Well, if I understand the OP's need, it doesn't matter so long as it works with a USB keyboard. There's no need for supporting multiple devices, no need for supplying full hub power capabilities, etc. They need the physical connector, the electrical physical signaling and power so far as one keyboard is concerned, and the only that higher level protocol support that is specifically required by USB keyboards.
AIUI, anyway.
Jon
No. USB was designed to make the client relatively simple at the cost of making the host's job very tough. The expectation was that PC style device you serve as the host.
I wonder about that USB to PS2 adapter though.
Doesn't work like that. There was supposed to be such a simpler spec in the USB-on-the-go. But both devices have to participate for it to work.
If there were an easy host pare-down, then all of the micro manufacturers would have done it a long time ago.
BAJ
Well, color me stoopid, but I fail to see any reason somebody wants to use a complex USB system, with attendant reliability, protocol, and maintenance problems at both ends, rather than the simple, well tested, readily available PS2 interface for a keyboard.
You will have no problem finding PS2 keyboards for 2 to 3 USD at your local Salvation Army or Goodwill store. Other countries may need to look elsewhere, but they should still be more easily found (and cheaper) than USB keyboards.
-- "If you want to post a followup via groups.google.com, don't use the broken "Reply" link at the bottom of the article. Click on "show options" at the top of the article, then click on the "Reply" at the bottom of the article headers." - Keith Thompson
I can't say I've ever noticed any reliability, protocol or maintenance problems with any USB keyboard I've used. Is that really a problem with USB keyboards? Anyway, my question is not really about the complexity of USB, but about the complexity of using chips (and attendent vendor code) designed to do USB. I'm hoping that the state of the USB-chip art is such that they can now do their complex thing (in particular, their USB keyboard thing) with a minimum of our engineering fuss. I mean, we can put a man on the moon... :-\
As for PS/2 keyboards, I should look into the possibility. Technically they would be an easy solution. What is the likelyhood, however, that they will be available new (please, no Goodwill stores!) say, 10 years from now?
But we would have a problem including them with our product and expecting the customers to send us money, much less expecting them to buy more of our products in the future.
I'd much rather support a PS/2 or standard AT keyboard connector. However, as you may have noticed, the old larger-diameter AT keyboard connector type keyboards are getting harder to find. Especially new. A *lot* of USB keyboards, some PS/2 -- and a few of the PS/2 ones will include one of those adapters in them for the old connector.
Where does one buy an ISA bus system anymore?? And I like them for building protos of my own. I don't have the ability to work with PCI, as reflection wave bus design with 1 and 2 nanosecond clock to data skews, etc., are out of my hobbyist league. Plain and simple.
I guess these things are why I didn't question the OP's direction and instead took it on face value and went with it.
But for how long? Do you know where to go to get an ISA floppy disk controller? Or an ISA hard disk controller? Or one of those multi-I/O controllers for the ISA? I've got them still in boxes here, but there is no way I could source one commercially, these days.
When planning products for long term availability, I can understand someone desiring to look at the current situation and extrapolate a bit to see what is changing and to plan for it. At least, within reasonable bounds.
The question, I suppose, is whether or not planning in a USB keyboard connector is within reason. There probably would be no argument from you or others, were it the case that USB hosting was easy to do. The only reason there is an argument here is simply because there are real risks and so folks quibble over that choice in the context of various alternatives. Ultimately, the OP is going to have to decide. All we can do is kibitz.
Jon
I am glad for a direct answer (well, not so glad myself, but more vicariously for the OP.) But I have a hard time accepting this assertion without the details. Why? It just seems that keyboard support cannot possibly require every single feature of a general purpose USB host. I mean, each and every single one?? Really???
Hmm. Okay, so there was some thought to the idea. Just didn't happen?
Well, there's a point. The only thing that comes to mind to perhaps dispute this is that micro manufacturers don't know how their USB host support is going to be used by a customer, so they *need* to supply a general purpose support in order to reach a broader audience. And restricting the design they offer to some tiny, single purpose would also too narrowly restrict their customer base and make it unprofitable.
Jon
"Mike Silva" wrote
Mike,
I think its hard to look ten years down the road see what's going to be available, USB or not. I don't think that much of anything that was around ten years ago is too easy to find today, EXCEPT for the PS/2 keyboards.
Serial keyboards will certainly be around for another three or more years. If the product does well and you're still in business, and PS/2 keyboards start going out of style, you will still have an opportunity to buy a ton of USB to PS/2 converters, which will buy you plenty of time to redesign your product.
