Author Topic: Soap router - HW specs too good to be true?  (Read 97565 times)

0 Members and 1 Guest are viewing this topic.

Offline sync

  • Frequent Contributor
  • **
  • Posts: 799
  • Country: de
Re: Soap router - HW specs too good to be true?
« Reply #25 on: March 13, 2014, 03:13:37 pm »
I think you don't need a quad core to route 1Gb/s. Realtime malware scanning is of course something completely different. You need tons of computing power for it. Also to make it proper you need to decrypt and reencrypt encrypted connections.
I'm guessing the realtime scanning of the SoHo routers are very basic and limited. In case of the Soap router maybe an app which displays a nice animated "you are safe" icon is enough. :-DD
 

Offline sync

  • Frequent Contributor
  • **
  • Posts: 799
  • Country: de
Re: Soap router - HW specs too good to be true?
« Reply #26 on: March 13, 2014, 04:05:51 pm »
Out of curiosity. How fast are these Japanese 1Gb/s SOHO connections in real life?
 

Offline gabhex21

  • Newbie
  • Posts: 2
Re: Soap router - HW specs too good to be true?
« Reply #27 on: March 13, 2014, 04:19:02 pm »
Quote
We have been dealing with this issue and in testing speed we have used a mPCIE to USB3.0 and we have tested SATAII to USB3.0.

They seem to believe that you can take a USB -> SATA DEVICE cable and just swap the connectors.
What they are missing is that they actually need something that behaves as a SATA device and connects to an external USB3.0 mass storage device, not the other way around.
The stuff they keep making up would be funny if it wasn't for the real people who are going to lose money in this scam.
 

Offline dext0rb

  • Contributor
  • Posts: 49
  • Country: us
  • grind it out
Re: Soap router - HW specs too good to be true?
« Reply #28 on: March 13, 2014, 09:48:37 pm »
Quote
We have been dealing with this issue and in testing speed we have used a mPCIE to USB3.0 and we have tested SATAII to USB3.0.

They seem to believe that you can take a USB -> SATA DEVICE cable and just swap the connectors.
What they are missing is that they actually need something that behaves as a SATA device and connects to an external USB3.0 mass storage device, not the other way around.
The stuff they keep making up would be funny if it wasn't for the real people who are going to lose money in this scam.

I pledged them $1 just to ask the USB 3.0 question. Once I saw that answer ...   :palm:

I don't understand how Kickstarter allows their program to remain open when they obviously do not have a functional prototype.

But hey, I could use an extra $100K too...I have a TI MSP430 Launchpad, some STM32F4 Discovery boards, and an Embedded Artists LPC3250 board sitting at my desk. Gimme some money!  :-DD
EE and Tinkerer.
 

Offline marcan

  • Regular Contributor
  • *
  • Posts: 80
  • If it ain't broke I'll fix it anyway.
    • My blog
Re: Soap router - HW specs too good to be true?
« Reply #29 on: March 17, 2014, 06:25:08 pm »
I like their current claimed board design.



Let's count the fail:
- No mounting holes for the board
- SATA male connector, which needs a completely non-standard female-female data+power cable.
- mPCIe socket next to that, hanging off the board. No mounting holes for the card that goes there.
- Two radio modules but no antenna connectors for them
- Another radio module with a single SMA connector - but this is supposed to be a MIMO router
- No Ethernet switch chip - instead they have a PHY for each connector, but nowhere for that PHY to connect to

Nevermind the random rows of passives and general BS smell of the layout.

And their explanation for why they used to have the ripped off Novena screenshots is just gold:

Quote
We originally had a opensource board called Kosagi as our foundation for our final design but we changed this when cost was too high and it had to many features that we didn't need.

We have now updated the project with our new and most recent design below which is not using any opensource source files to build (like a Altium from Kosagi)

"A Altium". Right.


Edit: more lies:



That's this board: http://www.friendlyarm.net/products/tiny4412
Note the absence of SATA or a 12V power rail, and yet they have an HDD hanging off to the side...
« Last Edit: March 18, 2014, 03:01:03 am by marcan »
 

Offline Fsck

  • Super Contributor
  • ***
  • Posts: 1157
  • Country: ca
  • sleep deprived
Re: Soap router - HW specs too good to be true?
« Reply #30 on: March 18, 2014, 09:11:29 am »
I'm sort of curious how they'll draw the traces for those PHYs.
"This is a one line proof...if we start sufficiently far to the left."
 

Offline tom66

  • Super Contributor
  • ***
  • Posts: 6680
  • Country: gb
  • Electronics Hobbyist & FPGA/Embedded Systems EE
Re: Soap router - HW specs too good to be true?
« Reply #31 on: March 19, 2014, 07:14:13 pm »
I mentioned the PCB didn't look right but there's more.

