Yes, but there are plenty of people out there with no inkling whatsoever, who fall for marketing of devices like this, who would be better off spending $2K on something else. So IMO it's worth demonstrating repeatedly for those folks.
-
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!
DJM ACTIV Audio Ethernet EMI Filter Review (Bonus!)
- Thread starter amirm
- Start date
Hi Amir, I have to think that you didn't enter the panther piggy bank just because it was sent to you by the company, because this almost 2000 year old tool literally has no effect.
KSTR
Major Contributor
You might want to add a low mains leakage supply for the downstream side but other than that it doesn't get any better with regard to cost-effective common-mode filtering (actually true galvanic isolation in this case).Two TP Link MC200CM's ($25 each) plus a 62.5u SC/SC OM1 Fibre Lead ($10)... job done.
The MC200CM's come with integrated GBIC's (no need to bother with media convertors that use insertable ones)... really plug and play.
------:-----
To judge the effectiveness of a filter or isolation device we have to keep in mind that any EMC issue degrading audio performance needs both an aggressor and a victim. The aggressor is clear: high levels of common-mode RFI that would easily jump over the usual ethernet signal transformers.
The victim is much less clear, basically that could only be some analog device (DAC, amp, ....) downstream that is prone to RF demodulation artifacts. Only if the RF levels were brutally strong we would ever encounter errors on the digital side.
Amir does not test devices for RF demodulation artifacts as that is way beyond the scope of his gear and setup (it would require a state-of-the-art EMC lab), therefore it was completely unlikely from the start that he would find something with regular audio-domain tests as any competent gear is not prone to RF demodulation.
I voted fine because I'm 100% convinced the device performs as advertised, DJM are a serious tech company and professional supplier. It is a bit costly, thus no golfing panther.
Question:
I learned that in the digital world, stuff works perfetcly or doesn't work at all (considering correct setup).
So in times of TCP, what would be the worst case impact of EMI on ethernet? What would be the noticable impact? Huge bandwith reducation? Ping spikes?
Even if there is a heavy EMI impact on the cables, do you notice it at the endpoints?
I learned that in the digital world, stuff works perfetcly or doesn't work at all (considering correct setup).
So in times of TCP, what would be the worst case impact of EMI on ethernet? What would be the noticable impact? Huge bandwith reducation? Ping spikes?
Even if there is a heavy EMI impact on the cables, do you notice it at the endpoints?
Involved in hundreds of ethernet installations in all kinds of enviroments and full 300' runs of copper and never had an interference problem.
And if there is a problem, using shielded CAT cable is usually all thats necessary and if that doesn't do it, use fiber, completely immune.
And at a fraction of the cost of that filter.
And if there is a problem, using shielded CAT cable is usually all thats necessary and if that doesn't do it, use fiber, completely immune.
And at a fraction of the cost of that filter.
Last edited:
OP
- Thread Starter
- #86
It us a tough call in that there totally nonsense tweaks out there which do nothing. This box does what it says.
KSTR
Major Contributor
Basically it's a go / no-go scenario just as you say. At some point very strong EMI will break a connection because of too much low-level bit errors. Very very unlikely to happen in practice, I'd guess.
But again, the point here is not Ethernet data integrity, it is to protect a whole (audio) system from EMI ingress which could propagate along all kinds of cables (sans optical, of course) and may cause RF demodulation noise in analog (audio) circuits.
We all know or at least remember the rhythmic noise pattern of cell phone communication when an active cell phone is placed on gear or near a cable in an audio system, let's say, on the network cable from router to streamer... which is a classic example of RF demodulation. This filter would quite certainly stop it when inserted in the network cable connection which was subjected to the cell phone signal (on the router side).
As noted, you can have the same effective function for a fraction of the price with an optical bridge.
However, you *can* safely place your cell phone on an Ethernet cable and/or router/switch without creating EMI issues that interfere with data integrity. It's the analog part that may be impacted by that - not really the digital transmission path, which is pretty darn robust (don't forget very large data centers are not very clean environments, there are very long runs of copper ethernet at very high rates - copper is noticeably cheaper than optics in that environment). Audio is a very low bandwidth application these days. The amount of EMI you'd need to corrupt audio data on Ethernet would make you quickly forget about audio dropouts... :-D
keith_h
Member
- Joined
- Mar 23, 2021
- Messages
- 89
- Likes
- 92
Interesting review. Amir has said it does what the company claims it does at a cost. Others have identified more cost effective solutions to the same problem, noting its an extremely rare possibly non existent use case. I wonder if they actually sell any?