Right now, the PS/2 solution is simple, affordable, reliable and a good business decision. IMHO, USB was designed by the big boys to try and keep the small guy from getting a piece of the pie.
[OT] Are you of any relation to Jose?Mike
I have an IBM PS/2 Keyboard that was manufacured in 1980. So this specific protocol has been around for 25 years. You might have to order from a higher level supplier than Joe's Cheapest Computers down the road. One can buy the better keyboards from industrial suppliers with all sorts of interfaces.
[Snipped]
I think as long as you can supply the neede keyboard as an option, most people should be able to source a cheap PS/2 keyboard without to much problems for the next 10 years. If you force your customers to buy an "expensive" keyboard when they can get a US$2 around the corner will piss them off. Giving them the option keeps everybody happy. Just make it very clear in your spec overview that the keyboard is a stock standard PS/2 keyboard.
Regards Anton Erasmus
I suspect misunderstanding here. PS/2 does *not* mean PlayStation 2 in this context. It is the name of IBM's system introduced in '87:
You are mostly welcome.
Vadim
Someone
Well some crazy guy did something along those lines:
low speed only, connected to a avr.
There is no "x" in my email address.
Well he did a function, not a host, but I still tip my hat to anybody who can bit-bang a USB device. Dang!
Yes, thanks, I did understand the PS/2 reference, but I wasn't sure they were still available after my quick search. Now I see I just need to wade through the PlayStation 2 devices to find the PS/2 devices.
BTW, question for everybody: are all these PS/2 keyboards being sold for replacement use only, or are there still computers being sold that use PS/2 keyboards? If no PS/2-keyboard computers are now being sold, when did they stop selling them and switch completely to USB keyboard computers?
Yes, you're right of course, choosing PS/2 keyboards now does not lock us into them forever. At the same time, at least for my own knowledge, I need to pursue the USB angle a little more just to get a better feel for it. I got'sta know, as the man said.
Nope, afraid not.
Dang, you are right. I had another ...cesko... link that I THOUGHT was a host, but it is now a dead link. I am very aware of the diffence between a host and a device. Sorry for the noise on the newsgroup.
There is no "x" in my email address.
Not sure about branded computers, but every motherboard I looked at still support PS/2 k/b and floppy drive. Although the authority declared them dead some time ago, they are still there.
Not really. Download the USB HID (Human Interface Device) spec, and look up "Boot Protocol." IIRC, it's a minimal USB interface that keyboards must support to allow BIOS keyboard interface.
Regards,
-=Dave
-- Change is inevitable, progress is not.
I'm not into USB just now, but I might just look that up on your recommendation, anyway. But before I do, your last sentence here suggests you are talking from the perspective of what a keyboard must support. Presumably, this is also what the host side must also support? But that makes me wonder if the host side must also provide this (the idea being that although it *may* be a requirement for the keyboard to support it, it may not be the case that keyboards insist that the host go through these steps in order to function -- keep in mind that I don't believe our OP needs to support identification of device/subdevice for driver loading, etc. -- so while all that might be necessary in general for Windows, it isn't necessary as I read it for the OP.)
Can you address yourself to this? Or should I just go look and see for myself, anyway?
Thanks for the comments, by the way.
Jon
I understand how designing a USB host that will support multiple devices like a PC would is a daunting task indeed. But after designing a few USB devices, it would seem a USB Host that only supports a specific device (like a keyboard) wouldn't be incredibly difficult.
Host would have to:
-Wait for device detection (change of state on the D+ and D- lines)
-Read out descriptor from endpoint
-if (Endpoint != Keyboard)
- Ignore
-else
- start polling the endpoint where the keyboard data comes out.
The only BIG problem I see is the physical interface electronics, because no one makes a physical interface only. In addition, no one makes a simple host uC with the physical interface built in. Any thoughts or comments on this?
-J
I understand how designing a USB host that will support multiple devices like a PC would is a daunting task indeed. But after designing a few USB devices, it would seem a USB Host that only supports a specific device (like a keyboard) wouldn't be incredibly difficult.
Host would have to:
-Wait for device detection (change of state on the D+ and D- lines)
-Read out descriptor from endpoint
-if (Endpoint != Keyboard)
- Ignore
-else
- start polling the endpoint where the keyboard data comes out.
The only BIG problem I see is the physical interface electronics, because no one makes a physical interface only. In addition, no one makes a simple host uC with the physical interface built in. Any thoughts or comments on this?
-J
How about the makers of pxa270, at91rm9200 and ep9315? Are they nobody making no USB hosted micro?
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.