viswesr writes:
Cypress has a new USB 3.0 peripheral controller EZ-USB FX3 – CYUSB3014 with ARM926EJS core. It has an architecture which enables data transfers of 320 MBps from GPIF II (a 32bit Port) to USB 3.0 interface.This value includes protocol overheads. This is a promising device for Open Logic analyzers, DSOs and any data acquisition system based on USB 3.0 streaming interface. The associated promo video mentions about support of ARM GCC toolchain, ideal for Open Hardware project. With availability of PCIe USB 3.0 expansion cards costing just $22, we could easily use USB 3.0 devices.
The downside :
1) at present they are only sampling with FBGA footprint, might require 4 layer PCB :( . 2) Require costly USB VID 3)No price info is available
Via the contact form.
Yeah, saw a note about this in Elector a couple of days ago and had a look at the specs. It’s an 11×11 0.8mm pitch BGA package, which might just be within the limits of what can be achieved on Laen’s 4-layer PCB orders, though I’m not sure anyone’s really done finer than 1.0mm BGA on that.
Other neat features include USB 2.0 OTG high-speed support (possibly useful for interfacing to Android phones?) and charger detection support. I’m guessing Cypress are targetting battery-powered portable applications for this again.
As far as the VID thing goes, I’ve often thought about this, especially after reading the old LUFA blogs, where he talks about the absurdity of the USB VID situation.
I really wonder what would be the downside of the open hardware/hacker community just hijacking a couple VID’s and dolling out PID’s. If you don’t advertise your product as usb-compliant, I don’t think there’s any real legal problems, and I don’t think they can sue you for letting people reserve a 16 bit number. No hardware vendor is going to want those VID’s eventually anyway, so there’s not much worry about collision with legit products, either.
I guess somebody needs to take a risk on setting up a site initially though. Someone outside the US I expect.
In fact they do sue, or threaten too, even when there is no legal basis (some Dutch guys experienced this). Now there is a long contract that spells out penalties for stuff like this :(
But they were a legit company with a purchased VID that were reselling PID’s, weren’t they? So they are bound by some contract. As you said I think the original contract was less carefully worded, so some guys are (or were) getting away with reselling PIDs.
I’m just talking about picking an unused one at random and dolling out ‘magic 16 bit numbers with no guarantees’. I think whoever ran such a site would be pretty safe but if someone tried to sell commercial products using them they could get in trouble if they even mention ‘USB’.
So again I’m only talking about for free/open stuff.
There was no contract at the beginning, but USB forum started harassing and suing individual and business resellers anyways.
I have also considered squatting an (already abandoned?) USB ID. As long as you never sign a contract with them I doubt there is much they can do. Given their history though, you might need a lawyer anyways to reiterate that fact for you.
Yeah theres no licenses for the ICs or open software stacks, and a VID is just a number, so the only thing they have actual control over is the trademarks, and these days the term usb is so ubiquitous that probably only applies to the actual certified compatible logos.
But yeah if you’re shipping product with a vid you dont own expect to talk to a lawyer eventually. But empty hardware with firmware available online… there is tons of usb “example” code out there using random vids as is.
Cypress development boards tend to be a little bit (or hella-) pricey, so getting started can be difficult/expensive. If it goes like the PSoC parts did, getting a development board with engineering samples on it is the only way that you can get parts any parts whatsoever until they go into production — sometimes a year after the parts are announced. My guess is that samples for parts like these are only going to go to the bigger players for a while.
They definitely do some interesting work though.
If the 320 MBps value includes protocol overheads, then what is the actual payload bandwidth? I know where to look up the overhead for USB 2.0, but I haven’t bothered to investigate USB 3.0 because it seems like such a kludge of a specification.
> because it seems like such a kludge of a specification.
Backwards-compatibility sometimes isn’t pretty, but it damn sure pays the bills.
I don’t know if “backwards-compatible” is entirely accurate. As far as I understand, USB3 is nothing more than a cable with USB2 and eSATA/PCIe crammed into the same sheath. That’s more like squeezing two different standards into one and calling it “new.” Specifically, I’ve heard that the USB2 part is exactly USB2, not anything more.
If you’re already running out of USB ports on your laptop, then imagine if your eSATA and USB ports were rolled into one – then plugging either type of device will block “both” ports.
In two words; “yes please”. It has a frickin ARM9 in there! No more 8052! Yay!
It’s Mr Peripheral.