[hpsdr] ATLAS PCB fab status ??

[Ray Anderson]

>From: Eric Ellison [mailto:ecellison at comcast.net] 
>Sent: Tuesday, March 28, 2006 3:29 PM
>To: ray.anderson; 'High Performance Software Defined Radio Discussion
List'
>Subject: RE: [hpsdr] ATLAS PCB fab status ??

>Ray

>You'll have one of the boards. I agree that electrical integrity rather
>than the more sophisticated tests is in order for meshing with the
desire >of folks who have ordered the now "Production Model" is the most
important.
>However, if you notice in your testing that we might include
active/passive
>termination, we could either go with a second alpha or discuss to see
if it
>makes this fundamental product in the chain better than it is. A week
or 2
>turn around for boards is great, and worth the test to avoid the "Blue
or
>Yellow (if you are KD5TFD) obligatory jumpers. I'd rather have the
'added'
>jumper just for show, rather than really needed to make it work!


>Thanks
>Eric

Eric-

I don't think we'll need to make provisions for termination on the PCB
for a couple reasons:

	1) The current intent is to not run any real high-speed fast
rise time signals on the backplane (like LVDS clocks). 

	However if it were desired to do that in lieu of adding an
external coax jumper between boards it could be done without too much
degradation of the LVDS signals by ensuring the driving board is in one
end slot and that the receiving board is in the opposite end slot and
providing the LVDS termination on the receiving board.

	2) The RLC parasitics involved in providing a selectable
termination on the backplane would probably be enough to negate a lot of
the benefits from the termination.

	3) Unless the use of multiple LVDS drivers on the buss is
contemplated I think the scheme of paragraph 1b will be adequate.

	4) The remainder of the currently envisioned signals (I2C, I2S,
SPI, audio, low speed control signals and the like) are slow enough
(from a rise time viewpoint) that the small (5.5 inch) backplane won't
necessitate terminations.

I'll consider this some more, but at this time I'm thinking terminations
on the backplane are probably not something we really want to do.

If anyone has good arguments to the contrary please speak up.

-Ray	WB6TPU





 1143592812.0