- I see them mostly using large format ceramic caps which are really expensive.
- And they don't have any easily visible clock/xtal for the processor
- If they use the i.MX don't they need the i.MX power/reset controller with the various core supplies?
- Amusing random headers in the middle of the board no idea what they do.
- Insufficient clearance for the SATA header to install a HDD and miniPCI card unless extension cables are used, pricey.
 

Offline dext0rb

  • Contributor
  • Posts: 49
  • Country: us
  • grind it out
Re: Soap router - HW specs too good to be true?
« Reply #32 on: March 20, 2014, 03:47:45 pm »
Suddenly they have added an internal battery, which of course, is nowhere to be seen on the "prototype" PCB...

https://www.kickstarter.com/projects/soaprouter/soap-first-smart-router-w-touch-display-powered-by/comments?cursor=6237539#comment-6237538
Quote
Soap will have a small internal battery for this sort of thing we don't want it forgetting everything with the simple loss of power.

Maybe it will connect to one of the many BS headers.
EE and Tinkerer.
 

Offline amyk

  • Super Contributor
  • ***
  • Posts: 8240
Re: Soap router - HW specs too good to be true?
« Reply #33 on: March 22, 2014, 01:53:52 pm »
Is it that hard to Google "router PCB" and create your fake one based on existing layouts...? At least that would make it a little more convincing.

Quote
Out of curiosity. How fast are these Japanese 1Gb/s SOHO connections in real life?
I was there a few years ago. Under the right conditions, they are really fast: http://i61.tinypic.com/2b7cyh.png
 

Offline marcan

  • Regular Contributor
  • *
  • Posts: 80
  • If it ain't broke I'll fix it anyway.
    • My blog
Re: Soap router - HW specs too good to be true?
« Reply #34 on: March 23, 2014, 12:21:47 am »
- And they don't have any easily visible clock/xtal for the processor
That they do have - I think the small grey squares are supposed to be small SMT xtals. There are two near the CPU and one next to each ethernet PHY.

- Insufficient clearance for the SATA header to install a HDD and miniPCI card unless extension cables are used, pricey.
It's not even the right connector for that - it's male, that needs a cable no matter what. It'd have to be female to plug in directly to a drive. Not that that would work with their claimed form factor anyway. But then again, neither can the miniPCI card.
 

Offline CanadianAvenger

  • Regular Contributor
  • *
  • Posts: 179
Re: Soap router - HW specs too good to be true?
« Reply #35 on: March 23, 2014, 12:23:47 am »
That SATA header as you're calling it is a CFAST connector [SATA version of CompactFlash] The data signals are in the right order, IIRC, but the power part of the connector is mapped differently, with different voltages.
 

Offline tom66

  • Super Contributor
  • ***
  • Posts: 6680
  • Country: gb
  • Electronics Hobbyist & FPGA/Embedded Systems EE
Re: Soap router - HW specs too good to be true?
« Reply #36 on: March 24, 2014, 02:46:14 am »
Well, they got their bux, so now we wait and see if anything happens. :clap:
 

Offline BravoV

  • Super Contributor
  • ***
  • Posts: 7547
  • Country: 00
  • +++ ATH1
Re: Soap router - HW specs too good to be true?
« Reply #37 on: March 24, 2014, 02:52:18 am »
... so now we wait and see if anything happens. :clap:

Hopefully it won't get "MµTI-ed".

PS : "MµTI" stands for Mµ Thermal Imager  :-DD

Offline marcan

  • Regular Contributor
  • *
  • Posts: 80
  • If it ain't broke I'll fix it anyway.
    • My blog
Re: Soap router - HW specs too good to be true?
« Reply #38 on: March 24, 2014, 04:19:13 am »
That SATA header as you're calling it is a CFAST connector [SATA version of CompactFlash] The data signals are in the right order, IIRC, but the power part of the connector is mapped differently, with different voltages.

It's not CFAST. This is CFAST:


Note how the two slots are flat, while the wafers on the 3D render have a lip. It's just a SATA connector, of the kind you'd find on a drive:


 

Offline CanadianAvenger

  • Regular Contributor
  • *
  • Posts: 179
Re: Soap router - HW specs too good to be true?
« Reply #39 on: March 24, 2014, 02:52:24 pm »
Whups you're right... I did not notice the lip/key in the rendering earlier, thanks for correcting me.
« Last Edit: March 24, 2014, 02:55:29 pm by CanadianAvenger »
 

Offline Bud

  • Super Contributor
  • ***
  • Posts: 6877
  • Country: ca
Re: Soap router - HW specs too good to be true?
« Reply #40 on: May 15, 2014, 04:21:09 am »
I enjoyed reading the few last updates

"We have been putting together a large update with a lot more final information. The problem is we don't have the final information yet."



