Glennster wrote:
>  Hello everyone,
>  I am working on an RF design which has 2 boards, essentially a
>  "sandwich".
> The bottom board has the digital stuff and 18 relays.  The relays
>  have physical connections on the top and bottom of the case (Coil on
>  the bottom and contacts on the top). The top board has all the nasty
>  RF.
>  There is also a 4 pin header connecting the 2 boards together.
>  I have footprinted the relays in 2 parts, coil and contacts.
>  What is the best way to ensure that the 2 boards mate exactly.  When
>  assembled, the top board goes down on top of the relays with the
>  contacts sticking through the PCB and is soldered on the top along
>  with the header.

Don't know if it's too late to offer anything helpful (there have been 
so many great suggestions already), but I hope my own experience 
can be useful to you.

I once made an rf style board with only one relay, that although 
much simpler, had similar problems to solve. There was also a four 
pin header connecting the relay board and the "digital" base-board, 
although "sandwiching" wasn't involved due to the design of the 
relay, and having a single relay made it easier to deal with cross-
talk etc.

What I ended up doing was laying out the relay board first, double 
checking the critical component pad spacings with the actual 
components, then using that layout to produce a library component 
with the tracking stripped off, leaving only the desired mechanical 
and silkscreen layers, plus the critical component and fastener 

The mother board was then layed out around that component, and 
some "flotation" built into the alignment also, to reduce stress on 
the through-header. There were never any problems with the 
alignment of the two boards, even with pairs manufactured 

Its already been mentioned in another post that it's worth 
considering the future maintenance and repair side of a such a 
board and have some socketing somehow, and this is where I was 
caught out.

Myself and others eventually had to repair these boards at the 
component level, either if all the spares were used in the field, or to 
put back into service during maintenance. This caused wear and 
tear on the header pads, which at the time, for reliability reasons in 
a marine enviroment, I had decided to solder. 

Although the pads were oversize, and with a nice large hole, 
eventually even with a good vacuum desoldering machine, boards 
had to be scrapped, or relegated to 2nd class test or debug status.

In hindsight (wonderful thing that!), I would have used a plug/socket 
arrangement, and doubled/quadrupled up on the contacts for each 
signal (it was only digital control + power supply anyway), and 
made life a lot easier at repair time for everyone.

This is not so important if the board is only a one off, or if the whole 
assembly is considered a non-repairable (except in emergency) 

Aligning and trying to place a circuit board over 18 sets of 
component legs at assembly time could be tricky also, but some 
alignment pegs/fasteners, or cutting all the leads short, bar say 
one at each corner of the board could help. 

Sometimes there is no easy way out with mixed 
analogue/digital/mechanical all in the one assembly, and things 
have to be tweaked to suit the whole package, and/or the 
components available/essential to use.

I have seen a nice set up with a programmable attenuator from an 
older style signal generator.

The relays themselves are physically in two parts, a detacheable rf 
contact assembly, and an actuator and coil base. The driving 
circuits are on a board screwed to the relay block, connecting to 
each coil pin via a socket. The driver board can be removed just by 
loosening four or six screws. This isn't quite the same arrangement 
as a mother/daughter board setup but perhaps the idea is 
applicable in some way

Good Luck,

Guy Porritt

NIWA Ltd - National Institute of Water and Atmospheric Research
P.O. Box 14-901, Kilbirnie, Wellington, New Zealand.       
Ph.  ++64 4 383-0300  Fax. ++64 4 386-0574

* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
*  This message sent by: PROTEL EDA USERS MAILING LIST
*  Use the "reply" command in your email program to
*  respond to this message.
*  To unsubscribe from this mailing list use the form at
*  the Association web site. You will need to give the same
*  email address you originally used to subscribe (do not
*  give an alias unless it was used to subscribe).
*  Visit
*  to unsubscribe or to subscribe a new email address.
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *


To leave the EDAFORUM discussion list, send a email with
'leave edaforum' in the body to '[EMAIL PROTECTED]'

More Information :

Reply via email to