It can vary dramatically - you can have connections where bandwidth drops off to almost nothing due to constant retries... then drops off, then works again.... and at other times it works perfectly.
Typically the issues relate to particular equipment that emits loads of RF interference, but only when it is in use.... hence the variability of the connection for nearby ethernet links...
Often these can be "fixed" through the proper use of shielded cables and terminations... but not always
When TCP IP fails it is for traffic congestion or equipment failure in basically 100% of cases. Ethernet and TCP IP have a lot of resilience built into them.
Until you run an unshielded cable past a metal workshop using arc welders.... and then all your bandwidth gets used up in retries to get through the noise.... (I'm just finalising a build that includes such a setup, running shielded cables fixes most of the issue...)
OK... 99.999% :-D
OP
- Thread Starter
- #94
In the context of home LANs, that is very much true. But go on a weak cellular connection and you will absolutely experience dropped connections, etc. If the channel is noisy, all bets are off.
solderdude
Grand Contributor
As KSTR already mentioned. It is not so much about the data being corrupted but about leakage and RF (common mode signals opposite ground) that can enter into the audio path and become audible in several ways (mostly detection by some of the internal semiconductors).
This filter (as well as the suggested cheap multibox solutions which all could induce garbage from used SMPS) is quite likely to work wonders in that case.
Audiophiles are scared of audio gremlins and don't mind spending a lot of money on devices that prevent gremlins without affecting the incoming data.
The audio measurements Amir did showed there was no impact on the audio. Granted the cheap solutions would have given the exact same results as no 'garbage' was introduced in common mode to the incoming side.
For about 99.9% of audiophile cases it won't do anything but set the mind of these audiophiles at ease that it works and audio gremlins entering via the ethernet cable can't enter their system.
Most likely the same audiophiles will connect this device with ethernet cables that are even more expensive than this filter.
It does not. It doesn't do anything in the analog audio path, which is where noise matters.
100dB insertion loss that includes Ethernet frequency range?
I'm going to assume they are referring to common-mode isolation. In that case yes that's a thing.
If you really want to test this thing doing the job it's supposed to,
Baseline: Take an RJ45 cable. Select a twisted pair of wires. Send a single-ended signal on both wires (copy). Return ground goes to a separate cable. Use audio precision to measure this loopback. Adapters needed to convert from RJ45 to RCA etc.
Device under test: Same RJ45 but now there's a DJI ACTIV in the loop.
The signal amplitude should drop by 100dB
I'm going to assume they are referring to common-mode isolation. In that case yes that's a thing.
If you really want to test this thing doing the job it's supposed to,
Baseline: Take an RJ45 cable. Select a twisted pair of wires. Send a single-ended signal on both wires (copy). Return ground goes to a separate cable. Use audio precision to measure this loopback. Adapters needed to convert from RJ45 to RCA etc.
Device under test: Same RJ45 but now there's a DJI ACTIV in the loop.
The signal amplitude should drop by 100dB
solderdude
Grand Contributor
The device is also galvanically separating and depending on how the ethernet receiver cable is grounded internally and the audio path is grounded internally there COULD be a direct path (or even indirect path) to the audio. There should not be but as this is audio not all devices (certainly the audiophool ones) are designed properly.
Would I buy, use, or even recommend this device ? NEVER. There are cheaper solutions in the rare case an actual solution is needed.
Amir's test does not test for common mode issues and crap on the incoming Ethernet cable (which can be there, and could, in some devices, enter the common (audio) ground.
The only thing the test shows (just like any other device in the ethernet cable path) is that there is no influence nor benefit if everything works 'as expected'.
So for the vast majority of cases all the device does is cost money and offer no benefits. That does not mean that in some cases there could not be a benefit. And when that case is there it can be solved equally well for a fraction of the price.
Last edited:
KSTR
Major Contributor
For me it's clear that one main target group for this filter (and the other filters they offer, USB and HDMI) are EMC (pre-)test engineers setting up a small EMC test cabinet that needs several clean digital data pathways, besides clean power etc., for testing many kinds of consumer gear.
And if they can sell a few to audiophiles striving for peace of mind, why not?
This "audio" filter is the only one they quote a price for on their website. I believe they are testing new markets for their technology. I am sure they would tell you if you need it or not, would you dare to ask. I am also sure they would be able to provide measurements on what the device accomplishes. This is not the case with snakeoil companies.
Specs should be close to this: https://www.djmelectronics.com/docs/ACTIV 10G Data Sheet.pdf
Specs should be close to this: https://www.djmelectronics.com/docs/ACTIV 10G Data Sheet.pdf
Similar threads
- Poll