This is hilarious. Make sure you read last couple dozen comments, too. 

https://www.kickstarter.com/projects/soaprouter/soap-first-smart-router-w-touch-display-powered-by/posts

Facebook-free life and Rigol-free shack.
 

Offline madires

  • Super Contributor
  • ***
  • Posts: 7695
  • Country: de
  • A qualified hobbyist ;)
Re: Soap router - HW specs too good to be true?
« Reply #41 on: May 15, 2014, 10:49:34 am »
From Update #32:
Quote
-Premium apps from Soap and subscriptions available and discounted price for KS backers only

Let's call it iRouter  >:D There are already a few SOHO routers with dual core ARM on the market right now and we'll see much more, also dropping prices. As soon as some of them are supported by OpenWRT or DD-WRT Soap doesn't stand any chance. Some could argue that Soap got that nice touch screen but I think it's a silly idea. Either you want to connect ethernet cables and optimize the position of the WLAN antennas for best performance or you want to carry the touch screen around. I'd buy a SOHO router, place it on a shelf, wire up everything and leave it there. And if I need a touch screen I'd get a pad.
 

Offline Kean

  • Supporter
  • ****
  • Posts: 2054
  • Country: au
  • Embedded systems & IT consultant
    • Kean Electronics
Re: Soap router - HW specs too good to be true?
« Reply #42 on: May 15, 2014, 03:26:13 pm »
As is typical, there is almost nothing in the recent updates... except this nugget from the latest update:
Quote
SOM Vendor for CPU (For any hardware guy this was a obvious change that needed to happen for development)

Ummm... huh?  So they've ditched all their previous (faked?) PCB layouts and switched to a System-on-Module?   :-//

Don't know how that addresses any of their actual deliverable issues - multiple gigabit ethernet ports, high network throughput, Android drivers, kitchen-sink wireless standards, etc.
 

Offline tom66

  • Super Contributor
  • ***
  • Posts: 6680
  • Country: gb
  • Electronics Hobbyist & FPGA/Embedded Systems EE
Re: Soap router - HW specs too good to be true?
« Reply #43 on: May 29, 2014, 11:36:59 pm »
May 21st update

https://www.kickstarter.com/projects/soaprouter/soap-first-smart-router-w-touch-display-powered-by/posts

wow... where do we start

...Inability to distinguish between MB/s and Mb/s, and MB and MB per second (not a typo - claims they will overcome "470 Megabyte" limit, "ethernet gigabyte speeds", etc.)

...USB 3.0 x 2 ports requiring 10GBits/s bandwidth, yet the i.MX6 Quad only has a single USB 2.0 host (if they did this it would be via PCI-Express but they only have one lane at 5Gbit/s so it's not possible anyway, unless using both ports halves the transfer rate.)

...Sudden late game switch to Maxwell  vs Quantenna, they were so sure and had prototype hardware allegedly...
 

Offline Rasz

  • Super Contributor
  • ***
  • Posts: 2616
  • Country: 00
    • My random blog.
Re: Soap router - HW specs too good to be true?
« Reply #44 on: May 30, 2014, 03:46:26 am »
Latest update actually looks realistic. Obviously they will use PLX PCIE switch chip for USB 3.0, ethernet and two wifi radios.
Something like PEX 8606

example of how it looks like
http://www.ebay.com/itm/PCI-e-express-1X-to-4port-1X-switch-multiplier-riser-card-for-diy-bitcoin-miner-/221390112653?pt=LH_DefaultDomain_0&hash=item338be1078d

Previous update gave me a mental image of 3 frat boys sitting with some MBA type debating how to best monetize the userbase and synergize back-end convergence (ass fucking said user base). App stores? nickel and dining for every code snippet while hardware still doesnt exist? Sure sounds like they are already pitching this thing to VCs to get that sweet piece of silicon valley startup pie.

Edit: Obviously the plan is to "make it up in volume". i.MX6 module alone is ~$200 in volume, PLX switch ~$15, probably ~$10 for wifi radios, usb ~$5, BT another 5, <1GHz radios $10? 20? sounds like its some off the shelf usb module.

All in all ~$300 in parts alone, and they "sold" them at $150-170. Another solid business plan in action.  Either they screw some sucker investment fund or not deliver at all.
« Last Edit: May 30, 2014, 04:06:45 am by Rasz »
Who logs in to gdm? Not I, said the duck.
My fireplace is on fire, but in all the wrong places.
 

Offline saironTopic starter

  • Contributor
  • Posts: 13
  • Country: cz
Re: Soap router - HW specs too good to be true?
« Reply #45 on: May 31, 2014, 09:49:50 am »
Still sounds bullshitty to me. What are those symmetric groups of components you can see four times on the bottom side and once left from the mPCIe slots? Looks like a crystal oscillator with some diodes and capacitors around, but it doesn't make sense to have this five times randomly spawned around the board. And why is there an unequipped place for THT crystal next to the clock battery? Also, I don't see any connector for the LCD and the layout doesn't correspond to the layout of connectors in their previous 3D render (then why did they do it?).

