[Hpsdr] Call for Comments - ALEX

[Graham Haddock]

Chris:

Comments below.

--- Graham / KE9H

==
Chris Albertson wrote:
> The Eurocard size looked good to me too
> they fit inside those extruded aluminum boxes.  You could loose
> the 10 standoffs and use the card guides in the extruded box to hold
> the cards.   The box is completely RF tight.  You could mount BNC
> connectors on the box's end panel and then use BNC jumpers to route
> the signals to user supplied connectors on his HPSDR enclose.
> Use PC mounted BNC connectors that protrude through holes drilled
> in the end panels.  There is no need for the "shield board" if the
> rx and tx boards go inside a box.
>
> These boxes work well and don't add much cost when you concider
> you will not need the shield card.  
> http://www.nutech-eng.com/english/enclosurecases/cases/product/e006.asp
>   
There is a USA supplier of what appears to be the identical box as you 
describe.
(Hammond Manufacturing; stocked at Mouser.)
http://www.hammondmfg.com/1455.htm
I am currently packaging a homebrew CW transmitter in this case. 
Although the
Hammond literature doesn't show it clearly, I think they are identical, 
including
the removable top panel, if you ignore the Hammond plastic end gasket.

I'll keep this as a backup plan, if the three board sandwich is not 
enough shielding.
Since we are using toroids, they are somewhat self-shielding.
As you say, the boards as designed will just slide in to this size case.

It could be a real easy user packaging option.  Hammond sells
nice mounting brackets and other accessories for the case, also.

Thanks for the idea.

> Is the TR Relay fast enough to keep up with someone sending
> CW at a high rate?  
>   
The T-R relay I plan to use, Omron G6J-Y Series, is rated at 3 ms. max 
(<2 ms. typical)
switching time, including bounce, and it is an RF rated relay. There 
will be a pair of 1 Amp
contacts in parallel. Since the recommended rise and fall times for CW 
transmissions is
about 5 ms., I think this will work fine.
Switching command time on the SPI bus will be tens of microseconds.
The relay should not be "hot-switched" with 100 watts on it, but there 
is nothing to
say that you can't listen during part of the the actuation and bounce 
time.  So, with
dumb software, you will loose 3 ms at the start and end of each 
element.  With clever
software, you could reduce this to 1.6 ms mute at the start and end of 
each element,
even though you are driving the relay with 3 ms. switching allowance.  
Either should
give nice QSK, and be equal to or better than the reed relays Ten-Tec 
used in their
early transceivers that made them famous for good QSK.

> It would be good if these were designed so it could be hand soldered.
> A 0.05 inch lead pitch is easy for most people
>   
The Hittite HMC226 switches on the Receiver board are 0.037 inch (0.95 
mm) pitch, and there are
about 16 of them.  Everything else can be 0.05 inch pitch.  Since there 
are so many toroids, I was
going to make the relays thru-hole, also.

--- Graham / KE9H




 1185379922.0