UpTone Audio EtherREGEN Switch Review

[ajawamnet]

As to ethernet cabling:

This is what my drummer uses to certify cable plants -

https://www.amazon.com/Fluke-Networks-DSX-5000-120-Permanent/dp/B00DGHAOM8
about $13,000.

Here's some screenshots of him doing some tests for our company using an earlier version - the DTX-1800 with the options for twisted pair. Interesting in that - just like a vector network analyzer - the cable from the DTX to the 8P8C is actually individual waveguide/coax since you null the reference plane.

Note - a lot of people ref to it as RJ45 tho that's actually an FCC term for a registered jack. Also note that this was for 100 so the custom only had two pair.

- a best buy cat 5 (TEST REF) and a "custom" cat 3:

Best Buy 25':

Custom cable 25':

 

[amirm]

SOtM makes similar devices and are probably more popular, especially worldwide. It would be interesting to review some of theirs. At least then you couldn't be accused of being on a vendetta against Uptone.

Satisfying their paranoia and anxiety is not worth me spending money on such products. In time though, I am sure I will get more products to test.

 

[amirm]

FYI, the EtherRegen is packed and I will be shipping it shortly.

I also heard from the owner that it has the "old" firmware.

 

[g29]

FYI, the EtherRegen is packed and I will be shipping it shortly.

I also heard from the owner that it has the "old" firmware.

@amirm , So you aren't going to do a bandwidth test to see if it is a descent switch or not ?

 

[Soniclife]

@amirm , So you aren't going to do a bandwidth test to see if it is a descent switch or not ?

He does not need to do that, any owner with the skills can do it, no AP required. Plus he has already stated his network works perfectly, like mine and many others, so if could only be the same or worse, you cannot improve perfect.

 

[g29]

He does not need to do that, any owner with the skills can do it, no AP required. Plus he has already stated his network works perfectly, like mine and many others, so if could only be the same or worse, you cannot improve perfect.

LOL, you forget the cost to make that test is over $600 for an unknown commodity. No thank you. The review would be more complete if it included validation of the switch's basic, primary function, a baseline throughput functionality benchmark. ASR measures and documents, why omit the primary measurement ?

Based on your reasoning, an AP is not required to do a listening test either, but it is included in the review none the less.

 

[phoenixdogfan]

"Built on a costly six layer circuit board."

Six layer circuit boards are even better than five layer bean dips!

 

[Jinjuku]

LOL, you forget the cost to make that test is over $600 for an unknown commodity. No thank you. The review would be more complete if it included validation of the switch's basic, primary function, a baseline throughput functionality benchmark. ASR measures and documents, why omit the primary measurement ?

Based on your reasoning, an AP is not required to do a listening test either, but it is included in the review none the less.

Throughput, delay, BER, improvements aren't listed on the UpTone website. Improved fidelity on the output of a DAC is and was tested.

 

[g29]

Throughput, delay, BER, improvements aren't listed on the UpTone website. Improved fidelity on the output of a DAC is and was tested.

UpTone also published it improves the sound which apparently it does NOT. They imply (5X) that this is a Gigabit switch on their website. Why would you believe their published specs on anything with their historical track record of misinformation ? Isn't the whole reason for ASR to validate (or debunk) manufacturer's published claims ???

See Also: Uptone Audio ISO Regen Measurement and Review

Update: Main web page implies Gigabit 5X, but their board illustration shows 4 - 1GB RJ45 ports and 1 - 100Mpbs RJ45 port going to the music PC.

 

[amirm]

@amirm , So you aren't going to do a bandwidth test to see if it is a descent switch or not ?

There just wasn't enough time this go around as the owner needs the unit immediately. Maybe in the future we get another chance to run such tests.

 

[ajawamnet]

"Built on a costly six layer circuit board."

Six layer circuit boards are even better than five layer bean dips!

But 7 is even better... actually,most PCB houses prefer even numbers - stack ups of prepreg and cores.

Here's the breakdown of a short run of 50 PCB's on a 1 week spin for the same 1/16" thick PCB same 4 x 4 size (the size of my 9 ports about that - PC104 sized with:

2 layer
4 layer
6 layer

1 oz copper inner/outer.

Not much diff... I design about 100-130 PCB's a year; have to fab and proto build about a third of them.

2 layer is not really applicable, since diff pairs in ethernet really like having a contiguous ground plane. But I've seen it done.

And note this is a domestic supplier (Sunstone - Value Proto service ) . Off shore, there's even less of a difference.

In fact, if you have to add a layer pair in order to have all the parts on one side, its a significant savings (only one pass thru pik and place).

It's obvious that this guy with the " ... 6 layer ...." BS has no clue

 

[ajawamnet]

UpTone also published it improves the sound which apparently it does NOT. They imply (5X) that this is a Gigabit switch on their website. Why would you believe their published specs on anything with their historical track record of misinformation ? Isn't the whole reason for ASR to validate (or debunk) manufacturer's published claims ???

See Also: Uptone Audio ISO Regen Measurement and Review

Amirm can test to see if there's any truth to the recovered audio, hopefully all the link partners (what they call each side of an ethernet connection) are similar.

As to a switch's ethernet performance see my earlier post on how they test that -

https://www.audiosciencereview.com/...io-etherregen-switch-review.10232/post-280694

 

[g29]

There just wasn't enough time this go around as the owner needs the unit immediately. Maybe in the future we get another chance to run such tests.

Thanks, I would really like to know what its real throughput is since they imply that it is a Gigabit switch by mentioning use of Gigabit parts, but omitting the actual throughput.

Maybe the owner can run that test and post the results to this thread.

 

[graz_lag]

Fraud and deception is a dirty business ...

Is it so complicated to understand?
Really?

 

[g29]

...
As to a switch's ethernet performance see my earlier post on how they test that -

https://www.audiosciencereview.com/...io-etherregen-switch-review.10232/post-280694

Sorry, I did not see any throughput numbers for the specific EtherREGEN switch in your previous post. Did you or someone actually benchmark the UpTone switch ?

 

[pozz]

Sorry, I did not see any throughput numbers for the specific EtherREGEN switch in your previous post. Did you or someone actually benchmark the UpTone switch ?

It would be interesting to see, but just as a sidebar to the testing already done.

 

[g29]

From UpTone's illustration, it appears the unit has 4 - Gigabit ports on one side of the moat and 1 - 100Mbps port going to the music PC on the other side of the moat. The 100Mbps port would be a bottleneck for today's ISP speeds.

 

[Wes]

At least it has actual semi-conductor devices in it - not all BS Audio products have in the past

 

[decoRyder]

Looking at that board, I think they've could have done it a lot easier, and cheaper if the purpose was to outright scam people. I still don't understand how something that happens on the physical layer on the switch is supposed to affect anything on the application layer on the 'Music PC', not to mention the physical layer of a connected DAC, and I've never seen a comprehensive explanation of the uptone proprietor on that subject.

 

[ajawamnet]

From UpTone's illustration, it appears the unit has 4 - Gigabit ports on one side of the moat and 1 - 100Mbps port going to the music PC on the other side of the moat. The 100Mbps port would be a bottleneck for today's ISP speeds.

So in that photo, I see no markings on the main switch IC.

Note that it has the PHYS integral - ie. not using MII/RMII GMII for external phys. In fact it looks a lot like a micrel/microchip eth switch - nothing special.
https://www.microchip.com/wwwproducts/en/KSZ9477

Might be a Marvell tho... again just a commodity switch IC.

J6 and J8 are programming headers for a J-Link. Most likely for ARM SWD. Used those on a design here:

I'm not sure why he has a "moat" with a Murata isolated DC-DC... maybe to keep the BNC off of the power gnd? Dunno...

I will say this - if the guy goes as far as to remove IC markings then there's a good chance that this is all BS...