I have been looking for any 802.11ac mPCIe card with Marvell chipset and haven't found one (after some Google searching nor in this list: https://wikidevi.com/wiki/List_of_802.11ac_Hardware), they plan do design their own?

Yeah, and the component prices still don't add up to their retail price, that's right. Still won't believe until I see anything real that's really working.
 

Offline rob77

  • Super Contributor
  • ***
  • Posts: 2085
  • Country: sk
Re: Soap router - HW specs too good to be true?
« Reply #46 on: May 31, 2014, 10:56:29 am »
ROFLMAO  :-DD :-DD :-DD :-DD :-DD :-DD

those guys are either scamers or extremely optimistic newbies knowing nothing about the things they're planning to do ;) considering all those features they would need a hardware at least an order of magnitude more powerful than today's high-end SOHO routers and several years for developing the software for it.

and who the hell would choose android for a router ? ok. the kernel is a modified linux kernel.. but one of the modification is the network stack which was made kind of restrictive to improve security of the android devices on the network.

most of things what they promise are pretty possible, but on a powerful hardware optimized for routers, not in the device they're presenting.
 

Offline Kean

  • Supporter
  • ****
  • Posts: 2054
  • Country: au
  • Embedded systems & IT consultant
    • Kean Electronics
Re: Soap router - HW specs too good to be true?
« Reply #47 on: May 31, 2014, 01:11:52 pm »
What are those symmetric groups of components you can see four times on the bottom side and once left from the mPCIe slots? Looks like a crystal oscillator with some diodes and capacitors around, but it doesn't make sense to have this five times randomly spawned around the board.
If you're talking about what I think, then in the 3D view they surround a QFP/QFN IC with a large thermal pad - maybe ethernet transceivers or DC/DC converters?  If ethernet, then they are weirdly positioned.  I see 4 very similar on the bottom, and one that could be the same on the top?

Also, at first I suspected the footprints that look like SMD crystals at the edge of the board next to the outer mPCIe slot were for the quad antenna connections, but surely they would come directly off the mPCIe module via u.fl connectors.

Also, I don't see any connector for the LCD and the layout doesn't correspond to the layout of connectors in their previous 3D render (then why did they do it?).
Yeah, where is the LCD connector?  I only see HDMI.  Surely the final prototype has to verify the LCD, if not the other unproven features like the audio and backup battery power.  They also seem to have misplaced some of the other wireless modules previously visible.

Yeah, and the component prices still don't add up to their retail price, that's right. Still won't believe until I see anything real that's really working.
They did say they were subsidising the cost of the KS devices - but why the hell would they do that???  Get a working unit on the market, and you can sell your first batch of super duper routers for a huge premium to all those people who must own the latest and greatest, who cares the cost!  Most likely they'll run out of money getting a partially working prototype and never deliver anything.

And even worse they're now using a third party SOM, which can only be cost effective if they've realised they will only making a few hundred units.
AND YET THEY'RE SELLING THEM AT A LOSS!

They also say
Quote
The SMA connectors are not on the board. We are going to run antenna wire from the case to the WiFi chipset. THis doesn't change anything for you guys but allows our EMI to pass with flying colors
Wow!  Pretty confident aren't they!
Most WiFi routers I've pulled apart do this anyway with short coax cables between the chassis mount RPSMA and PCB.
And this is a completely different connector layout to that of the earlier "production" housing renders, so any emmissions test results on this board aren't particularly relevant to a final production PCBA.

All that said, if they had shown this "prototype" early on in the KS timeline they might have got a few more belivers, rather than showing photos of Arduinos and silly low end ARM boards with faked SATA connections.  I will be very surprised if the iMX6 and Android bloatware environment will ever meet their stated throughput goals, but had I backed this project for more than $1, I'd only be doing it for a nice HA controller.
 

Offline Kean

  • Supporter
  • ****
  • Posts: 2054
  • Country: au
  • Embedded systems & IT consultant
    • Kean Electronics
Re: Soap router - HW specs too good to be true?
« Reply #48 on: May 31, 2014, 01:20:09 pm »
Oh, and I forgot - they've got one of the large electrolytic capacitors placed on top of a bunch of SMD passives (near the HDMI connector).
Maybe just a slip of the mouse during 3D mode in Altium... I know I've done it.
 

Offline madires

  • Super Contributor
  • ***
  • Posts: 7695
  • Country: de
  • A qualified hobbyist ;)
Re: Soap router - HW specs too good to be true?
« Reply #49 on: May 31, 2014, 02:20:56 pm »
I think that the Soap team totally underestimated the project and overestimated their skills. I wouldn't be suprised if their last project was something with an Arduino  >:D
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf