60hz glitch:

1675806699239.png


I also measured the crosstalk/isolation between the Left to Right channels for the A-B path. In this case the signal was applied to the Left input with the Right input terminated:

1675806724041.png


Results were similar with the A-C path. For my uses, I will power the PW6 using a USB battery pack, which gave very similar results to the powered usb hub that I used for the testing. From the manufacturer’s website, the inside of the PW6 looks like this:
1675806758491.png

Looks like a pretty simple design and pretty much just switching relay paths in and out, which accounts for the good performance as far as THD and Frequency response (IMHO). Since the relays are spec’d at 5A, you could deliver 100w/4Ω or 200w/8Ω, which would be enough for most of the A/B testing/switching one would want to do (at least for me).

All that was left to do was to try the PW6 out. For my 1st test, I hooked up a pair of Klipsch KG2 speakers to a pair of integrated amplifiers: a vintage NAD 3130 and the Aiyima A07. Both were fed with the same (split) signal from a Surfans F20 music player. Each amplifier’s gain was adjusted to give the same SPL from the KG2’s using REW to generate pink/white noise and measuring the SPL with a calibrated microphone. I started off with a 60dB SPL and heard no difference when switching between the amps (not really my intent, just a side note). All that told me I had switched amps was the sound of a relay clicking and the appropriate selection in the display. I recalibrated for a 70dB SPL, and still heard no difference or nuances when switching- I used several different types of music. The relays are spec’d for a 5ms switching time, but I was not able to notice that I had switched amps other than the sound of the relay clicking, which probably would not be heard if you were doing the test at a louder level. The display is very easy to read and the remote works well (only the ON/OFF,1 & 2 keys are functional). Granted the KG2’s may not be the discerning speakers, but they did allow for some listening tests.

I tried another test as I had a pair of “vintage, high end” monoblock amps as well as a “vintage” small footprint stereo amp (all are solid state). One of the vintage monoblocks sold for 10x the cost of the stereo amp. An RC11 RCA volume control (reviewed by Amir), connected between one of the preamps outputs and the Stereo amp, was used to match the levels as it had more gain. Two sets of speakers were used, as well as a variety of listening material. I was unable to hear any difference between the sounds of the amplifiers. At little louder listening levels, when you could not hear the switcher’s relay click, and you would not have known that you switched amps- probably not a surprise to ASR readers. I dumped as much as about 100w/4ohms into the speakers and the PW-6 did fine. Maybe younger ears would hear some small nuances…..

The next “test” for the PW-6 was to use it to A/B two (single) speakers, that were placed next to each other. A power amp that had gain controls to was used to balance each speaker using pink random noise with a calibrated microphone and REW software. The Right Channel was sent to the “C” output/speaker (a Wilson Watt/Puppy) and the Left Channel was sent to the” B” output/speaker (a KEF 107). Though this test could not compare imaging, it definitely was useful to determine the tonal differences between the two and to a certain extent, the soundstage deltas. The switching of the relay(s) is typically not heard if the volume of the music at a certain level.

Based on what I measured and heard, I think the NJ&FXAUDIO PW-6 would work well for some amplifier/speaker switching and basic A/B testing, particularly at its <$100 price point. It can certainly be used as a “truth detector” of sorts and allow someone to evaluate the sound between one amp/preamp/speaker vs. another!
 

ethernode

Member
Joined
Dec 27, 2022
Messages
55
Likes
14
Nice find about the USB, thanks. I'm currently using it exactly for these purposes and the device works really well although I did not measure it. However the LCD died on me after a few weeks of use (I keep it plugged all the time and bought it in December). The switching still works though.
 
OP
KEFCarver

KEFCarver

Active Member
Joined
Mar 29, 2021
Messages
114
Likes
132
Location
Tucson, AZ
Sorry to hear about your LED going out- Mine is only used every now and then. The USB battery pack I have is ideal for an afternoon of A/B testing. The noisy adapter that is provided may be just fine for some uses, but not for what I use it for. Not having the display working would drive me crazy, but at least it switches :)
 

ethernode

Member
Joined
Dec 27, 2022
Messages
55
Likes
14
Honestly there is still the little LEDs that indicate which set of in/outs is active. As I use this for blind a/b testing mostly that's not a big deal to me.
 
OP
KEFCarver

KEFCarver

Active Member
Joined
Mar 29, 2021
Messages
114
Likes
132
Location
Tucson, AZ
That is good. For what I have used it for so far, you really need to know whether you are on "B" or "C" and have not noticed if you can see the LEDs during the day from about 10'. Will pay more attention during my next test...
 

ethernode

Member
Joined
Dec 27, 2022
Messages
55
Likes
14
Nice find about the USB, thanks. I'm currently using it exactly for these purposes and the device works really well although I did not measure it. However the LCD died on me after a few weeks of use (I keep it plugged all the time and bought it in December). The switching still works though.
Little follow up, it turns out the LCD is not dead, but the remote control B6 button apparently allows to configure the LCD brightness between MAX, MINI and OFF positions... :facepalm:
 
OP
KEFCarver

KEFCarver

Active Member
Joined
Mar 29, 2021
Messages
114
Likes
132
Location
Tucson, AZ
Funny. I think I remember reading about that but forgot about it until you mentioned it. Good to know. My little group here is planning on doing an a/b test with it with some $8000 speaker cables- ones with arrow to tell you which way to hook them up :oops: -vs some basic 12 gauge cables. That will be interesting....
 

ethernode

Member
Joined
Dec 27, 2022
Messages
55
Likes
14
Just displays the current selected output
 

lessemm

New Member
Joined
Jul 17, 2023
Messages
4
Likes
2
I'm curious if anyone else who has one of these has had an issue with the remote function changing.

When I got this, it was set so that A-1 on the remote set the Input to A-B and A-2 set it to A-C.

Apparently holding down any of the remote buttons for 2+ seconds sets that button to a "Select" function, which toggles between A-B and A-C.

I did this accidentally while trying to get the remote function programmed into my Logitech Harmony base station. Now there is no way to select A-B or A-C. All I can do is toggle between them.

As anyone with a universal remote knows, having discrete input selection is far better than input select. I haven't found any way to set the device back to the original programmed state.

I've tried to let the device "reset" by leaving it unplugged for an extended period. I've opened it up and looked for a reset button. I found a jumper labeled "RST", but shorting that to "GND" didn't do anything.

I've written to the manufacturer, but I'm not expecting a response. Anyone else run into this? Have a fix for it?
 
OP
KEFCarver

KEFCarver

Active Member
Joined
Mar 29, 2021
Messages
114
Likes
132
Location
Tucson, AZ
Thanks sharing your predicament- I am sorry that your are just able to toggle- I did not know that was possible. Actually, I do not see that as a bad outcome- at least how I use mine, which is predominantly for A/B testing. It would simplify it a bit not having to "jump" between buttons. If you do get a reply from them, I hope you will share it.
 

lessemm

New Member
Joined
Jul 17, 2023
Messages
4
Likes
2
Thanks sharing your predicament- I am sorry that your are just able to toggle- I did not know that was possible. Actually, I do not see that as a bad outcome- at least how I use mine, which is predominantly for A/B testing. It would simplify it a bit not having to "jump" between buttons. If you do get a reply from them, I hope you will share it.
At the risk of losing your discrete selection functionality, you could try "programming" one of the other buttons -- like B-1 -- to the toggle feature. My guess is it will set B-1 to toggle and remove all functionality from A-1 and A-2, but maybe you'll get lucky and get all three functions.

To "program" I just hold down the target button while pointing at the switch. After a couple seconds the Power light will blink twice indicating the program state has changed.

Another thing that might be annoying for you: Once you do this, the display only shows the selected channel for a few seconds. Then it jumps to show the programmed remote button. For example, in its current state, remote button A-1 toggles my inputs. I hit A-1, the device shows A-B, then after about 3 seconds it shows A-1. Hit A-1 again, it shows A-C for a few seconds, then A-1 again. A bit annoying. At this point, I just turned off the display (remote button B-6) and rely on the LED.

It's a shame this thing is so fiddley with the remote. The default functionality of discrete input buttons and constant input display would work great for my setup.
 
OP
KEFCarver

KEFCarver

Active Member
Joined
Mar 29, 2021
Messages
114
Likes
132
Location
Tucson, AZ
I think I will just keep it the way it came. I put some small label below the buttons to correspond to B & C to make it easier to follow. I don't need to switch sooner than the time it takes me to move my thumb from one to another. I had an A/B listening test with a group of 8 friends a few weeks ago and no-one accidently reprogrammed the remote, so that is good. I would probably not get lucky if I tried to program another button...
 

solderdude

Grand Contributor
Joined
Jul 21, 2018
Messages
15,891
Likes
35,912
Location
The Neitherlands
Of course you did not really measure the box but just measured the QA limits + converter cables.

The actual distortion will be load dependent and the frequency response will be dead flat from DC to 1MHz or even higher.
Just like the Douk VU3 b.t.w. it won't change the frequency response.
These are just boxes with 4 additional relay contacts in series with the speaker cable and 4 additional binding posts.
The VU3, however, also provides a very small non-linear extra load (the meter circuit) which is measurable (with an AP) but not audible.
The reviewed speaker selector does not have that but instead has a display but that is not interfering with the audio path.
 
  • Like
Reactions: pma
OP
KEFCarver

KEFCarver

Active Member
Joined
Mar 29, 2021
Messages
114
Likes
132
Location
Tucson, AZ
Of course you did not really measure the box but just measured the QA limits + converter cables.

The actual distortion will be load dependent and the frequency response will be dead flat from DC to 1MHz or even higher.
Just like the Douk VU3 b.t.w. it won't change the frequency response.
These are just boxes with 4 additional relay contacts in series with the speaker cable and 4 additional binding posts.
The VU3, however, also provides a very small non-linear extra load (the meter circuit) which is measurable (with an AP) but not audible.
The reviewed speaker selector does not have that but instead has a display but that is not interfering with the audio path.
The Douk VU3 did alter the frequency response by about 1dB at 20hz if my memory serves me correctly, and that was caused by their VU ckt from what I gather. For the most part I would agree that what I measured was the limits of the QA analyzer and cables- the NJ&FX switch is pretty much just relays in the signal path as you mentioned. It does not have any "large" resistors across the outputs either. I am very pleased with the NJ&FX so far. It is very "eye opening" for A/B testing.
 

solderdude

Grand Contributor
Joined
Jul 21, 2018
Messages
15,891
Likes
35,912
Location
The Neitherlands
The Douk VU3 did alter the frequency response by about 1dB at 20hz if my memory serves me correctly, and that was caused by their VU ckt from what I gather

Impossible, they are just relay contacts.
Amir never measured the frequency response b.t.w. as this would be pointless. These are just relay contacts.
The load of the VU is too small to cause such a difference anyway.
 
  • Like
Reactions: pma
Top Bottom